From: Thomas Walker Lynch Date: Tue, 10 Jun 2025 07:34:10 +0000 (-0700) Subject: refactoring due to better understanding of bump buffer X-Git-Url: https://git.reasoningtechnology.com/usr/lib/python2.7/encodings/mac_greek.py?a=commitdiff_plain;h=81916b3f78ab08956cca368d9563fd910920e4cb;p=RT-gcc refactoring due to better understanding of bump buffer --- diff --git a/README.md b/README.md index 4f884c4..dab8163 100644 --- a/README.md +++ b/README.md @@ -1,139 +1,54 @@ # RT CPP Extensions and GCC Toolchain -> *Status: Scripts build a standalone GCC 12.4.1 on Debian 12.10 with an option for RT's CPP extensions. Directives #macro and #assign are included. The builtin macros are being incorporated, and will hopefully be released soon.* - -This repository provides a standalone GCC installation toolkit with optional support for the RT extensions to the C preprocessor. It includes reusable build scripts, macro programming infrastructure, and documentation for the extended CPP environment. - -The RT extensions are designed to modernize CPP with structural programming primitives, support for token/argument list transformations, and controlled macro evaluation. They enable the use of type templates and reusable include files in plain C, without altering standard CPP behavior. - ---- - -## Repository Structure - -* `developer/` — Core documentation, macro definitions, experiments, and examples -* `script_Deb-12.10_gcc-12.4.1🖉/` — Working scripts for building GCC 12.4.1 on Debian 12.10, with or without the RT extensions -* `script_gcc-15🖉/` — Incomplete scripts for GCC 15.x (experimental) -* `LICENSE.text` — MIT License +> **Status:** This project builds a standalone GCC 12.4.1 on Debian 12.10, with optional support for the RT Extensions to the C Preprocessor. Scripts are modular, reproducible, and under active refinement. --- -## How to Build - -To begin a build, first set up your environment: - -```bash -. env_developer -``` - -Each script directory contains its own `README.org` and a set of Bash scripts that can be run in order. They are idempotent and reflect the actual command sequence used during the build. - ---- - -## About the RT Extensions - -The RT CPP extensions introduce two new macro directives: - -### `#macro` - -A multiline, ISO-style function macro with literal token bodies and a strict parameter list. Enables readable macro logic that spans lines without escape characters. - -### `#assign` - -Similar behavior to a non-function, declarative, form of `#define`. It has both a directive and macro form. In macro form it expands to nothing. When brackets are used instead of parenthesis, the clauses are expanded before the named macro is defined. - -```c -#assign (FOO) (42) // literal -_ASSIGN [BAR] [BAZ] // expanded name and body -``` - -Hence in this second case, `BAR` will be expanded and must result in an identifier name, if it is not macro, then it is taken literally. Also `BAZ` will be expanded if it is a macro, and the result will be the body of the `[BAZ]` macro. - -## Built-in Macros +## Overview -RT introduces functional built-in macros to extend CPP’s token manipulation capabilities. All built-ins are prefixed with `_` and operate without side effects, except `_ASSIGN`. +This repository provides a structured toolchain for building GCC with optional support for the RT Extensions to the C Preprocessor (CPP). These extensions modernize the C macro system to support reusable structural templates, controlled evaluation, and functional-style transformations — all within the semantics of standard C. -### _ASSIGN - -A macro form of `#assign`. Takes two *clauses* — either parenthesized for literal input, or bracketed for expanded input. - -```c - _ASSIGN (FOO) (42) // literal name and body - _ASSIGN [BAR] [BAZ] // both name and body expanded -``` - -- ( ... ) ⇒ literal clause (no expansion) -- [ ... ] ⇒ expanded clause (macros are recursively expanded) - -This expands to nothing but defines (or overwrites) a macro binding. New macros created via _ASSIGN are *colored* (i.e., inert within the macro that created them). +The project is designed for clarity and reproducibility. Each script series targets a specific GCC version and platform, enabling consistent builds for both development and experimentation. --- -### List Conversions +## RT CPP Extensions -These help translate between CPP’s two list types: space-separated `token_list` and comma-separated `argument_list`. +The RT Extensions introduce new preprocessor directives and built-in macros that enable: -```c - _TO_ARG_LIST(tokens...) // a b c → a,b,c - _TO_TOKEN_LIST(a,b,c) // a,b,c → a b c -``` +- Multiline macros and declarative definitions +- Token and argument list transformations +- Structural and functional macro programming +- Compatibility with standard CPP behavior ---- +These capabilities allow type templates, associative sets, conditional logic, and macro-driven abstraction layers — all expressed in standard C headers. -### Token List Operators - -```c - _FIRST(tokens...) // first token - _REST(tokens...) // all but first - _MAP(f, tokens...) // f(a) f(b) f(c) - _AL_MAP(f, a1, a2, ...) // f(a1) f(a2) f(a3) -``` - -- _MAP operates on a space-separated token list. -- _AL_MAP operates on a comma-separated argument list. +Detailed documentation is located in the `developer/` directory. --- -### Logic Primitives - -These treat token presence as boolean truth. - -```c - _IF(p, a, b) // if p is non-empty → a else → b - _NOT(p) // if p is empty → TRUE else → - _AND(a1, a2, ...) // all must be non-empty - _OR(a1, a2, ...) // returns first non-empty -``` +## Repository Structure -Notes: -- _AND() expands to nothing if any argument is empty. -- _OR() short-circuits to the first non-empty argument. +- `developer/` — Core documentation, macro definitions, experiments, and examples +- `script_Deb-12.10_gcc-12.4.1🖉/` — Working scripts for building GCC 12.4.1 on Debian 12.10 +- `script_gcc-15🖉/` — Preliminary scripts for GCC 15.x (experimental) +- `LICENSE.text` — MIT License --- -### Token-Level Predicate - -```c - _IS_IDENTIFIER(x) // if x is a valid identifier → x else → -``` - ---- +## Getting Started -### Token Construction +To configure your shell environment for building and testing: +```bash +. env_developer ``` - _PASTE(a, b, c) // pastes into single token -``` - -All tokens are concatenated into one; result must be a valid CPP token (e.g., identifier). Invalid pastes result in a compiler error. ---- - -These primitives enable set manipulation, conditional macros, structural mapping, and functional programming patterns — within standard C preprocessor constraints. +Each script directory contains its own `README.org` and a set of executable scripts representing the build sequence. These scripts are idempotent and document the actual commands used to construct the toolchain. --- ## License -This project is licensed under the **MIT License**. -See the `LICENSE.text` file for full terms. - +This project is licensed under the **MIT License**. +See `LICENSE.text` for full terms. diff --git "a/developer/document\360\237\226\211/.gitignore" "b/developer/document\360\237\226\211/.gitignore" new file mode 100644 index 0000000..8e44d75 --- /dev/null +++ "b/developer/document\360\237\226\211/.gitignore" @@ -0,0 +1,3 @@ + +*.html +*.pdf diff --git "a/developer/document\360\237\226\211/doc_draft.html" "b/developer/document\360\237\226\211/doc_draft.html" deleted file mode 100644 index 4a14063..0000000 --- "a/developer/document\360\237\226\211/doc_draft.html" +++ /dev/null @@ -1,534 +0,0 @@ - - - - - - - -RT Extensions to the C Preprocessor - - - - - -
-

RT Extensions to the C Preprocessor

- -
-

1. Overview

-
-

-The RT extensions modernize the C preprocessor (CPP) to support structural programming primitives, sets, associative maps, and token/argument list manipulation. -

- -

-These extensions are intended to support C include files that are type templated, and can be included multiple times with different type template bindings. However, these functions add flexibility to CPP that will be appreciated in many contexts. -

- -

-The RT extensions do not change existing CPP behavior. -

- -

-The RT extensions also do not facilitate recursive programming. During a given evaluation, expanded macros are colored, and thereafter their names are taken literally. On a conciliatory note, among the new built-ins is a map function. -

-
-
- -
-

2. Review of Conventional CPP

-
-

-CPP has two explicit list types: -

- -
    -
  • `tokenlist`: A stream of tokens as produced by the lexer.
    • -
  • `argumentlist`: A comma-separated list of `tokenlist` elements.
    • -
- -

-Although CPP has these two types of lists, there are few operators in the language for supporting this structure. Hence, the RT extensions contain some. -

- -

-CPP also supports macro binding via `#define`, which has two major forms: object-like macros and function-like macros. However, `#define` is limited in how it can manage evaluation control, bind late, or perform structural reasoning. The RT extensions provide two new directives to address these limitations. -

-
-
- - -
-

3. `#macro`

-
-

-The `#macro` directive defines a function-style macro with named parameters and a token body enclosed in parentheses. The parameter list is required, but may be empty. -

- -
-
#macro NAME(arg1 ,arg2 ,...) ( token sequence )
-
-
-
- -
-

3.1. BNF for `#macro`

-
-
-
directive     ::= "#macro" name params body ;
-
-name          ::= identifier ;
-
-params        ::= "(" param_list? ")" ;
-param_list    ::= identifier ("," identifier)* ;
-
-body          ::= paren_clause ;
-paren_clause  ::= "(" literal? ")" ;
-
-literal       ::= ; sequence parsed into tokens without expansion
-
-; whitespace, including newlines, is ignored
-
-
- -

-Unlike `#define`, which requires the entire macro to reside on a single line, `#macro` allows its body to span multiple lines with no end-of-line escapes being required. -

- -

-As with `#define`, the `#macro` name is not evaluated, and the body is stored as a literal token sequence and is not expanded at the time of definition. -

- -

-The body-delimiting parentheses need to balance only at the time the definition is scanned by the lexer. If the programmer desires to include unbalanced parentheses within the body, then they should first create a macro such as `#define OPEN (` or `#define CLOSE )`, and then use those in the macro body. -

-
-
-
- - -
-

4. `#assign`

-
-

-The `#define` directive has both a declarative and function form. The `#assign` variation of `#define` does not have a function form. -

- -

-With `#define`, neither the name nor the body are expanded at time of definition. With `#assign`, the name and/or the body can optionally be expanded by using brackets for the respective clause, rather than parentheses. -

- -
-
#assign (literal NAME) (literal body)
-#assign (literal NAME) [expanded body]
-#assign [expanded NAME] (literal body)
-#assign [expanded NAME] [expanded body]
-
-
- -

-As with `#macro`, the clauses in `#assign` can be multi-line. The clauses are contained within the specified delimiters, which must balance at the time of definition, when the lexer scans them. This happens before any expansion. -

-
- -
-

4.1. BNF for `#assign`

-
-
-
cmd        ::= "#assign" name body ;
-
-name       ::= clause ;
-body       ::= clause ;
-
-clause     ::= "(" literal? ")" | "[" expr? "]" ;
-
-literal    ::= ; sequence parsed into tokens
-expr       ::= ; sequence parsed into tokens with recursive expansion of each token
-
-; white space, including new lines, is ignored.
-
-
- -

-The name clause must resolve to a valid name identifier. -

- -

-There is a corresponding built-in macro form `ASSIGN name body` which expands to nothing, but has the side effect of creating a definition. -

-
-
-
- -
-

5. New Built-in Macros

-
-

-Each new built-in macro has a `_` prefix. -

-
- -
-

5.1. `ASSIGN`

-
-

-The same syntax and evaluation semantics as for `#assign` apply. -

- -
-
_ASSIGN (literal NAME) (literal body)
-_ASSIGN (literal NAME) [expanded body]
-_ASSIGN [expanded NAME] (literal body)
-_ASSIGN [expanded NAME] [expanded body]
-
-
- -

-The `ASSIGN` macro mirrors the behavior of the `#assign` directive and, unlike other macros, is followed by two delineated clauses. Each clause may be parenthesized (for a literal) or bracketed (for an expanded expression). -

- -

-The `ASSIGN` macro expands to nothing. Its only effect is to create or update a macro binding. -

- -

-When a macro is created with _ASSIGN, it is colored. Hence, the newly created macro can not be expanded within the same macro where it was defined. -

-
-
-
- - -
-

6. List Transformations

-
-
-
_TO_ARG_LIST( <token_list> )      // Converts tokens into an argument list
-_TO_TOKEN_LIST( <argument_list> ) // Flattens argument list into token stream
-
-
- -

-These form the basis for transitioning between structure-aware and structure-neutral representations. -

-
-
- -
-

7. Token List Operators

-
-
-
_FIRST( <token_list> )            // First token
-_REST( <token_list> )             // All tokens after the first
-_MAP(f ,<token_list>)             // f(token1) f(token2) ...
-_AL_MAP(f ,<argument_list>)       // f(arg1) f(arg2) ...
-
-
- -

-These enable iteration-like behavior over static macro arguments. -

-
-
- -
-

8. Logic Primitives

-
-
-
_IF(p ,a ,b)                      // If p is present, expand to a; else b
-_NOT(p)                           // Expand to TRUE if p is empty, else empty
-
-
- -

-Logical presence is defined by token existence, not token content. Empty argument lists are `false`. All others are `true`. -

- -
-
_AND( <arg1> ,<arg2> ,... )       // Returns nothing if any arg is empty
-_OR( <arg1> ,<arg2> ,... )        // Returns first non-empty argument
-
-
- -

-Missing trailing arguments are not detected by `AND`. Appending a `TRUE` token is recommended to guard edge cases. -

-
-
- -
-

9. Token-Level Checks

-
-
-
_IS_IDENTIFIER( <arg> )          // If arg is an identifier, expands to it; else nothing
-
-
-
-
- -
-

10. Pasting and Construction

-
-
-
_PASTE( <token1> ,<token2> ,... ) // Combines all tokens into a single token
-
-
- -

-Pasting is only valid if the result is a valid preprocessing token (e.g., identifier). Invalid pastes will trigger compiler diagnostics. -

-
-
- -
-

11. Usage Example

-
-

-(To be added) -

-
-
- -
-

12. Sets

-
-

-Sets are macros whose membership is defined by naming convention. All set operations are evaluated through presence of a named binding. -

- -
-
#define _SET_ADD(set ,x) ASSIGN( _PASTE(__SET_ ,set ,_ ,x), )
-#define _SET_IN(set ,x)  _NOT(_PASTE(__SET_ ,set ,_ ,x))
-
-
- -

-A macro such as `_SETmyset_foo` existing in the preprocessor environment marks `foo` as a member of `myset`. -

-
-
- -
-

13. Associative Sets (ASet)

-
-

-Associative sets allow storing values indexed by a key: -

- -
-
#assign _ASET_ADD(set ,x ,y) (
-  ASSIGN(_PASTE(__ASET_ ,set ,_ ,x))
-  ASSIGN(_PASTE(__ASET_ITEM_ ,set ,_ ,x))
-)
-
-#assign _ASET_GET(set ,x) (
-  _IF(
-    _PASTE(__ASET_ ,set ,_ ,x)
-    ,_PASTE(__ASET_ITEM_ ,set ,_ ,x)
-    ,
-  )
-)
-
-
- -

-If `x` is in the set, `ASETGET` returns the associated value; else nothing. -

-
-
- -
-

14. (End of manual — more modules to be added as completed.)

-
-
-
-

Author: Thomas Walker Lynch

-

Created: 2025-06-06 Fri 03:00

-

Validate

-
- - diff --git "a/developer/document\360\237\226\211/doc_draft.org" "b/developer/document\360\237\226\211/doc_draft.org" deleted file mode 100644 index 53e1281..0000000 --- "a/developer/document\360\237\226\211/doc_draft.org" +++ /dev/null @@ -1,203 +0,0 @@ -#+TITLE: RT Extensions to the C Preprocessor -#+AUTHOR: Thomas Walker Lynch -#+OPTIONS: toc:nil -#+OPTIONS: ^:nil - -* Overview - -The RT extensions modernize the C preprocessor (CPP) to support structural programming primitives, sets, associative maps, and token/argument list manipulation. - -These extensions are intended to support C include files that are type templated, and can be included multiple times with different type template bindings. However, these functions add flexibility to CPP that will be appreciated in many contexts. - -The RT extensions do not change existing CPP behavior. - -The RT extensions also do not facilitate recursive programming. During a given evaluation, expanded macros are colored, and thereafter their names are taken literally. On a conciliatory note, among the new built-ins is a map function. - -* Review of Conventional CPP - -CPP has two explicit list types: - -- `token_list`: A stream of tokens as produced by the lexer. -- `argument_list`: A comma-separated list of `token_list` elements. - -Although CPP has these two types of lists, there are few operators in the language for supporting this structure. Hence, the RT extensions contain some. - -CPP's `#define` is limited in that the name and body are not evaluated, and that it an entire definition must be on line line. The RT extensions provide two new directives to address these limitations. - -* `#macro` - -The `#macro` directive defines a function-style macro with named parameters and a token body enclosed in balanced parentheses. The parameter list is required, but may be empty. - -#+BEGIN_SRC c -#macro NAME(arg1 ,arg2 ,...) ( token sequence ) -#+END_SRC - -** BNF for `#macro` - -#+BEGIN_SRC bnf -directive ::= "#macro" name params body ; - -name ::= identifier ; - -params ::= "(" param_list? ")" ; -param_list ::= identifier ("," identifier)* ; - -body ::= paren_clause ; -paren_clause ::= "(" literal? ")" ; - -literal ::= ; sequence parsed into tokens without expansion - -; whitespace, including newlines, is ignored -#+END_SRC - -Unlike `#define`, which requires the entire macro to reside on a single line, `#macro` allows its body clause to span multiple lines with no end-of-line escapes being required. - -As with `#define`, the `#macro` name is not evaluated, and the body is stored as a literal token sequence that will only be expanded when the macro is called. - -The body-delimiting parentheses need to balance only at the time the definition is scanned by the lexer. If the programmer desires to include unbalanced parentheses within the body, then the programmer should first create a macro such as `#define OPEN (` or `#define CLOSE )`, and then use those in the body clause. - - -* `#assign` - -The `#define` directive has both a declarative and function form. The `#assign` variation of `#define` does not have a function form. - -With `#define`, neither the name nor the body are expanded at time of definition. With `#assign`, the name and/or the body can optionally be expanded by using brackets for the respective clause, rather than parentheses. - -#+BEGIN_SRC c -#assign (literal NAME) (literal body) -#assign (literal NAME) [expanded body] -#assign [expanded NAME] (literal body) -#assign [expanded NAME] [expanded body] -#+END_SRC - -As with `#macro`, the clauses in `#assign` can be multi-line. The clauses are contained within the specified delimiters, which must balance at the time of definition, when the lexer scans them. This happens before any expansion. - -** BNF for `#assign` - -#+BEGIN_SRC bnf -cmd ::= "#assign" name body ; - -name ::= clause ; -body ::= clause ; - -clause ::= "(" literal? ")" | "[" expr? "]" ; - -literal ::= ; sequence parsed into tokens -expr ::= ; sequence parsed into tokens with recursive expansion of each token - -; white space, including new lines, is ignored. -#+END_SRC - -The name clause must resolve to a valid name identifier. - -There is a corresponding built-in macro form `_ASSIGN name body` which expands to nothing, but has the side effect of creating a definition. - -* New Built-in Macros - -Each new built-in macro has a `_` prefix. - -** `_ASSIGN` - -The same syntax and evaluation semantics as for `#assign` apply. - -#+BEGIN_SRC c -_ASSIGN (literal NAME) (literal body) -_ASSIGN (literal NAME) [expanded body] -_ASSIGN [expanded NAME] (literal body) -_ASSIGN [expanded NAME] [expanded body] -#+END_SRC - -The `_ASSIGN` macro has the same behavior as the `#assign` directive and, unlike other macros, is followed by, not one, but two delineated clauses. Each clause may be parenthesized (for a literal) or bracketed (for an expanded). - -The `_ASSIGN` macro expands to nothing. Its only effect is to create or update a macro binding. - -When a macro is created with `_ASSIGN`, it is colored. Hence, the newly created macro cannot be expanded within the same macro where it was defined. - - -* List Transformations - -#+BEGIN_SRC c -_TO_ARG_LIST( ) // Converts tokens into an argument list -_TO_TOKEN_LIST( ) // Flattens argument list into token stream -#+END_SRC - -These form the basis for transitioning between structure-aware and structure-neutral representations. - -* Token List Operators - -#+BEGIN_SRC c -_FIRST( ) // First token -_REST( ) // All tokens after the first -_MAP(f ,) // f(token1) f(token2) ... -_AL_MAP(f ,) // f(arg1) f(arg2) ... -#+END_SRC - -These enable iteration-like behavior over static macro arguments. - -* Logic Primitives - -#+BEGIN_SRC c -_IF(p ,a ,b) // If p is present, expand to a; else b -_NOT(p) // Expand to TRUE if p is empty, else empty -#+END_SRC - -Logical presence is defined by *token existence*, not token content. Empty argument lists are `false`. All others are `true`. - -#+BEGIN_SRC c -_AND( , ,... ) // Returns nothing if any arg is empty -_OR( , ,... ) // Returns first non-empty argument -#+END_SRC - -Missing trailing arguments are *not* detected by `_AND`. Appending a `TRUE` token is recommended to guard edge cases. - -* Token-Level Checks - -#+BEGIN_SRC c -_IS_IDENTIFIER( ) // If arg is an identifier, expands to it; else nothing -#+END_SRC - -* Pasting and Construction - -#+BEGIN_SRC c -_PASTE( , ,... ) // Combines all tokens into a single token -#+END_SRC - -Pasting is only valid if the result is a valid preprocessing token (e.g., identifier). Invalid pastes will trigger compiler diagnostics. - -* Usage Example - -(To be added) - -* Sets - -Sets are macros whose membership is defined by naming convention. All set operations are evaluated through presence of a named binding. - -#+BEGIN_SRC c -#define _SET_ADD(set ,x) ASSIGN( _PASTE(__SET_ ,set ,_ ,x), ) -#define _SET_IN(set ,x) _NOT(_PASTE(__SET_ ,set ,_ ,x)) -#+END_SRC - -A macro such as `__SET_myset__foo` existing in the preprocessor environment marks `foo` as a member of `myset`. - -* Associative Sets (ASet) - -Associative sets allow storing values indexed by a key: - -#+BEGIN_SRC c -#assign _ASET_ADD(set ,x ,y) ( - ASSIGN(_PASTE(__ASET_ ,set ,_ ,x)) - ASSIGN(_PASTE(__ASET_ITEM_ ,set ,_ ,x)) -) - -#assign _ASET_GET(set ,x) ( - _IF( - _PASTE(__ASET_ ,set ,_ ,x) - ,_PASTE(__ASET_ITEM_ ,set ,_ ,x) - , - ) -) -#+END_SRC - -If `x` is in the set, `_ASET_GET` returns the associated value; else nothing. - -* (End of manual — more modules to be added as completed.) diff --git "a/developer/document\360\237\226\211/notes.org" "b/developer/document\360\237\226\211/notes.org" deleted file mode 100644 index 0cc91bc..0000000 --- "a/developer/document\360\237\226\211/notes.org" +++ /dev/null @@ -1,113 +0,0 @@ - - - -#There are two list structures explicitly supported by CPP: -* token list -* argument list - -A tokens are gotten from the lexer, hence the token list is a data structure. Each token has a type, such as identifier or punctuation. - -An argument list is a common separated list of token lists. Though sometimes special requirements are made, such that a list member be single token, or a single token of identifier type. - -#Operations to support list structures - -* _TO_ARG_LIST( ) -* _TO_TOKEN_LIST( ) - -#Token List operators - -* _FIRST( ) -* _REST( ) -* _MAP( f , ) -* _AL_MAP( f , ) - -`_FIRST` and `_REST` expand as expected. - -`_MAP` expands as `f(token) ...` -`_AL_MAP` expands to `f(arg) ...` - -We also add: - -* _IF( p ,a ,b) - -_IF can be implemented with #define. However, we still provide a built-in version. - -If `p` exists, expands to `a`, otherwise expands to `b`. Any of the arguments can be missing, in which case they are taken to be false. Note, - -We build this in: - -* _NOT( p ) - -Which is the same as: - -``` -#define NOT(x) _IF(x ,,TRUE) -``` - -If `p` is a token, a token list, or an argument list, `_NOT( p )` expands to nothing. Otherwise it expands to `TRUE`. This is the only operator among those discussed here that introduces a token not already found among its inputs. Here `TRUE` is a non-identifying statement of existence. - -As false is the absence of anything, logical connectors such as AND and OR can not be defined for token lists. However, they can be defined for argument lists, as argument lists can represent the absence of an argument. Hence we have built-in: - -* _AND( ) -* _OR( ) - -`_AND` expands to nothing if there are any absent members of the argument list. There is an interesting end case however, a missing last member will not be detected. A programmer can append TRUE to the end of a list to avoid this effect. - -`_OR` expands to the first argument that is present in the argument list. If there is no first argument, then it expands to nothing. There is no problem with missing argument off the end of the list. - -#Paste connects tokens to create a single token. - -Only Identifier tokens can be pasted. Giving a non-identifer token results the error: - -``` -paste_1.c:9:46: error: pasting "x" and "y" does not give a valid preprocessing token -``` - -The CPP paste operator is found between tokens. There is no operator for pasting all the tokens in a list. So we add: - -* _PASTE( ) - -#Token operators - -* _IS_IDENTIFIER( ) - -If is an argument list, or a tokens list, or the token is not an identifier, expands to nothing. Otherwise expands to the given identifier. - -#Set - -``` -#define _SET_ADD(set ,x) ASSIGN( _PASTE( __SET_ ,set ,_ ,x), ) -``` - -``` -#define _SET_IN(set ,x) _NOT(_PASTE( __SET_ ,set ,_ ,x)) -``` - -If the member is in the set, expands to TRUE, otherwise expands to nothing. - -#ASet - - -``` -#assign _ASET_ADD(set ,x ,y) ( - ASSIGN( _PASTE( __ASET_ ,set ,_ ,x )) - ASSIGN( _PASTE( __ASET_ITEM_ ,set ,_ ,x)) - ) -``` - -``` -#assign _ASET_GET(set ,x) ( - _IF( - _PASTE( __ASET_ ,set ,_ ,x) - , _PASTE( __ASET_ITEM ,set ,_ ,x) - , - ) -``` - - -#List - -``` -#define _LIST_CONNECT(list ,node ,x) ASSIGN( _PASTE( __LIST_ ,list ,_ ,x), ) -#define _LIST_NEXT(list ,x) ASSIGN( _PASTE( __LIST_ ,list ,_ ,x), ) -``` diff --git "a/developer/document\360\237\226\211/user_manual.org" "b/developer/document\360\237\226\211/user_manual.org" new file mode 100644 index 0000000..69f9450 --- /dev/null +++ "b/developer/document\360\237\226\211/user_manual.org" @@ -0,0 +1,204 @@ +#+TITLE: RT Extensions to the C Preprocessor +#+AUTHOR: Thomas Walker Lynch +#+OPTIONS: toc:nil +#+OPTIONS: ^:nil + +* Overview + +The RT extensions modernize the C preprocessor (CPP) to support structural programming primitives, sets, associative maps, and token/argument list manipulation. + +These extensions are intended to support C include files that are type templated, and can be included multiple times with different type template bindings. However, these functions add flexibility to CPP that will be appreciated in many contexts. + +The RT extensions do not change existing CPP behavior. + +The RT extensions also do not facilitate recursive programming. During a given evaluation, expanded macros are colored, and thereafter their names are taken literally. On a conciliatory note, among the new built-ins is a map function. + +* Review of Conventional CPP + +CPP has two explicit list types: + +- `token_list`: A stream of tokens as produced by the lexer. +- `argument_list`: A comma-separated list of `token_list` elements. + +Although CPP has these two types of lists, there are few operators in the language for supporting this structure. Hence, the RT extensions contain some. + +CPP's `#define` is limited in that the name and body are not evaluated, and that it an entire definition must be on line line. The RT extensions provide two new directives to address these limitations. + +* `#macro` + +The `#macro` directive defines a function-style macro with named parameters and a token body enclosed in balanced parentheses. The parameter list is required, but may be empty. + +#+BEGIN_SRC c +#macro NAME(arg1 ,arg2 ,...) ( token sequence ) +#+END_SRC + +** BNF for `#macro` + +#+BEGIN_SRC bnf +directive ::= "#macro" name params body ; + +name ::= identifier ; + +params ::= "(" param_list? ")" ; +param_list ::= identifier ("," identifier)* ; + +body ::= paren_clause ; +paren_clause ::= "(" literal? ")" ; + +literal ::= ; sequence parsed into tokens without expansion + +; whitespace, including newlines, is ignored +#+END_SRC + +Unlike `#define`, which requires the entire macro to reside on a single line, `#macro` allows its body clause to span multiple lines with no end-of-line escapes being required. + +As with `#define`, the `#macro` name is not evaluated, and the body is stored as a literal token sequence that will only be expanded when the macro is called. + +The body-delimiting parentheses need to balance only at the time the definition is scanned by the lexer. If the programmer desires to include unbalanced parentheses within the body, then the programmer should first create a macro such as `#define OPEN (` or `#define CLOSE )`, and then use those in the body clause. + + +* `#assign` + +The `#define` directive has both a declarative and function form. The `#assign` variation of `#define` does not have a function form. + +With `#define`, neither the name nor the body are expanded at time of definition. With `#assign`, the name and/or the body can optionally be expanded by using brackets for the respective clause, rather than parentheses. + +#+BEGIN_SRC c +#assign (literal NAME) (literal body) +#assign (literal NAME) [expanded body] +#assign [expanded NAME] (literal body) +#assign [expanded NAME] [expanded body] +#+END_SRC + +As with `#macro`, the clauses in `#assign` can be multi-line. The clauses are contained within the specified delimiters, which must balance at the time of definition, when the lexer scans them. This happens before any expansion. + +** BNF for `#assign` + +#+BEGIN_SRC bnf +cmd ::= "#assign" name body ; + +name ::= clause ; +body ::= clause ; + +clause ::= "(" literal? ")" | "[" expr? "]" ; + +literal ::= ; sequence parsed into tokens +expr ::= ; sequence parsed into tokens with recursive expansion of each token + +; white space, including new lines, is ignored. +#+END_SRC + +The name clause must resolve to a valid name identifier. + +There is a corresponding built-in macro form `_ASSIGN name body` which expands to nothing, but has the side effect of creating a definition. + +* New Built-in Macros + +Each new built-in macro has a `_` prefix. + +** `_ASSIGN` + +The same syntax and evaluation semantics as for `#assign` apply. + +#+BEGIN_SRC c +_ASSIGN (literal NAME) (literal body) +_ASSIGN (literal NAME) [expanded body] +_ASSIGN [expanded NAME] (literal body) +_ASSIGN [expanded NAME] [expanded body] +#+END_SRC + +The `_ASSIGN` macro has the same behavior as the `#assign` directive and, unlike other macros, is followed by, not one, but two delineated clauses. Each clause may be parenthesized (for a literal) or bracketed (for an expanded). + +The `_ASSIGN` macro expands to nothing. Its only effect is to create or update a macro binding. + +When a macro is created with `_ASSIGN`, it is colored. Hence, the newly created macro cannot be expanded within the same macro where it was defined. + + +* List Transformations + +#+BEGIN_SRC c +_TO_ARG_LIST( ) // Converts tokens into an argument list +_TO_TOKEN_LIST( ) // Flattens argument list into token stream +#+END_SRC + +These form the basis for transitioning between structure-aware and structure-neutral representations. + +* Token List Operators + +#+BEGIN_SRC c +_FIRST( ) // First token +_REST( ) // All tokens after the first +_MAP(f ,) // f(token1) f(token2) ... +_AL_MAP(f ,) // f(arg1) f(arg2) ... +#+END_SRC + +These enable iteration-like behavior over static macro arguments. + +* Logic Primitives + +#+BEGIN_SRC c +_IF(p ,a ,b) // If p is present, expand to a; else b +_NOT(p) // Expand to TRUE if p is empty, else empty +#+END_SRC + +Logical presence is defined by *token existence*, not token content. Empty argument lists are `false`. All others are `true`. + +#+BEGIN_SRC c +_AND( , ,... ) // Returns nothing if any arg is empty +_OR( , ,... ) // Returns first non-empty argument +#+END_SRC + +Missing trailing arguments are *not* detected by `_AND`. Appending a `TRUE` token is recommended to guard edge cases. + +* Token-Level Checks + +#+BEGIN_SRC c +_IS_IDENTIFIER( ) // If arg is an identifier, expands to it; else nothing +_IS_NAME( ) // If arg is an identifier, expands to it; else nothing +#+END_SRC + +* Pasting and Construction + +#+BEGIN_SRC c +_PASTE( , ,... ) // Combines all tokens into a single token +#+END_SRC + +Pasting is only valid if the result is a valid preprocessing token (e.g., identifier). Invalid pastes will trigger compiler diagnostics. + +* Usage Example + +(To be added) + +* Sets + +Sets are macros whose membership is defined by naming convention. All set operations are evaluated through presence of a named binding. + +#+BEGIN_SRC c +#define _SET_ADD(set ,x) ASSIGN( _PASTE(__SET_ ,set ,_ ,x), ) +#define _SET_IN(set ,x) _NOT(_PASTE(__SET_ ,set ,_ ,x)) +#+END_SRC + +A macro such as `__SET_myset__foo` existing in the preprocessor environment marks `foo` as a member of `myset`. + +* Associative Sets (ASet) + +Associative sets allow storing values indexed by a key: + +#+BEGIN_SRC c +#assign _ASET_ADD(set ,x ,y) ( + ASSIGN(_PASTE(__ASET_ ,set ,_ ,x)) + ASSIGN(_PASTE(__ASET_ITEM_ ,set ,_ ,x)) +) + +#assign _ASET_GET(set ,x) ( + _IF( + _PASTE(__ASET_ ,set ,_ ,x) + ,_PASTE(__ASET_ITEM_ ,set ,_ ,x) + , + ) +) +#+END_SRC + +If `x` is in the set, `_ASET_GET` returns the associated value; else nothing. + +* (End of manual — more modules to be added as completed.) diff --git "a/developer/experiment\360\237\226\211/.gitignore" "b/developer/experiment\360\237\226\211/.gitignore" deleted file mode 100644 index cba7efc..0000000 --- "a/developer/experiment\360\237\226\211/.gitignore" +++ /dev/null @@ -1 +0,0 @@ -a.out diff --git "a/developer/experiment\360\237\226\211/RT_CAT_1.c" "b/developer/experiment\360\237\226\211/RT_CAT_1.c" deleted file mode 100644 index 17663b1..0000000 --- "a/developer/experiment\360\237\226\211/RT_CAT_1.c" +++ /dev/null @@ -1,6 +0,0 @@ -#include - -int main(void){ - printf( "The answer is: %s\n", RT_CAT ); - return 0; -} diff --git "a/developer/experiment\360\237\226\211/assign_1.c" "b/developer/experiment\360\237\226\211/assign_1.c" deleted file mode 100644 index f58320c..0000000 --- "a/developer/experiment\360\237\226\211/assign_1.c" +++ /dev/null @@ -1,8 +0,0 @@ -#include - -#assign (ANSWER) (42) - -int main(void){ - printf( "The answer is: %d\n", ANSWER ); - return 0; -} diff --git "a/developer/experiment\360\237\226\211/assign_2.c" "b/developer/experiment\360\237\226\211/assign_2.c" deleted file mode 100644 index 3a41a63..0000000 --- "a/developer/experiment\360\237\226\211/assign_2.c" +++ /dev/null @@ -1,22 +0,0 @@ -#include - -#if 0 -#define STRINGIFY(x) #x -#define TOSTRING(x) STRINGIFY(x) -#define SHOW_MACRO(x) _Pragma(TOSTRING(message(#x " → " TOSTRING(x)))) - -SHOW_MACRO(a) -SHOW_MACRO(b) - -SHOW_MACRO($a) -SHOW_MACRO($b) -#endif - -#define a 2 -#define b 3 -#assign (ADD) [a + b] - -int main(void){ - printf("2 + 3 = %d\n", ADD); - return 0; -} diff --git "a/developer/experiment\360\237\226\211/macro_1.c" "b/developer/experiment\360\237\226\211/macro_1.c" deleted file mode 100644 index 2ba3e31..0000000 --- "a/developer/experiment\360\237\226\211/macro_1.c" +++ /dev/null @@ -1,8 +0,0 @@ -#include - -#rt_macro I(x) (x) - -int main(void){ - printf("5: %x" ,I(5)); - return 0; -} diff --git "a/developer/experiment\360\237\226\211/macro_2.c" "b/developer/experiment\360\237\226\211/macro_2.c" deleted file mode 100644 index 421f176..0000000 --- "a/developer/experiment\360\237\226\211/macro_2.c" +++ /dev/null @@ -1,11 +0,0 @@ -#include - -#rt_macro Q(f ,...)( - printf(f ,__VA_ARGS__) -) - -int main(void){ - Q("%x %x %x" ,1 ,2 ,3); - putchar('\n'); - return 0; -} diff --git "a/developer/experiment\360\237\226\211/paste_1.c" "b/developer/experiment\360\237\226\211/paste_1.c" deleted file mode 100644 index d438c6b..0000000 --- "a/developer/experiment\360\237\226\211/paste_1.c" +++ /dev/null @@ -1,10 +0,0 @@ -#include - -#define CAT2(a ,b) a##b -#define STRING_1(a) #a -#define STRING(a) STRING_1(a) - -int main(){ - printf("ab: %s." ,STRING(CAT2(a ,b))); - printf("gcc -E this one: %s." ,STRING(CAT2(a ,+))); -} diff --git "a/developer/experiment\360\237\226\211/recursive_define_0.c" "b/developer/experiment\360\237\226\211/recursive_define_0.c" deleted file mode 100644 index da2919a..0000000 --- "a/developer/experiment\360\237\226\211/recursive_define_0.c" +++ /dev/null @@ -1,52 +0,0 @@ -#include - - -#define X Q()B()(C) -#define Q() A -#define A(x) 21 -#define B() - -int main(){ - printf("X %d" ,X); -} - -/* - -So cpp truly does walk the token list with `cpp_get_token_1(pfile, &loc);` expanding each token one by one left to right. We should be able to get the desired expansion by adding another layer of macros, so that the argument expansion catches the final A(c). - -2025-05-12T02:53:59Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> gcc test_recursive_define_0.c -test_recursive_define_0.c: In function ‘main’: -test_recursive_define_0.c:5:13: warning: implicit declaration of function ‘A’ [-Wimplicit-function-declaration] - 5 | #define Q() A - | ^ -test_recursive_define_0.c:4:12: note: in expansion of macro ‘Q’ - 4 | #define X Q()B()(C) - | ^ -test_recursive_define_0.c:10:18: note: in expansion of macro ‘X’ - 10 | printf("X %d" ,X); - | ^ -test_recursive_define_0.c:4:19: error: ‘C’ undeclared (first use in this function) - 4 | #define X Q()B()(C) - | ^ -test_recursive_define_0.c:10:18: note: in expansion of macro ‘X’ - 10 | printf("X %d" ,X); - | ^ -test_recursive_define_0.c:4:19: note: each undeclared identifier is reported only once for each function it appears in - 4 | #define X Q()B()(C) - | ^ -test_recursive_define_0.c:10:18: note: in expansion of macro ‘X’ - 10 | printf("X %d" ,X); - | ^ - -2025-05-12T02:54:17Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> gcc -E -P test_recursive_define_0.c | tail -n 5 -extern int __overflow (FILE *, int); - -int main(){ - printf("X %d" ,A(C)); -} - -*? diff --git "a/developer/experiment\360\237\226\211/recursive_define_1.c" "b/developer/experiment\360\237\226\211/recursive_define_1.c" deleted file mode 100644 index a7c6335..0000000 --- "a/developer/experiment\360\237\226\211/recursive_define_1.c" +++ /dev/null @@ -1,44 +0,0 @@ -#include - - -#define X Q()B()(C) -#define Q() A -#define A(x) 21 -#define B() - -#define P(x) x - -int main(){ - printf("X %d" ,P(X)); -} - -/* -2025-05-12T03:00:19Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> /bin/gcc test_recursive_define_1.c - -2025-05-12T03:00:51Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> ./a.out -X 21 -2025-05-12T03:00:55Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> gcc test_recursive_define_1.c - -2025-05-12T03:01:03Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> ./a.out -X 21 -2025-05-12T03:01:05Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> gcc -E -P test_recursive_define_1.c | tail -n 5 -extern int __overflow (FILE *, int); - -int main(){ - printf("X %d" ,21); -} - -2025-05-12T03:01:11Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> -*/ diff --git "a/developer/experiment\360\237\226\211/recursive_define_2.c" "b/developer/experiment\360\237\226\211/recursive_define_2.c" deleted file mode 100644 index ecd87cc..0000000 --- "a/developer/experiment\360\237\226\211/recursive_define_2.c" +++ /dev/null @@ -1,58 +0,0 @@ -#include - - -#define X Q()B()(C) -#define Q() A -#define A(x) 21 -#define B() - -#define P X - -int main(){ - printf("X %d" ,P); -} - -/* - This does not work because it does not cause an argument evaluation, as did test_recursive_define_1.c - -2025-05-12T03:01:11Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> gcc test_recursive_define_2.c -test_recursive_define_2.c: In function ‘main’: -test_recursive_define_2.c:5:13: warning: implicit declaration of function ‘A’ [-Wimplicit-function-declaration] - 5 | #define Q() A - | ^ -test_recursive_define_2.c:4:12: note: in expansion of macro ‘Q’ - 4 | #define X Q()B()(C) - | ^ -test_recursive_define_2.c:9:11: note: in expansion of macro ‘X’ - 9 | #define P X - | ^ -test_recursive_define_2.c:12:18: note: in expansion of macro ‘P’ - 12 | printf("X %d" ,P); - | ^ -test_recursive_define_2.c:4:19: error: ‘C’ undeclared (first use in this function) - 4 | #define X Q()B()(C) - | ^ -test_recursive_define_2.c:9:11: note: in expansion of macro ‘X’ - 9 | #define P X - | ^ -test_recursive_define_2.c:12:18: note: in expansion of macro ‘P’ - 12 | printf("X %d" ,P); - | ^ -test_recursive_define_2.c:4:19: note: each undeclared identifier is reported only once for each function it appears in - 4 | #define X Q()B()(C) - | ^ -test_recursive_define_2.c:9:11: note: in expansion of macro ‘X’ - 9 | #define P X - | ^ -test_recursive_define_2.c:12:18: note: in expansion of macro ‘P’ - 12 | printf("X %d" ,P); - | ^ - -2025-05-12T03:02:36Z[developer] -Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ -> -*/ - - diff --git "a/developer/experiment\360\237\226\211/temp" "b/developer/experiment\360\237\226\211/temp" deleted file mode 100644 index 32b29d9..0000000 --- "a/developer/experiment\360\237\226\211/temp" +++ /dev/null @@ -1,9 +0,0 @@ -git mv assign_test_1.c assign_1.c -git mv assign_test_2.c assign_2.c -git mv macro_test_1.c macro_1.c -git mv macro_test_2.c macro_2.c -git mv recursive_define_0.c recursive_define_0.c -git mv recursive_define_1.c recursive_define_1.c -git mv recursive_define_2.c recursive_define_2.c -git mv RT_CAT_test_1.c RT_CAT_1.c -git mv va_arg_test_1.c va_arg_1.c diff --git "a/developer/experiment\360\237\226\211/va_arg_test_1.c" "b/developer/experiment\360\237\226\211/va_arg_test_1.c" deleted file mode 100644 index 8870054..0000000 --- "a/developer/experiment\360\237\226\211/va_arg_test_1.c" +++ /dev/null @@ -1,9 +0,0 @@ -#include - -#define A(...) (__VA_ARGS__) - - -int main(void){ - printf( "The answer is: %d\n", A(1,2,3) ); - return 0; -} diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/bump_buffer.org" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/bump_buffer.org" new file mode 100644 index 0000000..7eddbac --- /dev/null +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/bump_buffer.org" @@ -0,0 +1,101 @@ +#+TITLE: CPP bump buffer +#+AUTHOR: Thomas +#+DATE: 2025-06-08 +#+OPTIONS: toc:nil ^:nil + +* References +https://www.chiark.greenend.org.uk/doc/cpp-4.3-doc/cppinternals.html + +#+begin_quote +**References** + +- *libcpp/lex.cc* and *libcpp/macro.cc* (GCC ≥ 12) +- Ian Lance Taylor, “*Inside libcpp*”, GNU Cauldron 2018 slides +#+end_quote + +* The arena +/libcpp/ keeps a bump-pointer arena inside =cpp_reader=. Two routines are used to directly interface with it: =_cpp_reserve_room()=, and =_cpp_commit_buff()=. + + + +| Helper | Exact responsibility | +|--------------------------------+--------------------------------------------------------| +| =_cpp_reserve_room()= | Provide tentative space without moving the cursor | +| =_cpp_commit_buff()= | Finalise the previous reservation and move the cursor | + +* _cpp_reserve_room + +Defined in `internal.h`: + +#+begin_src c +static inline void *_cpp_reserve_room (cpp_reader *pfile, size_t have, size_t extra){ + if (BUFF_ROOM (pfile->a_buff) < (have + extra)) + _cpp_extend_buff (pfile, &pfile->a_buff, extra); + return BUFF_FRONT (pfile->a_buff); +} +#+end_src + +Example invocation found in `macro.cc::create_iso_definition`: + +#+begin_src c + macro = _cpp_new_macro( + pfile + ,cmk_macro + ,_cpp_reserve_room(pfile, 0, sizeof (cpp_macro)) + ); +#+end_src + +=_cpp_reserve_room()= returns a pointer to a buffer of the specified size, `have` + `extra`. The `have` argument is interesting, as it can be changed from call to call, hence it is more for the programmer than for the function. + +The buffer returned is not generally a unique buffer. It might be the same one that was returned on the previous invocation of =_cpp_reserve_room()=, or it will be a new buffer if `_cpp_extend_buff` was called during allocation. In this latter case, a new larger buffer is made, and the old buffer is copied into it. The new buffer is then returned. + +Hence, each call to =_cpp_reserve_room()= requires that all pointers into the buffer that existed before the call be considered stale. After the call, the only valid pointer to the buffer is the one that =_cpp_reserve_room()= returns. + + +* `_cpp_commit_buff` + +Defined in `lex.cc`: + +#+begin_src c +void *_cpp_commit_buff (cpp_reader *pfile, size_t size){ + void *ptr = BUFF_FRONT (pfile->a_buff); + + if (pfile->hash_table->alloc_subobject) + { + void *copy = pfile->hash_table->alloc_subobject (size); + memcpy (copy, ptr, size); + ptr = copy; + } + else + BUFF_FRONT (pfile->a_buff) += size; + + return ptr; +} +#+end_src + +A call to =_cpp_commit_buff()= finalizes the accumulation of reserved room, and sets the buffer aside as a dedicated allocation. Once =_cpp_commit_buff()= is called, a subsequent call to =_cpp_reserve_room()= will start fresh with a new buffer. + +When `! pfile->hash_table->alloc_subobject`, the buffer will be committed with `BUFF_FRONT (pfile->a_buff) += size;` + +However, when `pfile->hash_table->alloc_subobject` is true, the buffer will be copied when it is committed. Hence, it is important to consider all pointers into the buffer created during the one or more calls to =_cpp_reserve_room()= to be stale after a call to =_cpp_commit_buff()=, and to subsequently only use the pointer that =_cpp_commit_buff()= returns. + +According to `GPT o3`, which hasn't gotten much right thus far,"`alloc_subobject` is non-null only while installing a macro into the hash table (create_definition etc.)." + +* Deallocating a `_cpp_reserve_room` buffer yet to be committed + +Each subsequent call to `_cpp_reserve_room`, even if it is unrelated to the prior one, either uses the buffer it finds at `pfile->a_buff`, or if it is not large enough, replaces it with a larger copy. Consequently, there is no concept of 'deallocating a `_cpp_reserve_room` buffer'. + +However, according to `o3`, if one desired: + +#+begin_src c + void *mark = _cpp_get_buff (pfile); /* snapshot cursor */ + ... + _cpp_release_buff (pfile, mark); /* rewind to mark */ +#+end_src + + +* Deallocating a committed `_cpp_reserve_room` buffer + +There is no facility for this. Detritus collects until CPP moves on to the next translation unit and then everything is released. + + diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/custom_directives_macros.org" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/custom_directives_macros.org" new file mode 100644 index 0000000..a38289d --- /dev/null +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/custom_directives_macros.org" @@ -0,0 +1,73 @@ +#+TITLE: Adding Custom Directives and Built-in Macros in libcpp +#+AUTHOR: Thomas +#+DATE: 2025-06-10 +#+OPTIONS: toc:t + +* Custom Preprocessor Directives + +To add a new `#directive` (e.g. `#assign`, `#macro`), changes must be made in several files within libcpp. + +** `directives.cc` + +- Midway through the file, there is a static table mapping directive names to their handlers. +- Add an entry here for your new directive: + #+begin_src c + { "assign", RT_ASSIGN, false }, + { "macro", RT_MACRO, false }, + #+end_src + +- Define your handler in the switch for `handle_directive`. Place the handler logic (e.g., `handle_assign_directive`) in the section titled: + #+begin_quote + /* RT Extensions */ + #+end_quote + +** Tip: + Group your RT extensions near the end of the file for easy maintenance. + +--- + +* Custom Built-in Macros + +To define new built-in macros (e.g. `_CAT`, `_MAP`), several files need coordinated updates. + +** `include/cpplib.h` + +- Add a new entry to the `enum cpp_builtin_type`: + #+begin_src c + BT_CAT, + #+end_src + +- This enum distinguishes the macro’s kind in logic dispatch. + +** `init.cc` + +- Extend the `builtin_array[]` definition: + #+begin_src c + B("_CAT", BT_CAT, true), + #+end_src + +- The third parameter (`true`) indicates that a redefinition warning should be issued if the macro is redefined in user code. +- *Important:* The final entries in the table (`__DATE__`, `__TIME__`, etc.) are position-sensitive. Insert new macros near the top or middle, not the end. + +** `macros.cc` + +- This is where the actual macro expansion logic resides. +- Your `handle_builtin_macro` function will receive a `cpp_builtin_type` enum and dispatch accordingly. +- Add your implementation (e.g., for `_CAT`) in the RT Extensions section near the end of `macros.cc`. + + For example: + #+begin_src c + case BT_CAT: + return expand_cat_macro(pfile, ...); + #+end_src + +- If your macro uses token splicing, argument unpacking, or nesting, consider isolating each macro as a separate function for clarity and testability. + +--- + +* Notes + +- All RT Extension code is grouped under labeled sections (e.g., `/* RT Extensions */`) for both directives and built-in macros. This helps maintain a clean separation from upstream GCC code. +- If your extensions modify memory allocation, token arena behavior, or call back into the lexer, be sure to test under both expanded and unexpanded modes. + +Let me know if you'd like this converted into a template or included in your developer README. diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/fetching_a_token.org" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/fetching_a_token.org" new file mode 100644 index 0000000..30f4988 --- /dev/null +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/fetching_a_token.org" @@ -0,0 +1,92 @@ +#+TITLE: Token Fetching Routines in libcpp +#+AUTHOR: Thomas +#+DATE: 2025-06-10 +#+OPTIONS: toc:t + +* Getting a Token + +There are many routines for getting a token in libcpp, each tuned for different phases of preprocessing. This guide summarizes the key interfaces, grouped by their expansion behavior. + +Note that the term *expansion* is overloaded in libcpp: + +1. One meaning refers to **macro expansion** — replacing a macro call with its definition (possibly recursively). +2. Another refers to the **macro definition buffer** itself: macros store an array of tokens as their "expansion". Adding a token to this array is said to "add an expansion token", which is unrelated to expanding a macro. + +--- + +* No Macro Expansion + +** `lex.cc::_cpp_lex_token(pfile)` + +- Returns a pointer to the next token without macro expansion. +- It does handle important preprocessor logic: + - Skipping gated code (`#if 0`) + - Handling `#line` or `#pragma` + - Deferred pragma state +- This is the general-purpose low-level token fetch used during parsing — but it still obeys the logical flow of the file. + +** `macro.cc::lex_expansion_token` + +- Lexes a token and places it into the macro's `expansion` array. +- The token is *not* processed in the usual way — it does *not* honor `#if` skipping, macro argument parsing, or directive detection. +- Used by `create_iso_definition()` when parsing macro bodies. + +** `lex.cc::_cpp_lex_direct(pfile)` + +- Very low-level lexer function. +- Called *only* by `_cpp_lex_token`. +- Assumes `pfile->cur_token` is pre-set; writes directly into that memory. +- Bypasses all macro, directive, or skipping logic. +- Should not be used outside of tightly controlled contexts — *not* a public interface. + +--- + +* With Macro Expansion + +** `macro.cc::cpp_get_token_1(pfile, &src_loc)` + +- Returns the next token with full macro expansion. +- Also returns `location_t` via an out parameter. +- This is the recommended entry point for frontends or extensions that need both token and position. +- Internally calls `_cpp_lex_token` and handles expansion. + +** `macro.cc::cpp_get_token(pfile)` + +- Thin wrapper around `cpp_get_token_1`, but discards the location. +- Used in internal code where location is unnecessary. + +** `macro.cc::cpp_get_token_no_padding(pfile)` + +- Calls `cpp_get_token`, but skips tokens of type `CPP_PADDING`. +- Useful when padding tokens are irrelevant (e.g. while collecting macro arguments). + +--- + +* RT Extension + +** `rt_extensions.cc::get_token_noexpand(pfile)` + +- Returns one unexpanded token by value. +- Does *not* use the regular token stream; instead, it temporarily sets `pfile->cur_token` to a local buffer, calls `_cpp_lex_direct`, and restores the original pointer. +- Unlike `lex_expansion_token`, it does *not* store the result in a macro; it simply returns the token. +- This is appropriate for previewing the next token *without affecting* the token stream or macro state. + +However, for parsing structured clauses such as macro bodies — especially those that may include `#if`/`#endif` constructs — it is better to use `_cpp_lex_token`, to match the conditional skipping behavior of `cpp_get_token_1`. + +--- + +* Summary Table + +| Function | Expansion | Skips `#if 0` | Stores Token | Returns Token | Returns Location | +|-----------------------------+-----------+---------------+---------------+----------------+-------------------| +| `lex_expansion_token` | ❌ | ❌ | ✔ (into macro) | ❌ | ✔ | +| `_cpp_lex_direct` | ❌ | ❌ | ✔ (in-place) | ❌ | ✔ | +| `_cpp_lex_token` | ❌ | ✔ | ✔ | ✔ | ✔ | +| `cpp_get_token` | ✔ | ✔ | ✔ | ✔ | ❌ | +| `cpp_get_token_1` | ✔ | ✔ | ✔ | ✔ | ✔ | +| `cpp_get_token_no_padding` | ✔ | ✔ | ✔ | ✔ (no padding) | ❌ | +| `get_token_noexpand` | ❌ | ❌ | ❌ | ✔ (by value) | ✔ | + +--- + +Let me know if you'd like to cross-link these to their actual definitions or integrate them into a larger RT CPP developer manual. diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/todo.org" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/todo.org" new file mode 100644 index 0000000..40142c6 --- /dev/null +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/document\360\237\226\211/todo.org" @@ -0,0 +1,23 @@ +2025-05-00 + + - Add the call back and warn logic for #assign in the macro.cc::name_clause_is_name function. + + - The name is currently () or [], probably should allow a single name ID or [] instead. + or perhaps in general, allow for evaluated, not evaluated, or single identifier options. + + - When this matures, should replace the capture/install with diff and patch. + + +2025-05-17 in maco.cc, seems the end cases in `parse_clause_literal()` should be included in `parse_clause_expand()`. + +2025-05-18 It would have been better perhaps to send in a pointer to a token allocation, + instead of to a src_loc, and and terminal type. + + +2025-06-08 + + - I wonder if _ASSIGN can be contrived, perhaps along with a count, to create an + infinite number of uniquely named macros. This might be worth exploring. + + + diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_capture.sh" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_capture.sh" index ab0813f..7e3ad86 100755 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_capture.sh" +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_capture.sh" @@ -5,7 +5,7 @@ set -euo pipefail source "$(dirname "$0")/environment.sh" -echo "⚠️ You probably don't want to run this script. The files in \$ROOT/script_Deb-12.10_gcc-12.4.1🖉/library are intended to be the authoritative copies." +echo "⚠️ You probably don't want to run this script. The files in \$DEVELOPER/script_Deb-12.10_gcc-12.4.1🖉/library are intended to be the authoritative copies." echo "So you did the bad thing and edited the files directly in the GCC source tree? Then this script is for you. ;-)" echo @@ -18,8 +18,8 @@ else fi -if [[ -z "${ROOT:-}" ]]; then - echo "❌ ROOT environment variable is not set. Aborting." +if [[ -z "${DEVELOPER:-}" ]]; then + echo "❌ DEVELOPER environment variable is not set. Aborting." exit 1 fi if [[ -z "${SCRIPT_DIR:-}" ]]; then @@ -30,8 +30,8 @@ SRCDIR="library/" DESTDIR="$GCC_SRC/libcpp/" -SRCDIR="$ROOT/source/gcc-12.2.0/libcpp" -DESTDIR="$ROOT/script_Deb-12.10_gcc-12.4.1🖉/library" +SRCDIR="$DEVELOPER/source/gcc-12.2.0/libcpp" +DESTDIR="$DEVELOPER/script_Deb-12.10_gcc-12.4.1🖉/library" if [[ ! -d "$SRCDIR" ]]; then echo "❌ Source directory '$SRCDIR' does not exist." diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_diff.sh" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_diff.sh" index 31dfb00..9d75cc9 100755 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_diff.sh" +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_diff.sh" @@ -5,8 +5,8 @@ set -euo pipefail source "$(dirname "$0")/environment.sh" # Check required env vars -if [[ -z "${ROOT:-}" ]]; then - echo "❌ ROOT environment variable is not set. Aborting." +if [[ -z "${DEVELOPER:-}" ]]; then + echo "❌ DEVELOPER environment variable is not set. Aborting." exit 1 fi if [[ -z "${SCRIPT_DIR:-}" ]]; then diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_install.sh" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_install.sh" index 33033a5..a25a026 100755 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_install.sh" +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_install.sh" @@ -1,31 +1,31 @@ #!/bin/bash -# transfer.sh – Install RT library files into GCC libcpp source tree. +# ext_isntall.sh – Install RT library files into GCC libcpp source tree. # Usage: -# ./transfer.sh → transfers all files in RT_CPP_FILES -# ./transfer.sh init.cc → transfers only init.cc -# ./transfer.sh init.cc macro.cc → transfers just those +# ./ext_isntall.sh → ext_isntalls all files in RT_CPP_FILES +# ./ext_isntall.sh init.cc → ext_isntalls only init.cc +# ./ext_isntall.sh init.cc macro.cc → ext_isntalls just those set -euo pipefail -# provides: $ROOT, $RT_CPP_FILES, $GCC_SRC +# provides: $DEVELOPER, $RT_CPP_FILES, $GCC_SRC source "$(dirname "$0")/environment.sh" SRCDIR="library" DESTDIR="$GCC_SRC/libcpp" # Validate environment -[[ -z "${ROOT:-}" ]] && { echo "❌ ROOT is not set. Aborting."; exit 1; } +[[ -z "${DEVELOPER:-}" ]] && { echo "❌ DEVELOPER is not set. Aborting."; exit 1; } [[ ! -d "$SRCDIR" ]] && { echo "❌ Source directory '$SRCDIR' missing."; exit 1; } [[ ! -d "$DESTDIR" ]] && { echo "❌ Destination directory '$DESTDIR' missing."; exit 1; } -# Determine list of files to transfer +# Determine list of files to ext_isntall if [[ $# -eq 0 ]]; then file_list=("${RT_CPP_FILES[@]}") else file_list=("$@") fi -echo "📋 Transferring files to $DESTDIR..." +echo "📋 Ext_Isntallring files to $DESTDIR..." for file in "${file_list[@]}"; do src="$SRCDIR/$file" @@ -55,4 +55,4 @@ for file in "${file_list[@]}"; do fi done -echo "✅ Transfer complete." +echo "✅ Ext_Isntall complete." diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_save.sh" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_save.sh" index fe2a212..2869048 100755 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_save.sh" +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/ext_save.sh" @@ -8,8 +8,8 @@ source "$(dirname "$0")/environment.sh" # Appends _orig after the .cc extension (e.g., macro.cc → macro.cc_orig) # Files remain in place but can be manually diffed or restored if needed -if [[ -z "${ROOT:-}" ]]; then - echo "❌ ROOT environment variable is not set. Aborting." +if [[ -z "${DEVELOPER:-}" ]]; then + echo "❌ DEVELOPER environment variable is not set. Aborting." exit 1 fi if [[ -z "${SCRIPT_DIR:-}" ]]; then diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/#macro.cc#" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/#macro.cc#" deleted file mode 100644 index 0909d55..0000000 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/#macro.cc#" +++ /dev/null @@ -1,5374 +0,0 @@ -/* Part of CPP library. (Macro and #define handling.) - Copyright (C) 1986-2022 Free Software Foundation, Inc. - Written by Per Bothner, 1994. - Based on CCCP program by Paul Rubin, June 1986 - Adapted to ANSI C, Richard Stallman, Jan 1987 - -This program is free software; you can redistribute it and/or modify it -under the terms of the GNU General Public License as published by the -Free Software Foundation; either version 3, or (at your option) any -later version. - -This program is distributed in the hope that it will be useful, -but WITHOUT ANY WARRANTY; without even the implied warranty of -MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -GNU General Public License for more details. - -You should have received a copy of the GNU General Public License -along with this program; see the file COPYING3. If not see -. - - In other words, you are welcome to use, share and improve this program. - You are forbidden to forbid anyone else to use, share and improve - what you give them. Help stamp out software-hoarding! */ - -#pragma GCC diagnostic ignored "-Wparentheses" - - -#include "config.h" -#include "system.h" -#include "cpplib.h" -#include "internal.h" - -// RT extension -static const uchar *evaluate_TO_ARG_LIST(cpp_reader *pfile); - - -typedef struct macro_arg macro_arg; -/* This structure represents the tokens of a macro argument. These - tokens can be macro themselves, in which case they can be either - expanded or unexpanded. When they are expanded, this data - structure keeps both the expanded and unexpanded forms. */ -struct macro_arg -{ - const cpp_token **first; /* First token in unexpanded argument. */ - const cpp_token **expanded; /* Macro-expanded argument. */ - const cpp_token *stringified; /* Stringified argument. */ - unsigned int count; /* # of tokens in argument. */ - unsigned int expanded_count; /* # of tokens in expanded argument. */ - location_t *virt_locs; /* Where virtual locations for - unexpanded tokens are stored. */ - location_t *expanded_virt_locs; /* Where virtual locations for - expanded tokens are - stored. */ -}; - -/* The kind of macro tokens which the instance of - macro_arg_token_iter is supposed to iterate over. */ -enum macro_arg_token_kind { - MACRO_ARG_TOKEN_NORMAL, - /* This is a macro argument token that got transformed into a string - literal, e.g. #foo. */ - MACRO_ARG_TOKEN_STRINGIFIED, - /* This is a token resulting from the expansion of a macro - argument that was itself a macro. */ - MACRO_ARG_TOKEN_EXPANDED -}; - -/* An iterator over tokens coming from a function-like macro - argument. */ -typedef struct macro_arg_token_iter macro_arg_token_iter; -struct macro_arg_token_iter -{ - /* Whether or not -ftrack-macro-expansion is used. */ - bool track_macro_exp_p; - /* The kind of token over which we are supposed to iterate. */ - enum macro_arg_token_kind kind; - /* A pointer to the current token pointed to by the iterator. */ - const cpp_token **token_ptr; - /* A pointer to the "full" location of the current token. If - -ftrack-macro-expansion is used this location tracks loci across - macro expansion. */ - const location_t *location_ptr; -#if CHECKING_P - /* The number of times the iterator went forward. This useful only - when checking is enabled. */ - size_t num_forwards; -#endif -}; - -/* Saved data about an identifier being used as a macro argument - name. */ -struct macro_arg_saved_data { - /* The canonical (UTF-8) spelling of this identifier. */ - cpp_hashnode *canonical_node; - /* The previous value & type of this identifier. */ - union _cpp_hashnode_value value; - node_type type; -}; - -static const char *vaopt_paste_error = - N_("'##' cannot appear at either end of __VA_OPT__"); - -static void expand_arg (cpp_reader *, macro_arg *); - -/* A class for tracking __VA_OPT__ state while iterating over a - sequence of tokens. This is used during both macro definition and - expansion. */ -class vaopt_state { - - public: - - enum update_type - { - ERROR, - DROP, - INCLUDE, - BEGIN, - END - }; - - /* Initialize the state tracker. ANY_ARGS is true if variable - arguments were provided to the macro invocation. */ - vaopt_state (cpp_reader *pfile, bool is_variadic, macro_arg *arg) - : m_pfile (pfile), - m_arg (arg), - m_variadic (is_variadic), - m_last_was_paste (false), - m_stringify (false), - m_state (0), - m_paste_location (0), - m_location (0), - m_update (ERROR) - { - } - - /* Given a token, update the state of this tracker and return a - boolean indicating whether the token should be be included in the - expansion. */ - update_type update (const cpp_token *token) - { - /* If the macro isn't variadic, just don't bother. */ - if (!m_variadic) - return INCLUDE; - - if (token->type == CPP_NAME - && token->val.node.node == m_pfile->spec_nodes.n__VA_OPT__) - { - if (m_state > 0) - { - cpp_error_at (m_pfile, CPP_DL_ERROR, token->src_loc, - "__VA_OPT__ may not appear in a __VA_OPT__"); - return ERROR; - } - ++m_state; - m_location = token->src_loc; - m_stringify = (token->flags & STRINGIFY_ARG) != 0; - return BEGIN; - } - else if (m_state == 1) - { - if (token->type != CPP_OPEN_PAREN) - { - cpp_error_at (m_pfile, CPP_DL_ERROR, m_location, - "__VA_OPT__ must be followed by an " - "open parenthesis"); - return ERROR; - } - ++m_state; - if (m_update == ERROR) - { - if (m_arg == NULL) - m_update = INCLUDE; - else - { - m_update = DROP; - if (!m_arg->expanded) - expand_arg (m_pfile, m_arg); - for (unsigned idx = 0; idx < m_arg->expanded_count; ++idx) - if (m_arg->expanded[idx]->type != CPP_PADDING) - { - m_update = INCLUDE; - break; - } - } - } - return DROP; - } - else if (m_state >= 2) - { - if (m_state == 2 && token->type == CPP_PASTE) - { - cpp_error_at (m_pfile, CPP_DL_ERROR, token->src_loc, - vaopt_paste_error); - return ERROR; - } - /* Advance states before further considering this token, in - case we see a close paren immediately after the open - paren. */ - if (m_state == 2) - ++m_state; - - bool was_paste = m_last_was_paste; - m_last_was_paste = false; - if (token->type == CPP_PASTE) - { - m_last_was_paste = true; - m_paste_location = token->src_loc; - } - else if (token->type == CPP_OPEN_PAREN) - ++m_state; - else if (token->type == CPP_CLOSE_PAREN) - { - --m_state; - if (m_state == 2) - { - /* Saw the final paren. */ - m_state = 0; - - if (was_paste) - { - cpp_error_at (m_pfile, CPP_DL_ERROR, token->src_loc, - vaopt_paste_error); - return ERROR; - } - - return END; - } - } - return m_update; - } - - /* Nothing to do with __VA_OPT__. */ - return INCLUDE; - } - - /* Ensure that any __VA_OPT__ was completed. If ok, return true. - Otherwise, issue an error and return false. */ - bool completed () - { - if (m_variadic && m_state != 0) - cpp_error_at (m_pfile, CPP_DL_ERROR, m_location, - "unterminated __VA_OPT__"); - return m_state == 0; - } - - /* Return true for # __VA_OPT__. */ - bool stringify () const - { - return m_stringify; - } - - private: - - /* The cpp_reader. */ - cpp_reader *m_pfile; - - /* The __VA_ARGS__ argument. */ - macro_arg *m_arg; - - /* True if the macro is variadic. */ - bool m_variadic; - /* If true, the previous token was ##. This is used to detect when - a paste occurs at the end of the sequence. */ - bool m_last_was_paste; - /* True for #__VA_OPT__. */ - bool m_stringify; - - /* The state variable: - 0 means not parsing - 1 means __VA_OPT__ seen, looking for "(" - 2 means "(" seen (so the next token can't be "##") - >= 3 means looking for ")", the number encodes the paren depth. */ - int m_state; - - /* The location of the paste token. */ - location_t m_paste_location; - - /* Location of the __VA_OPT__ token. */ - location_t m_location; - - /* If __VA_ARGS__ substitutes to no preprocessing tokens, - INCLUDE, otherwise DROP. ERROR when unknown yet. */ - update_type m_update; -}; - -/* Macro expansion. */ - -static cpp_macro *get_deferred_or_lazy_macro (cpp_reader *, cpp_hashnode *, - location_t); -static int enter_macro_context (cpp_reader *, cpp_hashnode *, - const cpp_token *, location_t); -static int builtin_macro (cpp_reader *, cpp_hashnode *, - location_t, location_t); -static void push_ptoken_context (cpp_reader *, cpp_hashnode *, _cpp_buff *, - const cpp_token **, unsigned int); -static void push_extended_tokens_context (cpp_reader *, cpp_hashnode *, - _cpp_buff *, location_t *, - const cpp_token **, unsigned int); -static _cpp_buff *collect_args (cpp_reader *, const cpp_hashnode *, - _cpp_buff **, unsigned *); -static cpp_context *next_context (cpp_reader *); -static const cpp_token *padding_token (cpp_reader *, const cpp_token *); -static const cpp_token *new_string_token (cpp_reader *, uchar *, unsigned int); -static const cpp_token *stringify_arg (cpp_reader *, const cpp_token **, - unsigned int); -static void paste_all_tokens (cpp_reader *, const cpp_token *); -static bool paste_tokens (cpp_reader *, location_t, - const cpp_token **, const cpp_token *); -static void alloc_expanded_arg_mem (cpp_reader *, macro_arg *, size_t); -static void ensure_expanded_arg_room (cpp_reader *, macro_arg *, size_t, size_t *); -static void delete_macro_args (_cpp_buff*, unsigned num_args); -static void set_arg_token (macro_arg *, const cpp_token *, - location_t, size_t, - enum macro_arg_token_kind, - bool); -static const location_t *get_arg_token_location (const macro_arg *, - enum macro_arg_token_kind); -static const cpp_token **arg_token_ptr_at (const macro_arg *, - size_t, - enum macro_arg_token_kind, - location_t **virt_location); - -static void macro_arg_token_iter_init (macro_arg_token_iter *, bool, - enum macro_arg_token_kind, - const macro_arg *, - const cpp_token **); -static const cpp_token *macro_arg_token_iter_get_token -(const macro_arg_token_iter *it); -static location_t macro_arg_token_iter_get_location -(const macro_arg_token_iter *); -static void macro_arg_token_iter_forward (macro_arg_token_iter *); -static _cpp_buff *tokens_buff_new (cpp_reader *, size_t, - location_t **); -static size_t tokens_buff_count (_cpp_buff *); -static const cpp_token **tokens_buff_last_token_ptr (_cpp_buff *); -static inline const cpp_token **tokens_buff_put_token_to (const cpp_token **, - location_t *, - const cpp_token *, - location_t, - location_t, - const line_map_macro *, - unsigned int); - -static const cpp_token **tokens_buff_add_token (_cpp_buff *, - location_t *, - const cpp_token *, - location_t, - location_t, - const line_map_macro *, - unsigned int); -static inline void tokens_buff_remove_last_token (_cpp_buff *); -static void replace_args (cpp_reader *, cpp_hashnode *, cpp_macro *, - macro_arg *, location_t); -static _cpp_buff *funlike_invocation_p (cpp_reader *, cpp_hashnode *, - _cpp_buff **, unsigned *); -static cpp_macro *create_iso_definition (cpp_reader *); - -/* #define directive parsing and handling. */ - -static cpp_macro *lex_expansion_token (cpp_reader *, cpp_macro *); -static bool parse_params (cpp_reader *, unsigned *, bool *); -static void check_trad_stringification (cpp_reader *, const cpp_macro *, - const cpp_string *); -static bool reached_end_of_context (cpp_context *); -static void consume_next_token_from_context (cpp_reader *pfile, - const cpp_token **, - location_t *); -static const cpp_token* cpp_get_token_1 (cpp_reader *, location_t *); - -static cpp_hashnode* macro_of_context (cpp_context *context); - -/* Statistical counter tracking the number of macros that got - expanded. */ -unsigned num_expanded_macros_counter = 0; -/* Statistical counter tracking the total number tokens resulting - from macro expansion. */ -unsigned num_macro_tokens_counter = 0; - -/* Wrapper around cpp_get_token to skip CPP_PADDING tokens - and not consume CPP_EOF. */ -static const cpp_token * -cpp_get_token_no_padding (cpp_reader *pfile) -{ - for (;;) - { - const cpp_token *ret = cpp_peek_token (pfile, 0); - if (ret->type == CPP_EOF) - return ret; - ret = cpp_get_token (pfile); - if (ret->type != CPP_PADDING) - return ret; - } -} - -/* Handle meeting "__has_include" builtin macro. */ - -static int -builtin_has_include (cpp_reader *pfile, cpp_hashnode *op, bool has_next) -{ - int result = 0; - - if (!pfile->state.in_directive) - cpp_error (pfile, CPP_DL_ERROR, - "\"%s\" used outside of preprocessing directive", - NODE_NAME (op)); - - pfile->state.angled_headers = true; - const cpp_token *token = cpp_get_token_no_padding (pfile); - bool paren = token->type == CPP_OPEN_PAREN; - if (paren) - token = cpp_get_token_no_padding (pfile); - else - cpp_error (pfile, CPP_DL_ERROR, - "missing '(' before \"%s\" operand", NODE_NAME (op)); - pfile->state.angled_headers = false; - - bool bracket = token->type != CPP_STRING; - char *fname = NULL; - if (token->type == CPP_STRING || token->type == CPP_HEADER_NAME) - { - fname = XNEWVEC (char, token->val.str.len - 1); - memcpy (fname, token->val.str.text + 1, token->val.str.len - 2); - fname[token->val.str.len - 2] = '\0'; - } - else if (token->type == CPP_LESS) - fname = _cpp_bracket_include (pfile); - else - cpp_error (pfile, CPP_DL_ERROR, - "operator \"%s\" requires a header-name", NODE_NAME (op)); - - if (fname) - { - /* Do not do the lookup if we're skipping, that's unnecessary - IO. */ - if (!pfile->state.skip_eval - && _cpp_has_header (pfile, fname, bracket, - has_next ? IT_INCLUDE_NEXT : IT_INCLUDE)) - result = 1; - - XDELETEVEC (fname); - } - - if (paren - && cpp_get_token_no_padding (pfile)->type != CPP_CLOSE_PAREN) - cpp_error (pfile, CPP_DL_ERROR, - "missing ')' after \"%s\" operand", NODE_NAME (op)); - - return result; -} - -/* Emits a warning if NODE is a macro defined in the main file that - has not been used. */ -int -_cpp_warn_if_unused_macro (cpp_reader *pfile, cpp_hashnode *node, - void *v ATTRIBUTE_UNUSED) -{ - if (cpp_user_macro_p (node)) - { - cpp_macro *macro = node->value.macro; - - if (!macro->used - && MAIN_FILE_P (linemap_check_ordinary - (linemap_lookup (pfile->line_table, - macro->line)))) - cpp_warning_with_line (pfile, CPP_W_UNUSED_MACROS, macro->line, 0, - "macro \"%s\" is not used", NODE_NAME (node)); - } - - return 1; -} - -/* Allocates and returns a CPP_STRING token, containing TEXT of length - LEN, after null-terminating it. TEXT must be in permanent storage. */ -static const cpp_token * -new_string_token (cpp_reader *pfile, unsigned char *text, unsigned int len) -{ - cpp_token *token = _cpp_temp_token (pfile); - - text[len] = '\0'; - token->type = CPP_STRING; - token->val.str.len = len; - token->val.str.text = text; - token->flags = 0; - return token; -} - -static const char * const monthnames[] = -{ - "Jan", "Feb", "Mar", "Apr", "May", "Jun", - "Jul", "Aug", "Sep", "Oct", "Nov", "Dec" -}; - -/* Helper function for builtin_macro. Returns the text generated by - a builtin macro. */ -const uchar * -_cpp_builtin_macro_text (cpp_reader *pfile, cpp_hashnode *node, - location_t loc) -{ - const uchar *result = NULL; - linenum_type number = 1; - - switch (node->value.builtin) - { - default: - cpp_error (pfile, CPP_DL_ICE, "invalid built-in macro \"%s\"", - NODE_NAME (node)); - break; - - case BT_TIMESTAMP: - { - if (CPP_OPTION (pfile, warn_date_time)) - cpp_warning (pfile, CPP_W_DATE_TIME, "macro \"%s\" might prevent " - "reproducible builds", NODE_NAME (node)); - - cpp_buffer *pbuffer = cpp_get_buffer (pfile); - if (pbuffer->timestamp == NULL) - { - /* Initialize timestamp value of the assotiated file. */ - struct _cpp_file *file = cpp_get_file (pbuffer); - if (file) - { - /* Generate __TIMESTAMP__ string, that represents - the date and time of the last modification - of the current source file. The string constant - looks like "Sun Sep 16 01:03:52 1973". */ - struct tm *tb = NULL; - struct stat *st = _cpp_get_file_stat (file); - if (st) - tb = localtime (&st->st_mtime); - if (tb) - { - char *str = asctime (tb); - size_t len = strlen (str); - unsigned char *buf = _cpp_unaligned_alloc (pfile, len + 2); - buf[0] = '"'; - strcpy ((char *) buf + 1, str); - buf[len] = '"'; - pbuffer->timestamp = buf; - } - else - { - cpp_errno (pfile, CPP_DL_WARNING, - "could not determine file timestamp"); - pbuffer->timestamp = UC"\"??? ??? ?? ??:??:?? ????\""; - } - } - } - result = pbuffer->timestamp; - } - break; - case BT_FILE: - case BT_FILE_NAME: - case BT_BASE_FILE: - { - unsigned int len; - const char *name; - uchar *buf; - - if (node->value.builtin == BT_FILE - || node->value.builtin == BT_FILE_NAME) - { - name = linemap_get_expansion_filename (pfile->line_table, - pfile->line_table->highest_line); - if ((node->value.builtin == BT_FILE_NAME) && name) - name = lbasename (name); - } - else - { - name = _cpp_get_file_name (pfile->main_file); - if (!name) - abort (); - } - if (pfile->cb.remap_filename) - name = pfile->cb.remap_filename (name); - len = strlen (name); - buf = _cpp_unaligned_alloc (pfile, len * 2 + 3); - result = buf; - *buf = '"'; - buf = cpp_quote_string (buf + 1, (const unsigned char *) name, len); - *buf++ = '"'; - *buf = '\0'; - } - break; - - case BT_INCLUDE_LEVEL: - /* The line map depth counts the primary source as level 1, but - historically __INCLUDE_DEPTH__ has called the primary source - level 0. */ - number = pfile->line_table->depth - 1; - break; - - case BT_SPECLINE: - /* If __LINE__ is embedded in a macro, it must expand to the - line of the macro's invocation, not its definition. - Otherwise things like assert() will not work properly. - See WG14 N1911, WG21 N4220 sec 6.5, and PR 61861. */ - if (CPP_OPTION (pfile, traditional)) - loc = pfile->line_table->highest_line; - else - loc = linemap_resolve_location (pfile->line_table, loc, - LRK_MACRO_EXPANSION_POINT, NULL); - number = linemap_get_expansion_line (pfile->line_table, loc); - break; - - /* __STDC__ has the value 1 under normal circumstances. - However, if (a) we are in a system header, (b) the option - stdc_0_in_system_headers is true (set by target config), and - (c) we are not in strictly conforming mode, then it has the - value 0. (b) and (c) are already checked in cpp_init_builtins. */ - case BT_STDC: - if (_cpp_in_system_header (pfile)) - number = 0; - else - number = 1; - break; - - case BT_DATE: - case BT_TIME: - if (CPP_OPTION (pfile, warn_date_time)) - cpp_warning (pfile, CPP_W_DATE_TIME, "macro \"%s\" might prevent " - "reproducible builds", NODE_NAME (node)); - if (pfile->date == NULL) - { - /* Allocate __DATE__ and __TIME__ strings from permanent - storage. We only do this once, and don't generate them - at init time, because time() and localtime() are very - slow on some systems. */ - time_t tt; - auto kind = cpp_get_date (pfile, &tt); - - if (kind == CPP_time_kind::UNKNOWN) - { - cpp_errno (pfile, CPP_DL_WARNING, - "could not determine date and time"); - - pfile->date = UC"\"??? ?? ????\""; - pfile->time = UC"\"??:??:??\""; - } - else - { - struct tm *tb = (kind == CPP_time_kind::FIXED - ? gmtime : localtime) (&tt); - - pfile->date = _cpp_unaligned_alloc (pfile, - sizeof ("\"Oct 11 1347\"")); - sprintf ((char *) pfile->date, "\"%s %2d %4d\"", - monthnames[tb->tm_mon], tb->tm_mday, - tb->tm_year + 1900); - - pfile->time = _cpp_unaligned_alloc (pfile, - sizeof ("\"12:34:56\"")); - sprintf ((char *) pfile->time, "\"%02d:%02d:%02d\"", - tb->tm_hour, tb->tm_min, tb->tm_sec); - } - } - - if (node->value.builtin == BT_DATE) - result = pfile->date; - else - result = pfile->time; - break; - - case BT_COUNTER: - if (CPP_OPTION (pfile, directives_only) && pfile->state.in_directiveg) - cpp_error (pfile, CPP_DL_ERROR, - "__COUNTER__ expanded inside directive with -fdirectives-only"); - number = pfile->counter++; - break; - - case BT_HAS_ATTRIBUTE: - number = pfile->cb.has_attribute (pfile, false); - break; - - case BT_HAS_STD_ATTRIBUTE: - number = pfile->cb.has_attribute (pfile, true); - break; - - case BT_HAS_BUILTIN: - number = pfile->cb.has_builtin (pfile); - break; - - case BT_HAS_INCLUDE: - case BT_HAS_INCLUDE_NEXT: - number = builtin_has_include (pfile, node, - node->value.builtin == BT_HAS_INCLUDE_NEXT); - break; - - case BT_RT_CAT: - result = evaluate_RT_CAT(pfile); - break; - - - } - - if (result == NULL) - { - /* 21 bytes holds all NUL-terminated unsigned 64-bit numbers. */ - result = _cpp_unaligned_alloc (pfile, 21); - sprintf ((char *) result, "%u", number); - } - - return result; -} - -/* Get an idempotent date. Either the cached value, the value from - source epoch, or failing that, the value from time(2). Use this - during compilation so that every time stamp is the same. */ -CPP_time_kind -cpp_get_date (cpp_reader *pfile, time_t *result) -{ - if (!pfile->time_stamp_kind) - { - int kind = 0; - if (pfile->cb.get_source_date_epoch) - { - /* Try reading the fixed epoch. */ - pfile->time_stamp = pfile->cb.get_source_date_epoch (pfile); - if (pfile->time_stamp != time_t (-1)) - kind = int (CPP_time_kind::FIXED); - } - - if (!kind) - { - /* Pedantically time_t (-1) is a legitimate value for - "number of seconds since the Epoch". It is a silly - time. */ - errno = 0; - pfile->time_stamp = time (nullptr); - /* Annoyingly a library could legally set errno and return a - valid time! Bad library! */ - if (pfile->time_stamp == time_t (-1) && errno) - kind = errno; - else - kind = int (CPP_time_kind::DYNAMIC); - } - - pfile->time_stamp_kind = kind; - } - - *result = pfile->time_stamp; - if (pfile->time_stamp_kind >= 0) - { - errno = pfile->time_stamp_kind; - return CPP_time_kind::UNKNOWN; - } - - return CPP_time_kind (pfile->time_stamp_kind); -} - -/* Convert builtin macros like __FILE__ to a token and push it on the - context stack. Also handles _Pragma, for which a new token may not - be created. Returns 1 if it generates a new token context, 0 to - return the token to the caller. LOC is the location of the expansion - point of the macro. */ -static int -builtin_macro (cpp_reader *pfile, cpp_hashnode *node, - location_t loc, location_t expand_loc) -{ - const uchar *buf; - size_t len; - char *nbuf; - - if (node->value.builtin == BT_PRAGMA) - { - /* Don't interpret _Pragma within directives. The standard is - not clear on this, but to me this makes most sense. - Similarly, don't interpret _Pragma inside expand_args, we might - need to stringize it later on. */ - if (pfile->state.in_directive || pfile->state.ignore__Pragma) - return 0; - - return _cpp_do__Pragma (pfile, loc); - } - - buf = _cpp_builtin_macro_text (pfile, node, expand_loc); - len = ustrlen (buf); - nbuf = (char *) alloca (len + 1); - memcpy (nbuf, buf, len); - nbuf[len]='\n'; - - cpp_push_buffer (pfile, (uchar *) nbuf, len, /* from_stage3 */ true); - _cpp_clean_line (pfile); - - /* Set pfile->cur_token as required by _cpp_lex_direct. */ - pfile->cur_token = _cpp_temp_token (pfile); - cpp_token *token = _cpp_lex_direct (pfile); - /* We should point to the expansion point of the builtin macro. */ - token->src_loc = loc; - if (pfile->context->tokens_kind == TOKENS_KIND_EXTENDED) - { - /* We are tracking tokens resulting from macro expansion. - Create a macro line map and generate a virtual location for - the token resulting from the expansion of the built-in - macro. */ - location_t *virt_locs = NULL; - _cpp_buff *token_buf = tokens_buff_new (pfile, 1, &virt_locs); - const line_map_macro * map = - linemap_enter_macro (pfile->line_table, node, loc, 1); - tokens_buff_add_token (token_buf, virt_locs, token, - pfile->line_table->builtin_location, - pfile->line_table->builtin_location, - map, /*macro_token_index=*/0); - push_extended_tokens_context (pfile, node, token_buf, virt_locs, - (const cpp_token **)token_buf->base, - 1); - } - else - _cpp_push_token_context (pfile, NULL, token, 1); - if (pfile->buffer->cur != pfile->buffer->rlimit) - cpp_error (pfile, CPP_DL_ICE, "invalid built-in macro \"%s\"", - NODE_NAME (node)); - _cpp_pop_buffer (pfile); - - return 1; -} - -/* Copies SRC, of length LEN, to DEST, adding backslashes before all - backslashes and double quotes. DEST must be of sufficient size. - Returns a pointer to the end of the string. */ -uchar * -cpp_quote_string (uchar *dest, const uchar *src, unsigned int len) -{ - while (len--) - { - uchar c = *src++; - - switch (c) - { - case '\n': - /* Naked LF can appear in raw string literals */ - c = 'n'; - /* FALLTHROUGH */ - - case '\\': - case '"': - *dest++ = '\\'; - /* FALLTHROUGH */ - - default: - *dest++ = c; - } - } - - return dest; -} - -/* Convert a token sequence FIRST to FIRST+COUNT-1 to a single string token - according to the rules of the ISO C #-operator. */ -static const cpp_token * -stringify_arg (cpp_reader *pfile, const cpp_token **first, unsigned int count) -{ - unsigned char *dest; - unsigned int i, escape_it, backslash_count = 0; - const cpp_token *source = NULL; - size_t len; - - if (BUFF_ROOM (pfile->u_buff) < 3) - _cpp_extend_buff (pfile, &pfile->u_buff, 3); - dest = BUFF_FRONT (pfile->u_buff); - *dest++ = '"'; - - /* Loop, reading in the argument's tokens. */ - for (i = 0; i < count; i++) - { - const cpp_token *token = first[i]; - - if (token->type == CPP_PADDING) - { - if (source == NULL - || (!(source->flags & PREV_WHITE) - && token->val.source == NULL)) - source = token->val.source; - continue; - } - - escape_it = (token->type == CPP_STRING || token->type == CPP_CHAR - || token->type == CPP_WSTRING || token->type == CPP_WCHAR - || token->type == CPP_STRING32 || token->type == CPP_CHAR32 - || token->type == CPP_STRING16 || token->type == CPP_CHAR16 - || token->type == CPP_UTF8STRING || token->type == CPP_UTF8CHAR - || cpp_userdef_string_p (token->type) - || cpp_userdef_char_p (token->type)); - - /* Room for each char being written in octal, initial space and - final quote and NUL. */ - len = cpp_token_len (token); - if (escape_it) - len *= 4; - len += 3; - - if ((size_t) (BUFF_LIMIT (pfile->u_buff) - dest) < len) - { - size_t len_so_far = dest - BUFF_FRONT (pfile->u_buff); - _cpp_extend_buff (pfile, &pfile->u_buff, len); - dest = BUFF_FRONT (pfile->u_buff) + len_so_far; - } - - /* Leading white space? */ - if (dest - 1 != BUFF_FRONT (pfile->u_buff)) - { - if (source == NULL) - source = token; - if (source->flags & PREV_WHITE) - *dest++ = ' '; - } - source = NULL; - - if (escape_it) - { - _cpp_buff *buff = _cpp_get_buff (pfile, len); - unsigned char *buf = BUFF_FRONT (buff); - len = cpp_spell_token (pfile, token, buf, true) - buf; - dest = cpp_quote_string (dest, buf, len); - _cpp_release_buff (pfile, buff); - } - else - dest = cpp_spell_token (pfile, token, dest, true); - - if (token->type == CPP_OTHER && token->val.str.text[0] == '\\') - backslash_count++; - else - backslash_count = 0; - } - - /* Ignore the final \ of invalid string literals. */ - if (backslash_count & 1) - { - cpp_error (pfile, CPP_DL_WARNING, - "invalid string literal, ignoring final '\\'"); - dest--; - } - - /* Commit the memory, including NUL, and return the token. */ - *dest++ = '"'; - len = dest - BUFF_FRONT (pfile->u_buff); - BUFF_FRONT (pfile->u_buff) = dest + 1; - return new_string_token (pfile, dest - len, len); -} - -/* Try to paste two tokens. On success, return nonzero. In any - case, PLHS is updated to point to the pasted token, which is - guaranteed to not have the PASTE_LEFT flag set. LOCATION is - the virtual location used for error reporting. */ -static bool -paste_tokens (cpp_reader *pfile, location_t location, - const cpp_token **plhs, const cpp_token *rhs) -{ - unsigned char *buf, *end, *lhsend; - cpp_token *lhs; - unsigned int len; - - len = cpp_token_len (*plhs) + cpp_token_len (rhs) + 2; - buf = (unsigned char *) alloca (len); - end = lhsend = cpp_spell_token (pfile, *plhs, buf, true); - - /* Avoid comment headers, since they are still processed in stage 3. - It is simpler to insert a space here, rather than modifying the - lexer to ignore comments in some circumstances. Simply returning - false doesn't work, since we want to clear the PASTE_LEFT flag. */ - if ((*plhs)->type == CPP_DIV && rhs->type != CPP_EQ) - *end++ = ' '; - /* In one obscure case we might see padding here. */ - if (rhs->type != CPP_PADDING) - end = cpp_spell_token (pfile, rhs, end, true); - *end = '\n'; - - cpp_push_buffer (pfile, buf, end - buf, /* from_stage3 */ true); - _cpp_clean_line (pfile); - - /* Set pfile->cur_token as required by _cpp_lex_direct. */ - pfile->cur_token = _cpp_temp_token (pfile); - lhs = _cpp_lex_direct (pfile); - if (pfile->buffer->cur != pfile->buffer->rlimit) - { - location_t saved_loc = lhs->src_loc; - - _cpp_pop_buffer (pfile); - - unsigned char *rhsstart = lhsend; - if ((*plhs)->type == CPP_DIV && rhs->type != CPP_EQ) - rhsstart++; - - /* We have to remove the PASTE_LEFT flag from the old lhs, but - we want to keep the new location. */ - *lhs = **plhs; - *plhs = lhs; - lhs->src_loc = saved_loc; - lhs->flags &= ~PASTE_LEFT; - - /* Mandatory error for all apart from assembler. */ - if (CPP_OPTION (pfile, lang) != CLK_ASM) - cpp_error_with_line (pfile, CPP_DL_ERROR, location, 0, - "pasting \"%.*s\" and \"%.*s\" does not give " - "a valid preprocessing token", - (int) (lhsend - buf), buf, - (int) (end - rhsstart), rhsstart); - return false; - } - - lhs->flags |= (*plhs)->flags & (PREV_WHITE | PREV_FALLTHROUGH); - *plhs = lhs; - _cpp_pop_buffer (pfile); - return true; -} - -/* Handles an arbitrarily long sequence of ## operators, with initial - operand LHS. This implementation is left-associative, - non-recursive, and finishes a paste before handling succeeding - ones. If a paste fails, we back up to the RHS of the failing ## - operator before pushing the context containing the result of prior - successful pastes, with the effect that the RHS appears in the - output stream after the pasted LHS normally. */ -static void -paste_all_tokens (cpp_reader *pfile, const cpp_token *lhs) -{ - const cpp_token *rhs = NULL; - cpp_context *context = pfile->context; - location_t virt_loc = 0; - - /* We are expanding a macro and we must have been called on a token - that appears at the left hand side of a ## operator. */ - if (macro_of_context (pfile->context) == NULL - || (!(lhs->flags & PASTE_LEFT))) - abort (); - - if (context->tokens_kind == TOKENS_KIND_EXTENDED) - /* The caller must have called consume_next_token_from_context - right before calling us. That has incremented the pointer to - the current virtual location. So it now points to the location - of the token that comes right after *LHS. We want the - resulting pasted token to have the location of the current - *LHS, though. */ - virt_loc = context->c.mc->cur_virt_loc[-1]; - else - /* We are not tracking macro expansion. So the best virtual - location we can get here is the expansion point of the macro we - are currently expanding. */ - virt_loc = pfile->invocation_location; - - do - { - /* Take the token directly from the current context. We can do - this, because we are in the replacement list of either an - object-like macro, or a function-like macro with arguments - inserted. In either case, the constraints to #define - guarantee we have at least one more token. */ - if (context->tokens_kind == TOKENS_KIND_DIRECT) - rhs = FIRST (context).token++; - else if (context->tokens_kind == TOKENS_KIND_INDIRECT) - rhs = *FIRST (context).ptoken++; - else if (context->tokens_kind == TOKENS_KIND_EXTENDED) - { - /* So we are in presence of an extended token context, which - means that each token in this context has a virtual - location attached to it. So let's not forget to update - the pointer to the current virtual location of the - current token when we update the pointer to the current - token */ - - rhs = *FIRST (context).ptoken++; - /* context->c.mc must be non-null, as if we were not in a - macro context, context->tokens_kind could not be equal to - TOKENS_KIND_EXTENDED. */ - context->c.mc->cur_virt_loc++; - } - - if (rhs->type == CPP_PADDING) - { - if (rhs->flags & PASTE_LEFT) - abort (); - } - if (!paste_tokens (pfile, virt_loc, &lhs, rhs)) - { - _cpp_backup_tokens (pfile, 1); - break; - } - } - while (rhs->flags & PASTE_LEFT); - - /* Put the resulting token in its own context. */ - if (context->tokens_kind == TOKENS_KIND_EXTENDED) - { - location_t *virt_locs = NULL; - _cpp_buff *token_buf = tokens_buff_new (pfile, 1, &virt_locs); - tokens_buff_add_token (token_buf, virt_locs, lhs, - virt_loc, 0, NULL, 0); - push_extended_tokens_context (pfile, context->c.mc->macro_node, - token_buf, virt_locs, - (const cpp_token **)token_buf->base, 1); - } - else - _cpp_push_token_context (pfile, NULL, lhs, 1); -} - -/* Returns TRUE if the number of arguments ARGC supplied in an - invocation of the MACRO referenced by NODE is valid. An empty - invocation to a macro with no parameters should pass ARGC as zero. - - Note that MACRO cannot necessarily be deduced from NODE, in case - NODE was redefined whilst collecting arguments. */ -bool -_cpp_arguments_ok (cpp_reader *pfile, cpp_macro *macro, const cpp_hashnode *node, unsigned int argc) -{ - if (argc == macro->paramc) - return true; - - if (argc < macro->paramc) - { - /* In C++20 (here the va_opt flag is used), and also as a GNU - extension, variadic arguments are allowed to not appear in - the invocation at all. - e.g. #define debug(format, args...) something - debug("string"); - - This is exactly the same as if an empty variadic list had been - supplied - debug("string", ). */ - - if (argc + 1 == macro->paramc && macro->variadic) - { - if (CPP_PEDANTIC (pfile) && ! macro->syshdr - && ! CPP_OPTION (pfile, va_opt)) - { - if (CPP_OPTION (pfile, cplusplus)) - cpp_error (pfile, CPP_DL_PEDWARN, - "ISO C++11 requires at least one argument " - "for the \"...\" in a variadic macro"); - else - cpp_error (pfile, CPP_DL_PEDWARN, - "ISO C99 requires at least one argument " - "for the \"...\" in a variadic macro"); - } - return true; - } - - cpp_error (pfile, CPP_DL_ERROR, - "macro \"%s\" requires %u arguments, but only %u given", - NODE_NAME (node), macro->paramc, argc); - } - else - cpp_error (pfile, CPP_DL_ERROR, - "macro \"%s\" passed %u arguments, but takes just %u", - NODE_NAME (node), argc, macro->paramc); - - if (macro->line > RESERVED_LOCATION_COUNT) - cpp_error_at (pfile, CPP_DL_NOTE, macro->line, "macro \"%s\" defined here", - NODE_NAME (node)); - - return false; -} - -/* Reads and returns the arguments to a function-like macro - invocation. Assumes the opening parenthesis has been processed. - If there is an error, emits an appropriate diagnostic and returns - NULL. Each argument is terminated by a CPP_EOF token, for the - future benefit of expand_arg(). If there are any deferred - #pragma directives among macro arguments, store pointers to the - CPP_PRAGMA ... CPP_PRAGMA_EOL tokens into *PRAGMA_BUFF buffer. - - What is returned is the buffer that contains the memory allocated - to hold the macro arguments. NODE is the name of the macro this - function is dealing with. If NUM_ARGS is non-NULL, *NUM_ARGS is - set to the actual number of macro arguments allocated in the - returned buffer. */ -static _cpp_buff * -collect_args (cpp_reader *pfile, const cpp_hashnode *node, - _cpp_buff **pragma_buff, unsigned *num_args) -{ - _cpp_buff *buff, *base_buff; - cpp_macro *macro; - macro_arg *args, *arg; - const cpp_token *token; - unsigned int argc; - location_t virt_loc; - bool track_macro_expansion_p = CPP_OPTION (pfile, track_macro_expansion); - unsigned num_args_alloced = 0; - - macro = node->value.macro; - if (macro->paramc) - argc = macro->paramc; - else - argc = 1; - -#define DEFAULT_NUM_TOKENS_PER_MACRO_ARG 50 -#define ARG_TOKENS_EXTENT 1000 - - buff = _cpp_get_buff (pfile, argc * (DEFAULT_NUM_TOKENS_PER_MACRO_ARG - * sizeof (cpp_token *) - + sizeof (macro_arg))); - base_buff = buff; - args = (macro_arg *) buff->base; - memset (args, 0, argc * sizeof (macro_arg)); - buff->cur = (unsigned char *) &args[argc]; - arg = args, argc = 0; - - /* Collect the tokens making up each argument. We don't yet know - how many arguments have been supplied, whether too many or too - few. Hence the slightly bizarre usage of "argc" and "arg". */ - do - { - unsigned int paren_depth = 0; - unsigned int ntokens = 0; - unsigned virt_locs_capacity = DEFAULT_NUM_TOKENS_PER_MACRO_ARG; - num_args_alloced++; - - argc++; - arg->first = (const cpp_token **) buff->cur; - if (track_macro_expansion_p) - { - virt_locs_capacity = DEFAULT_NUM_TOKENS_PER_MACRO_ARG; - arg->virt_locs = XNEWVEC (location_t, - virt_locs_capacity); - } - - for (;;) - { - /* Require space for 2 new tokens (including a CPP_EOF). */ - if ((unsigned char *) &arg->first[ntokens + 2] > buff->limit) - { - buff = _cpp_append_extend_buff (pfile, buff, - ARG_TOKENS_EXTENT - * sizeof (cpp_token *)); - arg->first = (const cpp_token **) buff->cur; - } - if (track_macro_expansion_p - && (ntokens + 2 > virt_locs_capacity)) - { - virt_locs_capacity += ARG_TOKENS_EXTENT; - arg->virt_locs = XRESIZEVEC (location_t, - arg->virt_locs, - virt_locs_capacity); - } - - token = cpp_get_token_1 (pfile, &virt_loc); - - if (token->type == CPP_PADDING) - { - /* Drop leading padding. */ - if (ntokens == 0) - continue; - } - else if (token->type == CPP_OPEN_PAREN) - paren_depth++; - else if (token->type == CPP_CLOSE_PAREN) - { - if (paren_depth-- == 0) - break; - } - else if (token->type == CPP_COMMA) - { - /* A comma does not terminate an argument within - parentheses or as part of a variable argument. */ - if (paren_depth == 0 - && ! (macro->variadic && argc == macro->paramc)) - break; - } - else if (token->type == CPP_EOF - || (token->type == CPP_HASH && token->flags & BOL)) - break; - else if (token->type == CPP_PRAGMA && !(token->flags & PRAGMA_OP)) - { - cpp_token *newtok = _cpp_temp_token (pfile); - - /* CPP_PRAGMA token lives in directive_result, which will - be overwritten on the next directive. */ - *newtok = *token; - token = newtok; - do - { - if (*pragma_buff == NULL - || BUFF_ROOM (*pragma_buff) < sizeof (cpp_token *)) - { - _cpp_buff *next; - if (*pragma_buff == NULL) - *pragma_buff - = _cpp_get_buff (pfile, 32 * sizeof (cpp_token *)); - else - { - next = *pragma_buff; - *pragma_buff - = _cpp_get_buff (pfile, - (BUFF_FRONT (*pragma_buff) - - (*pragma_buff)->base) * 2); - (*pragma_buff)->next = next; - } - } - *(const cpp_token **) BUFF_FRONT (*pragma_buff) = token; - BUFF_FRONT (*pragma_buff) += sizeof (cpp_token *); - if (token->type == CPP_PRAGMA_EOL) - break; - token = cpp_get_token_1 (pfile, &virt_loc); - } - while (token->type != CPP_EOF); - - /* In deferred pragmas parsing_args and prevent_expansion - had been changed, reset it. */ - pfile->state.parsing_args = 2; - pfile->state.prevent_expansion = 1; - - if (token->type == CPP_EOF) - break; - else - continue; - } - set_arg_token (arg, token, virt_loc, - ntokens, MACRO_ARG_TOKEN_NORMAL, - CPP_OPTION (pfile, track_macro_expansion)); - ntokens++; - } - - /* Drop trailing padding. */ - while (ntokens > 0 && arg->first[ntokens - 1]->type == CPP_PADDING) - ntokens--; - - arg->count = ntokens; - /* Append an EOF to mark end-of-argument. */ - set_arg_token (arg, &pfile->endarg, token->src_loc, - ntokens, MACRO_ARG_TOKEN_NORMAL, - CPP_OPTION (pfile, track_macro_expansion)); - - /* Terminate the argument. Excess arguments loop back and - overwrite the final legitimate argument, before failing. */ - if (argc <= macro->paramc) - { - buff->cur = (unsigned char *) &arg->first[ntokens + 1]; - if (argc != macro->paramc) - arg++; - } - } - while (token->type != CPP_CLOSE_PAREN && token->type != CPP_EOF); - - if (token->type == CPP_EOF) - { - /* Unless the EOF is marking the end of an argument, it's a fake - one from the end of a file that _cpp_clean_line will not have - advanced past. */ - if (token == &pfile->endarg) - _cpp_backup_tokens (pfile, 1); - cpp_error (pfile, CPP_DL_ERROR, - "unterminated argument list invoking macro \"%s\"", - NODE_NAME (node)); - } - else - { - /* A single empty argument is counted as no argument. */ - if (argc == 1 && macro->paramc == 0 && args[0].count == 0) - argc = 0; - if (_cpp_arguments_ok (pfile, macro, node, argc)) - { - /* GCC has special semantics for , ## b where b is a varargs - parameter: we remove the comma if b was omitted entirely. - If b was merely an empty argument, the comma is retained. - If the macro takes just one (varargs) parameter, then we - retain the comma only if we are standards conforming. - - If FIRST is NULL replace_args () swallows the comma. */ - if (macro->variadic && (argc < macro->paramc - || (argc == 1 && args[0].count == 0 - && !CPP_OPTION (pfile, std)))) - args[macro->paramc - 1].first = NULL; - if (num_args) - *num_args = num_args_alloced; - return base_buff; - } - } - - /* An error occurred. */ - _cpp_release_buff (pfile, base_buff); - return NULL; -} - -/* Search for an opening parenthesis to the macro of NODE, in such a - way that, if none is found, we don't lose the information in any - intervening padding tokens. If we find the parenthesis, collect - the arguments and return the buffer containing them. PRAGMA_BUFF - argument is the same as in collect_args. If NUM_ARGS is non-NULL, - *NUM_ARGS is set to the number of arguments contained in the - returned buffer. */ -static _cpp_buff * -funlike_invocation_p (cpp_reader *pfile, cpp_hashnode *node, - _cpp_buff **pragma_buff, unsigned *num_args) -{ - const cpp_token *token, *padding = NULL; - - for (;;) - { - token = cpp_get_token (pfile); - if (token->type != CPP_PADDING) - break; - gcc_assert ((token->flags & PREV_WHITE) == 0); - if (padding == NULL - || padding->val.source == NULL - || (!(padding->val.source->flags & PREV_WHITE) - && token->val.source == NULL)) - padding = token; - } - - if (token->type == CPP_OPEN_PAREN) - { - pfile->state.parsing_args = 2; - return collect_args (pfile, node, pragma_buff, num_args); - } - - /* Back up. A CPP_EOF is either an EOF from an argument we're - expanding, or a fake one from lex_direct. We want to backup the - former, but not the latter. We may have skipped padding, in - which case backing up more than one token when expanding macros - is in general too difficult. We re-insert it in its own - context. */ - if (token->type != CPP_EOF || token == &pfile->endarg) - { - _cpp_backup_tokens (pfile, 1); - if (padding) - _cpp_push_token_context (pfile, NULL, padding, 1); - } - - return NULL; -} - -/* Return the real number of tokens in the expansion of MACRO. */ -static inline unsigned int -macro_real_token_count (const cpp_macro *macro) -{ - if (__builtin_expect (!macro->extra_tokens, true)) - return macro->count; - - for (unsigned i = macro->count; i--;) - if (macro->exp.tokens[i].type != CPP_PASTE) - return i + 1; - - return 0; -} - -/* Push the context of a macro with hash entry NODE onto the context - stack. If we can successfully expand the macro, we push a context - containing its yet-to-be-rescanned replacement list and return one. - If there were additionally any unexpanded deferred #pragma - directives among macro arguments, push another context containing - the pragma tokens before the yet-to-be-rescanned replacement list - and return two. Otherwise, we don't push a context and return - zero. LOCATION is the location of the expansion point of the - macro. */ -static int -enter_macro_context (cpp_reader *pfile, cpp_hashnode *node, - const cpp_token *result, location_t location) -{ - /* The presence of a macro invalidates a file's controlling macro. */ - pfile->mi_valid = false; - - pfile->state.angled_headers = false; - - /* From here to when we push the context for the macro later down - this function, we need to flag the fact that we are about to - expand a macro. This is useful when -ftrack-macro-expansion is - turned off. In that case, we need to record the location of the - expansion point of the top-most macro we are about to to expand, - into pfile->invocation_location. But we must not record any such - location once the process of expanding the macro starts; that is, - we must not do that recording between now and later down this - function where set this flag to FALSE. */ - pfile->about_to_expand_macro_p = true; - - if (cpp_user_macro_p (node)) - { - cpp_macro *macro = node->value.macro; - _cpp_buff *pragma_buff = NULL; - - if (macro->fun_like) - { - _cpp_buff *buff; - unsigned num_args = 0; - - pfile->state.prevent_expansion++; - pfile->keep_tokens++; - pfile->state.parsing_args = 1; - buff = funlike_invocation_p (pfile, node, &pragma_buff, - &num_args); - pfile->state.parsing_args = 0; - pfile->keep_tokens--; - pfile->state.prevent_expansion--; - - if (buff == NULL) - { - if (CPP_WTRADITIONAL (pfile) && ! node->value.macro->syshdr) - cpp_warning (pfile, CPP_W_TRADITIONAL, - "function-like macro \"%s\" must be used with arguments in traditional C", - NODE_NAME (node)); - - if (pragma_buff) - _cpp_release_buff (pfile, pragma_buff); - - pfile->about_to_expand_macro_p = false; - return 0; - } - - if (macro->paramc > 0) - replace_args (pfile, node, macro, - (macro_arg *) buff->base, - location); - /* Free the memory used by the arguments of this - function-like macro. This memory has been allocated by - funlike_invocation_p and by replace_args. */ - delete_macro_args (buff, num_args); - } - - /* Disable the macro within its expansion. */ - node->flags |= NODE_DISABLED; - - /* Laziness can only affect the expansion tokens of the macro, - not its fun-likeness or parameters. */ - _cpp_maybe_notify_macro_use (pfile, node, location); - if (pfile->cb.used) - pfile->cb.used (pfile, location, node); - - macro->used = 1; - - if (macro->paramc == 0) - { - unsigned tokens_count = macro_real_token_count (macro); - if (CPP_OPTION (pfile, track_macro_expansion)) - { - unsigned int i; - const cpp_token *src = macro->exp.tokens; - const line_map_macro *map; - location_t *virt_locs = NULL; - _cpp_buff *macro_tokens - = tokens_buff_new (pfile, tokens_count, &virt_locs); - - /* Create a macro map to record the locations of the - tokens that are involved in the expansion. LOCATION - is the location of the macro expansion point. */ - map = linemap_enter_macro (pfile->line_table, - node, location, tokens_count); - for (i = 0; i < tokens_count; ++i) - { - tokens_buff_add_token (macro_tokens, virt_locs, - src, src->src_loc, - src->src_loc, map, i); - ++src; - } - push_extended_tokens_context (pfile, node, - macro_tokens, - virt_locs, - (const cpp_token **) - macro_tokens->base, - tokens_count); - } - else - _cpp_push_token_context (pfile, node, macro->exp.tokens, - tokens_count); - num_macro_tokens_counter += tokens_count; - } - - if (pragma_buff) - { - if (!pfile->state.in_directive) - _cpp_push_token_context (pfile, NULL, - padding_token (pfile, result), 1); - do - { - unsigned tokens_count; - _cpp_buff *tail = pragma_buff->next; - pragma_buff->next = NULL; - tokens_count = ((const cpp_token **) BUFF_FRONT (pragma_buff) - - (const cpp_token **) pragma_buff->base); - push_ptoken_context (pfile, NULL, pragma_buff, - (const cpp_token **) pragma_buff->base, - tokens_count); - pragma_buff = tail; - if (!CPP_OPTION (pfile, track_macro_expansion)) - num_macro_tokens_counter += tokens_count; - - } - while (pragma_buff != NULL); - pfile->about_to_expand_macro_p = false; - return 2; - } - - pfile->about_to_expand_macro_p = false; - return 1; - } - - pfile->about_to_expand_macro_p = false; - /* Handle built-in macros and the _Pragma operator. */ - { - location_t expand_loc; - - if (/* The top-level macro invocation that triggered the expansion - we are looking at is with a function-like user macro ... */ - cpp_fun_like_macro_p (pfile->top_most_macro_node) - /* ... and we are tracking the macro expansion. */ - && CPP_OPTION (pfile, track_macro_expansion)) - /* Then the location of the end of the macro invocation is the - location of the expansion point of this macro. */ - expand_loc = location; - else - /* Otherwise, the location of the end of the macro invocation is - the location of the expansion point of that top-level macro - invocation. */ - expand_loc = pfile->invocation_location; - - return builtin_macro (pfile, node, location, expand_loc); - } -} - -/* De-allocate the memory used by BUFF which is an array of instances - of macro_arg. NUM_ARGS is the number of instances of macro_arg - present in BUFF. */ -static void -delete_macro_args (_cpp_buff *buff, unsigned num_args) -{ - macro_arg *macro_args; - unsigned i; - - if (buff == NULL) - return; - - macro_args = (macro_arg *) buff->base; - - /* Walk instances of macro_arg to free their expanded tokens as well - as their macro_arg::virt_locs members. */ - for (i = 0; i < num_args; ++i) - { - if (macro_args[i].expanded) - { - free (macro_args[i].expanded); - macro_args[i].expanded = NULL; - } - if (macro_args[i].virt_locs) - { - free (macro_args[i].virt_locs); - macro_args[i].virt_locs = NULL; - } - if (macro_args[i].expanded_virt_locs) - { - free (macro_args[i].expanded_virt_locs); - macro_args[i].expanded_virt_locs = NULL; - } - } - _cpp_free_buff (buff); -} - -/* Set the INDEXth token of the macro argument ARG. TOKEN is the token - to set, LOCATION is its virtual location. "Virtual" location means - the location that encodes loci across macro expansion. Otherwise - it has to be TOKEN->SRC_LOC. KIND is the kind of tokens the - argument ARG is supposed to contain. Note that ARG must be - tailored so that it has enough room to contain INDEX + 1 numbers of - tokens, at least. */ -static void -set_arg_token (macro_arg *arg, const cpp_token *token, - location_t location, size_t index, - enum macro_arg_token_kind kind, - bool track_macro_exp_p) -{ - const cpp_token **token_ptr; - location_t *loc = NULL; - - token_ptr = - arg_token_ptr_at (arg, index, kind, - track_macro_exp_p ? &loc : NULL); - *token_ptr = token; - - if (loc != NULL) - { - /* We can't set the location of a stringified argument - token and we can't set any location if we aren't tracking - macro expansion locations. */ - gcc_checking_assert (kind != MACRO_ARG_TOKEN_STRINGIFIED - && track_macro_exp_p); - *loc = location; - } -} - -/* Get the pointer to the location of the argument token of the - function-like macro argument ARG. This function must be called - only when we -ftrack-macro-expansion is on. */ -static const location_t * -get_arg_token_location (const macro_arg *arg, - enum macro_arg_token_kind kind) -{ - const location_t *loc = NULL; - const cpp_token **token_ptr = - arg_token_ptr_at (arg, 0, kind, (location_t **) &loc); - - if (token_ptr == NULL) - return NULL; - - return loc; -} - -/* Return the pointer to the INDEXth token of the macro argument ARG. - KIND specifies the kind of token the macro argument ARG contains. - If VIRT_LOCATION is non NULL, *VIRT_LOCATION is set to the address - of the virtual location of the returned token if the - -ftrack-macro-expansion flag is on; otherwise, it's set to the - spelling location of the returned token. */ -static const cpp_token ** -arg_token_ptr_at (const macro_arg *arg, size_t index, - enum macro_arg_token_kind kind, - location_t **virt_location) -{ - const cpp_token **tokens_ptr = NULL; - - switch (kind) - { - case MACRO_ARG_TOKEN_NORMAL: - tokens_ptr = arg->first; - break; - case MACRO_ARG_TOKEN_STRINGIFIED: - tokens_ptr = (const cpp_token **) &arg->stringified; - break; - case MACRO_ARG_TOKEN_EXPANDED: - tokens_ptr = arg->expanded; - break; - } - - if (tokens_ptr == NULL) - /* This can happen for e.g, an empty token argument to a - funtion-like macro. */ - return tokens_ptr; - - if (virt_location) - { - if (kind == MACRO_ARG_TOKEN_NORMAL) - *virt_location = &arg->virt_locs[index]; - else if (kind == MACRO_ARG_TOKEN_EXPANDED) - *virt_location = &arg->expanded_virt_locs[index]; - else if (kind == MACRO_ARG_TOKEN_STRINGIFIED) - *virt_location = - (location_t *) &tokens_ptr[index]->src_loc; - } - return &tokens_ptr[index]; -} - -/* Initialize an iterator so that it iterates over the tokens of a - function-like macro argument. KIND is the kind of tokens we want - ITER to iterate over. TOKEN_PTR points the first token ITER will - iterate over. */ -static void -macro_arg_token_iter_init (macro_arg_token_iter *iter, - bool track_macro_exp_p, - enum macro_arg_token_kind kind, - const macro_arg *arg, - const cpp_token **token_ptr) -{ - iter->track_macro_exp_p = track_macro_exp_p; - iter->kind = kind; - iter->token_ptr = token_ptr; - /* Unconditionally initialize this so that the compiler doesn't warn - about iter->location_ptr being possibly uninitialized later after - this code has been inlined somewhere. */ - iter->location_ptr = NULL; - if (track_macro_exp_p) - iter->location_ptr = get_arg_token_location (arg, kind); -#if CHECKING_P - iter->num_forwards = 0; - if (track_macro_exp_p - && token_ptr != NULL - && iter->location_ptr == NULL) - abort (); -#endif -} - -/* Move the iterator one token forward. Note that if IT was - initialized on an argument that has a stringified token, moving it - forward doesn't make sense as a stringified token is essentially one - string. */ -static void -macro_arg_token_iter_forward (macro_arg_token_iter *it) -{ - switch (it->kind) - { - case MACRO_ARG_TOKEN_NORMAL: - case MACRO_ARG_TOKEN_EXPANDED: - it->token_ptr++; - if (it->track_macro_exp_p) - it->location_ptr++; - break; - case MACRO_ARG_TOKEN_STRINGIFIED: -#if CHECKING_P - if (it->num_forwards > 0) - abort (); -#endif - break; - } - -#if CHECKING_P - it->num_forwards++; -#endif -} - -/* Return the token pointed to by the iterator. */ -static const cpp_token * -macro_arg_token_iter_get_token (const macro_arg_token_iter *it) -{ -#if CHECKING_P - if (it->kind == MACRO_ARG_TOKEN_STRINGIFIED - && it->num_forwards > 0) - abort (); -#endif - if (it->token_ptr == NULL) - return NULL; - return *it->token_ptr; -} - -/* Return the location of the token pointed to by the iterator.*/ -static location_t -macro_arg_token_iter_get_location (const macro_arg_token_iter *it) -{ -#if CHECKING_P - if (it->kind == MACRO_ARG_TOKEN_STRINGIFIED - && it->num_forwards > 0) - abort (); -#endif - if (it->track_macro_exp_p) - return *it->location_ptr; - else - return (*it->token_ptr)->src_loc; -} - -/* Return the index of a token [resulting from macro expansion] inside - the total list of tokens resulting from a given macro - expansion. The index can be different depending on whether if we - want each tokens resulting from function-like macro arguments - expansion to have a different location or not. - - E.g, consider this function-like macro: - - #define M(x) x - 3 - - Then consider us "calling" it (and thus expanding it) like: - - M(1+4) - - It will be expanded into: - - 1+4-3 - - Let's consider the case of the token '4'. - - Its index can be 2 (it's the third token of the set of tokens - resulting from the expansion) or it can be 0 if we consider that - all tokens resulting from the expansion of the argument "1+2" have - the same index, which is 0. In this later case, the index of token - '-' would then be 1 and the index of token '3' would be 2. - - The later case is useful to use less memory e.g, for the case of - the user using the option -ftrack-macro-expansion=1. - - ABSOLUTE_TOKEN_INDEX is the index of the macro argument token we - are interested in. CUR_REPLACEMENT_TOKEN is the token of the macro - parameter (inside the macro replacement list) that corresponds to - the macro argument for which ABSOLUTE_TOKEN_INDEX is a token index - of. - - If we refer to the example above, for the '4' argument token, - ABSOLUTE_TOKEN_INDEX would be set to 2, and CUR_REPLACEMENT_TOKEN - would be set to the token 'x', in the replacement list "x - 3" of - macro M. - - This is a subroutine of replace_args. */ -inline static unsigned -expanded_token_index (cpp_reader *pfile, cpp_macro *macro, - const cpp_token *cur_replacement_token, - unsigned absolute_token_index) -{ - if (CPP_OPTION (pfile, track_macro_expansion) > 1) - return absolute_token_index; - return cur_replacement_token - macro->exp.tokens; -} - -/* Copy whether PASTE_LEFT is set from SRC to *PASTE_FLAG. */ - -static void -copy_paste_flag (cpp_reader *pfile, const cpp_token **paste_flag, - const cpp_token *src) -{ - cpp_token *token = _cpp_temp_token (pfile); - token->type = (*paste_flag)->type; - token->val = (*paste_flag)->val; - if (src->flags & PASTE_LEFT) - token->flags = (*paste_flag)->flags | PASTE_LEFT; - else - token->flags = (*paste_flag)->flags & ~PASTE_LEFT; - *paste_flag = token; -} - -/* True IFF the last token emitted into BUFF (if any) is PTR. */ - -static bool -last_token_is (_cpp_buff *buff, const cpp_token **ptr) -{ - return (ptr && tokens_buff_last_token_ptr (buff) == ptr); -} - -/* Replace the parameters in a function-like macro of NODE with the - actual ARGS, and place the result in a newly pushed token context. - Expand each argument before replacing, unless it is operated upon - by the # or ## operators. EXPANSION_POINT_LOC is the location of - the expansion point of the macro. E.g, the location of the - function-like macro invocation. */ -static void -replace_args (cpp_reader *pfile, cpp_hashnode *node, cpp_macro *macro, - macro_arg *args, location_t expansion_point_loc) -{ - unsigned int i, total; - const cpp_token *src, *limit; - const cpp_token **first = NULL; - macro_arg *arg; - _cpp_buff *buff = NULL; - location_t *virt_locs = NULL; - unsigned int exp_count; - const line_map_macro *map = NULL; - int track_macro_exp; - - /* First, fully macro-expand arguments, calculating the number of - tokens in the final expansion as we go. The ordering of the if - statements below is subtle; we must handle stringification before - pasting. */ - - /* EXP_COUNT is the number of tokens in the macro replacement - list. TOTAL is the number of tokens /after/ macro parameters - have been replaced by their arguments. */ - exp_count = macro_real_token_count (macro); - total = exp_count; - limit = macro->exp.tokens + exp_count; - - for (src = macro->exp.tokens; src < limit; src++) - if (src->type == CPP_MACRO_ARG) - { - /* Leading and trailing padding tokens. */ - total += 2; - /* Account for leading and padding tokens in exp_count too. - This is going to be important later down this function, - when we want to handle the case of (track_macro_exp < - 2). */ - exp_count += 2; - - /* We have an argument. If it is not being stringified or - pasted it is macro-replaced before insertion. */ - arg = &args[src->val.macro_arg.arg_no - 1]; - - if (src->flags & STRINGIFY_ARG) - { - if (!arg->stringified) - arg->stringified = stringify_arg (pfile, arg->first, arg->count); - } - else if ((src->flags & PASTE_LEFT) - || (src != macro->exp.tokens && (src[-1].flags & PASTE_LEFT))) - total += arg->count - 1; - else - { - if (!arg->expanded) - expand_arg (pfile, arg); - total += arg->expanded_count - 1; - } - } - - /* When the compiler is called with the -ftrack-macro-expansion - flag, we need to keep track of the location of each token that - results from macro expansion. - - A token resulting from macro expansion is not a new token. It is - simply the same token as the token coming from the macro - definition. The new things that are allocated are the buffer - that holds the tokens resulting from macro expansion and a new - location that records many things like the locus of the expansion - point as well as the original locus inside the definition of the - macro. This location is called a virtual location. - - So the buffer BUFF holds a set of cpp_token*, and the buffer - VIRT_LOCS holds the virtual locations of the tokens held by BUFF. - - Both of these two buffers are going to be hung off of the macro - context, when the latter is pushed. The memory allocated to - store the tokens and their locations is going to be freed once - the context of macro expansion is popped. - - As far as tokens are concerned, the memory overhead of - -ftrack-macro-expansion is proportional to the number of - macros that get expanded multiplied by sizeof (location_t). - The good news is that extra memory gets freed when the macro - context is freed, i.e shortly after the macro got expanded. */ - - /* Is the -ftrack-macro-expansion flag in effect? */ - track_macro_exp = CPP_OPTION (pfile, track_macro_expansion); - - /* Now allocate memory space for tokens and locations resulting from - the macro expansion, copy the tokens and replace the arguments. - This memory must be freed when the context of the macro MACRO is - popped. */ - buff = tokens_buff_new (pfile, total, track_macro_exp ? &virt_locs : NULL); - - first = (const cpp_token **) buff->base; - - /* Create a macro map to record the locations of the tokens that are - involved in the expansion. Note that the expansion point is set - to the location of the closing parenthesis. Otherwise, the - subsequent map created for the first token that comes after the - macro map might have a wrong line number. That would lead to - tokens with wrong line numbers after the macro expansion. This - adds up to the memory overhead of the -ftrack-macro-expansion - flag; for every macro that is expanded, a "macro map" is - created. */ - if (track_macro_exp) - { - int num_macro_tokens = total; - if (track_macro_exp < 2) - /* Then the number of macro tokens won't take in account the - fact that function-like macro arguments can expand to - multiple tokens. This is to save memory at the expense of - accuracy. - - Suppose we have #define SQUARE(A) A * A - - And then we do SQUARE(2+3) - - Then the tokens 2, +, 3, will have the same location, - saying they come from the expansion of the argument A. */ - num_macro_tokens = exp_count; - map = linemap_enter_macro (pfile->line_table, node, - expansion_point_loc, - num_macro_tokens); - } - i = 0; - vaopt_state vaopt_tracker (pfile, macro->variadic, &args[macro->paramc - 1]); - const cpp_token **vaopt_start = NULL; - for (src = macro->exp.tokens; src < limit; src++) - { - unsigned int arg_tokens_count; - macro_arg_token_iter from; - const cpp_token **paste_flag = NULL; - const cpp_token **tmp_token_ptr; - - /* __VA_OPT__ handling. */ - vaopt_state::update_type vostate = vaopt_tracker.update (src); - if (__builtin_expect (vostate != vaopt_state::INCLUDE, false)) - { - if (vostate == vaopt_state::BEGIN) - { - /* Padding on the left of __VA_OPT__ (unless RHS of ##). */ - if (src != macro->exp.tokens && !(src[-1].flags & PASTE_LEFT)) - { - const cpp_token *t = padding_token (pfile, src); - unsigned index = expanded_token_index (pfile, macro, src, i); - /* Allocate a virtual location for the padding token and - append the token and its location to BUFF and - VIRT_LOCS. */ - tokens_buff_add_token (buff, virt_locs, t, - t->src_loc, t->src_loc, - map, index); - } - vaopt_start = tokens_buff_last_token_ptr (buff); - } - else if (vostate == vaopt_state::END) - { - const cpp_token **start = vaopt_start; - vaopt_start = NULL; - - paste_flag = tokens_buff_last_token_ptr (buff); - - if (vaopt_tracker.stringify ()) - { - unsigned int count - = start ? paste_flag - start : tokens_buff_count (buff); - const cpp_token **first - = start ? start + 1 - : (const cpp_token **) (buff->base); - unsigned int i, j; - - /* Paste any tokens that need to be pasted before calling - stringify_arg, because stringify_arg uses pfile->u_buff - which paste_tokens can use as well. */ - for (i = 0, j = 0; i < count; i++, j++) - { - const cpp_token *token = first[i]; - - if (token->flags & PASTE_LEFT) - { - location_t virt_loc = pfile->invocation_location; - const cpp_token *rhs; - do - { - if (i == count) - abort (); - rhs = first[++i]; - if (!paste_tokens (pfile, virt_loc, &token, rhs)) - { - --i; - break; - } - } - while (rhs->flags & PASTE_LEFT); - } - - first[j] = token; - } - if (j != i) - { - while (i-- != j) - tokens_buff_remove_last_token (buff); - count = j; - } - - const cpp_token *t = stringify_arg (pfile, first, count); - while (count--) - tokens_buff_remove_last_token (buff); - if (src->flags & PASTE_LEFT) - copy_paste_flag (pfile, &t, src); - tokens_buff_add_token (buff, virt_locs, - t, t->src_loc, t->src_loc, - NULL, 0); - continue; - } - if (start && paste_flag == start && (*start)->flags & PASTE_LEFT) - /* If __VA_OPT__ expands to nothing (either because __VA_ARGS__ - is empty or because it is __VA_OPT__() ), drop PASTE_LEFT - flag from previous token. */ - copy_paste_flag (pfile, start, &pfile->avoid_paste); - if (src->flags & PASTE_LEFT) - { - /* Don't avoid paste after all. */ - while (paste_flag && paste_flag != start - && *paste_flag == &pfile->avoid_paste) - { - tokens_buff_remove_last_token (buff); - paste_flag = tokens_buff_last_token_ptr (buff); - } - - /* With a non-empty __VA_OPT__ on the LHS of ##, the last - token should be flagged PASTE_LEFT. */ - if (paste_flag && (*paste_flag)->type != CPP_PADDING) - copy_paste_flag (pfile, paste_flag, src); - } - else - { - /* Otherwise, avoid paste on RHS, __VA_OPT__(c)d or - __VA_OPT__(c)__VA_OPT__(d). */ - const cpp_token *t = &pfile->avoid_paste; - tokens_buff_add_token (buff, virt_locs, - t, t->src_loc, t->src_loc, - NULL, 0); - } - } - continue; - } - - if (src->type != CPP_MACRO_ARG) - { - /* Allocate a virtual location for token SRC, and add that - token and its virtual location into the buffers BUFF and - VIRT_LOCS. */ - unsigned index = expanded_token_index (pfile, macro, src, i); - tokens_buff_add_token (buff, virt_locs, src, - src->src_loc, src->src_loc, - map, index); - i += 1; - continue; - } - - paste_flag = 0; - arg = &args[src->val.macro_arg.arg_no - 1]; - /* SRC is a macro parameter that we need to replace with its - corresponding argument. So at some point we'll need to - iterate over the tokens of the macro argument and copy them - into the "place" now holding the correspondig macro - parameter. We are going to use the iterator type - macro_argo_token_iter to handle that iterating. The 'if' - below is to initialize the iterator depending on the type of - tokens the macro argument has. It also does some adjustment - related to padding tokens and some pasting corner cases. */ - if (src->flags & STRINGIFY_ARG) - { - arg_tokens_count = 1; - macro_arg_token_iter_init (&from, - CPP_OPTION (pfile, - track_macro_expansion), - MACRO_ARG_TOKEN_STRINGIFIED, - arg, &arg->stringified); - } - else if (src->flags & PASTE_LEFT) - { - arg_tokens_count = arg->count; - macro_arg_token_iter_init (&from, - CPP_OPTION (pfile, - track_macro_expansion), - MACRO_ARG_TOKEN_NORMAL, - arg, arg->first); - } - else if (src != macro->exp.tokens && (src[-1].flags & PASTE_LEFT)) - { - int num_toks; - arg_tokens_count = arg->count; - macro_arg_token_iter_init (&from, - CPP_OPTION (pfile, - track_macro_expansion), - MACRO_ARG_TOKEN_NORMAL, - arg, arg->first); - - num_toks = tokens_buff_count (buff); - - if (num_toks != 0) - { - /* So the current parameter token is pasted to the previous - token in the replacement list. Let's look at what - we have as previous and current arguments. */ - - /* This is the previous argument's token ... */ - tmp_token_ptr = tokens_buff_last_token_ptr (buff); - - if ((*tmp_token_ptr)->type == CPP_COMMA - && macro->variadic - && src->val.macro_arg.arg_no == macro->paramc) - { - /* ... which is a comma; and the current parameter - is the last parameter of a variadic function-like - macro. If the argument to the current last - parameter is NULL, then swallow the comma, - otherwise drop the paste flag. */ - if (macro_arg_token_iter_get_token (&from) == NULL) - tokens_buff_remove_last_token (buff); - else - paste_flag = tmp_token_ptr; - } - /* Remove the paste flag if the RHS is a placemarker. */ - else if (arg_tokens_count == 0) - paste_flag = tmp_token_ptr; - } - } - else - { - arg_tokens_count = arg->expanded_count; - macro_arg_token_iter_init (&from, - CPP_OPTION (pfile, - track_macro_expansion), - MACRO_ARG_TOKEN_EXPANDED, - arg, arg->expanded); - - if (last_token_is (buff, vaopt_start)) - { - /* We're expanding an arg at the beginning of __VA_OPT__. - Skip padding. */ - while (arg_tokens_count) - { - const cpp_token *t = macro_arg_token_iter_get_token (&from); - if (t->type != CPP_PADDING) - break; - macro_arg_token_iter_forward (&from); - --arg_tokens_count; - } - } - } - - /* Padding on the left of an argument (unless RHS of ##). */ - if ((!pfile->state.in_directive || pfile->state.directive_wants_padding) - && src != macro->exp.tokens - && !(src[-1].flags & PASTE_LEFT) - && !last_token_is (buff, vaopt_start)) - { - const cpp_token *t = padding_token (pfile, src); - unsigned index = expanded_token_index (pfile, macro, src, i); - /* Allocate a virtual location for the padding token and - append the token and its location to BUFF and - VIRT_LOCS. */ - tokens_buff_add_token (buff, virt_locs, t, - t->src_loc, t->src_loc, - map, index); - } - - if (arg_tokens_count) - { - /* So now we've got the number of tokens that make up the - argument that is going to replace the current parameter - in the macro's replacement list. */ - unsigned int j; - for (j = 0; j < arg_tokens_count; ++j) - { - /* So if track_macro_exp is < 2, the user wants to - save extra memory while tracking macro expansion - locations. So in that case here is what we do: - - Suppose we have #define SQUARE(A) A * A - - And then we do SQUARE(2+3) - - Then the tokens 2, +, 3, will have the same location, - saying they come from the expansion of the argument - A. - - So that means we are going to ignore the COUNT tokens - resulting from the expansion of the current macro - argument. In other words all the ARG_TOKENS_COUNT tokens - resulting from the expansion of the macro argument will - have the index I. Normally, each of those tokens should - have index I+J. */ - unsigned token_index = i; - unsigned index; - if (track_macro_exp > 1) - token_index += j; - - index = expanded_token_index (pfile, macro, src, token_index); - const cpp_token *tok = macro_arg_token_iter_get_token (&from); - tokens_buff_add_token (buff, virt_locs, tok, - macro_arg_token_iter_get_location (&from), - src->src_loc, map, index); - macro_arg_token_iter_forward (&from); - } - - /* With a non-empty argument on the LHS of ##, the last - token should be flagged PASTE_LEFT. */ - if (src->flags & PASTE_LEFT) - paste_flag - = (const cpp_token **) tokens_buff_last_token_ptr (buff); - } - else if (CPP_PEDANTIC (pfile) && ! CPP_OPTION (pfile, c99) - && ! macro->syshdr && ! _cpp_in_system_header (pfile)) - { - if (CPP_OPTION (pfile, cplusplus)) - cpp_pedwarning (pfile, CPP_W_PEDANTIC, - "invoking macro %s argument %d: " - "empty macro arguments are undefined" - " in ISO C++98", - NODE_NAME (node), src->val.macro_arg.arg_no); - else if (CPP_OPTION (pfile, cpp_warn_c90_c99_compat)) - cpp_pedwarning (pfile, - CPP_OPTION (pfile, cpp_warn_c90_c99_compat) > 0 - ? CPP_W_C90_C99_COMPAT : CPP_W_PEDANTIC, - "invoking macro %s argument %d: " - "empty macro arguments are undefined" - " in ISO C90", - NODE_NAME (node), src->val.macro_arg.arg_no); - } - else if (CPP_OPTION (pfile, cpp_warn_c90_c99_compat) > 0 - && ! CPP_OPTION (pfile, cplusplus) - && ! macro->syshdr && ! _cpp_in_system_header (pfile)) - cpp_warning (pfile, CPP_W_C90_C99_COMPAT, - "invoking macro %s argument %d: " - "empty macro arguments are undefined" - " in ISO C90", - NODE_NAME (node), src->val.macro_arg.arg_no); - - /* Avoid paste on RHS (even case count == 0). */ - if (!pfile->state.in_directive && !(src->flags & PASTE_LEFT)) - { - const cpp_token *t = &pfile->avoid_paste; - tokens_buff_add_token (buff, virt_locs, - t, t->src_loc, t->src_loc, - NULL, 0); - } - - /* Add a new paste flag, or remove an unwanted one. */ - if (paste_flag) - copy_paste_flag (pfile, paste_flag, src); - - i += arg_tokens_count; - } - - if (track_macro_exp) - push_extended_tokens_context (pfile, node, buff, virt_locs, first, - tokens_buff_count (buff)); - else - push_ptoken_context (pfile, node, buff, first, - tokens_buff_count (buff)); - - num_macro_tokens_counter += tokens_buff_count (buff); -} - -/* Return a special padding token, with padding inherited from SOURCE. */ -static const cpp_token * -padding_token (cpp_reader *pfile, const cpp_token *source) -{ - cpp_token *result = _cpp_temp_token (pfile); - - result->type = CPP_PADDING; - - /* Data in GCed data structures cannot be made const so far, so we - need a cast here. */ - result->val.source = (cpp_token *) source; - result->flags = 0; - return result; -} - -/* Get a new uninitialized context. Create a new one if we cannot - re-use an old one. */ -static cpp_context * -next_context (cpp_reader *pfile) -{ - cpp_context *result = pfile->context->next; - - if (result == 0) - { - result = XNEW (cpp_context); - memset (result, 0, sizeof (cpp_context)); - result->prev = pfile->context; - result->next = 0; - pfile->context->next = result; - } - - pfile->context = result; - return result; -} - -/* Push a list of pointers to tokens. */ -static void -push_ptoken_context (cpp_reader *pfile, cpp_hashnode *macro, _cpp_buff *buff, - const cpp_token **first, unsigned int count) -{ - cpp_context *context = next_context (pfile); - - context->tokens_kind = TOKENS_KIND_INDIRECT; - context->c.macro = macro; - context->buff = buff; - FIRST (context).ptoken = first; - LAST (context).ptoken = first + count; -} - -/* Push a list of tokens. - - A NULL macro means that we should continue the current macro - expansion, in essence. That means that if we are currently in a - macro expansion context, we'll make the new pfile->context refer to - the current macro. */ -void -_cpp_push_token_context (cpp_reader *pfile, cpp_hashnode *macro, - const cpp_token *first, unsigned int count) -{ - cpp_context *context; - - if (macro == NULL) - macro = macro_of_context (pfile->context); - - context = next_context (pfile); - context->tokens_kind = TOKENS_KIND_DIRECT; - context->c.macro = macro; - context->buff = NULL; - FIRST (context).token = first; - LAST (context).token = first + count; -} - -/* Build a context containing a list of tokens as well as their - virtual locations and push it. TOKENS_BUFF is the buffer that - contains the tokens pointed to by FIRST. If TOKENS_BUFF is - non-NULL, it means that the context owns it, meaning that - _cpp_pop_context will free it as well as VIRT_LOCS_BUFF that - contains the virtual locations. - - A NULL macro means that we should continue the current macro - expansion, in essence. That means that if we are currently in a - macro expansion context, we'll make the new pfile->context refer to - the current macro. */ -static void -push_extended_tokens_context (cpp_reader *pfile, - cpp_hashnode *macro, - _cpp_buff *token_buff, - location_t *virt_locs, - const cpp_token **first, - unsigned int count) -{ - cpp_context *context; - macro_context *m; - - if (macro == NULL) - macro = macro_of_context (pfile->context); - - context = next_context (pfile); - context->tokens_kind = TOKENS_KIND_EXTENDED; - context->buff = token_buff; - - m = XNEW (macro_context); - m->macro_node = macro; - m->virt_locs = virt_locs; - m->cur_virt_loc = virt_locs; - context->c.mc = m; - FIRST (context).ptoken = first; - LAST (context).ptoken = first + count; -} - -/* Push a traditional macro's replacement text. */ -void -_cpp_push_text_context (cpp_reader *pfile, cpp_hashnode *macro, - const uchar *start, size_t len) -{ - cpp_context *context = next_context (pfile); - - context->tokens_kind = TOKENS_KIND_DIRECT; - context->c.macro = macro; - context->buff = NULL; - CUR (context) = start; - RLIMIT (context) = start + len; - macro->flags |= NODE_DISABLED; -} - -/* Creates a buffer that holds tokens a.k.a "token buffer", usually - for the purpose of storing them on a cpp_context. If VIRT_LOCS is - non-null (which means that -ftrack-macro-expansion is on), - *VIRT_LOCS is set to a newly allocated buffer that is supposed to - hold the virtual locations of the tokens resulting from macro - expansion. */ -static _cpp_buff* -tokens_buff_new (cpp_reader *pfile, size_t len, - location_t **virt_locs) -{ - size_t tokens_size = len * sizeof (cpp_token *); - size_t locs_size = len * sizeof (location_t); - - if (virt_locs != NULL) - *virt_locs = XNEWVEC (location_t, locs_size); - return _cpp_get_buff (pfile, tokens_size); -} - -/* Returns the number of tokens contained in a token buffer. The - buffer holds a set of cpp_token*. */ -static size_t -tokens_buff_count (_cpp_buff *buff) -{ - return (BUFF_FRONT (buff) - buff->base) / sizeof (cpp_token *); -} - -/* Return a pointer to the last token contained in the token buffer - BUFF. */ -static const cpp_token ** -tokens_buff_last_token_ptr (_cpp_buff *buff) -{ - if (BUFF_FRONT (buff) == buff->base) - return NULL; - return &((const cpp_token **) BUFF_FRONT (buff))[-1]; -} - -/* Remove the last token contained in the token buffer TOKENS_BUFF. - If VIRT_LOCS_BUFF is non-NULL, it should point at the buffer - containing the virtual locations of the tokens in TOKENS_BUFF; in - which case the function updates that buffer as well. */ -static inline void -tokens_buff_remove_last_token (_cpp_buff *tokens_buff) - -{ - if (BUFF_FRONT (tokens_buff) > tokens_buff->base) - BUFF_FRONT (tokens_buff) = - (unsigned char *) &((cpp_token **) BUFF_FRONT (tokens_buff))[-1]; -} - -/* Insert a token into the token buffer at the position pointed to by - DEST. Note that the buffer is not enlarged so the previous token - that was at *DEST is overwritten. VIRT_LOC_DEST, if non-null, - means -ftrack-macro-expansion is effect; it then points to where to - insert the virtual location of TOKEN. TOKEN is the token to - insert. VIRT_LOC is the virtual location of the token, i.e, the - location possibly encoding its locus across macro expansion. If - TOKEN is an argument of a function-like macro (inside a macro - replacement list), PARM_DEF_LOC is the spelling location of the - macro parameter that TOKEN is replacing, in the replacement list of - the macro. If TOKEN is not an argument of a function-like macro or - if it doesn't come from a macro expansion, then VIRT_LOC can just - be set to the same value as PARM_DEF_LOC. If MAP is non null, it - means TOKEN comes from a macro expansion and MAP is the macro map - associated to the macro. MACRO_TOKEN_INDEX points to the index of - the token in the macro map; it is not considered if MAP is NULL. - - Upon successful completion this function returns the a pointer to - the position of the token coming right after the insertion - point. */ -static inline const cpp_token ** -tokens_buff_put_token_to (const cpp_token **dest, - location_t *virt_loc_dest, - const cpp_token *token, - location_t virt_loc, - location_t parm_def_loc, - const line_map_macro *map, - unsigned int macro_token_index) -{ - location_t macro_loc = virt_loc; - const cpp_token **result; - - if (virt_loc_dest) - { - /* -ftrack-macro-expansion is on. */ - if (map) - macro_loc = linemap_add_macro_token (map, macro_token_index, - virt_loc, parm_def_loc); - *virt_loc_dest = macro_loc; - } - *dest = token; - result = &dest[1]; - - return result; -} - -/* Adds a token at the end of the tokens contained in BUFFER. Note - that this function doesn't enlarge BUFFER when the number of tokens - reaches BUFFER's size; it aborts in that situation. - - TOKEN is the token to append. VIRT_LOC is the virtual location of - the token, i.e, the location possibly encoding its locus across - macro expansion. If TOKEN is an argument of a function-like macro - (inside a macro replacement list), PARM_DEF_LOC is the location of - the macro parameter that TOKEN is replacing. If TOKEN doesn't come - from a macro expansion, then VIRT_LOC can just be set to the same - value as PARM_DEF_LOC. If MAP is non null, it means TOKEN comes - from a macro expansion and MAP is the macro map associated to the - macro. MACRO_TOKEN_INDEX points to the index of the token in the - macro map; It is not considered if MAP is NULL. If VIRT_LOCS is - non-null, it means -ftrack-macro-expansion is on; in which case - this function adds the virtual location DEF_LOC to the VIRT_LOCS - array, at the same index as the one of TOKEN in BUFFER. Upon - successful completion this function returns the a pointer to the - position of the token coming right after the insertion point. */ -static const cpp_token ** -tokens_buff_add_token (_cpp_buff *buffer, - location_t *virt_locs, - const cpp_token *token, - location_t virt_loc, - location_t parm_def_loc, - const line_map_macro *map, - unsigned int macro_token_index) -{ - const cpp_token **result; - location_t *virt_loc_dest = NULL; - unsigned token_index = - (BUFF_FRONT (buffer) - buffer->base) / sizeof (cpp_token *); - - /* Abort if we pass the end the buffer. */ - if (BUFF_FRONT (buffer) > BUFF_LIMIT (buffer)) - abort (); - - if (virt_locs != NULL) - virt_loc_dest = &virt_locs[token_index]; - - result = - tokens_buff_put_token_to ((const cpp_token **) BUFF_FRONT (buffer), - virt_loc_dest, token, virt_loc, parm_def_loc, - map, macro_token_index); - - BUFF_FRONT (buffer) = (unsigned char *) result; - return result; -} - -/* Allocate space for the function-like macro argument ARG to store - the tokens resulting from the macro-expansion of the tokens that - make up ARG itself. That space is allocated in ARG->expanded and - needs to be freed using free. */ -static void -alloc_expanded_arg_mem (cpp_reader *pfile, macro_arg *arg, size_t capacity) -{ - gcc_checking_assert (arg->expanded == NULL - && arg->expanded_virt_locs == NULL); - - arg->expanded = XNEWVEC (const cpp_token *, capacity); - if (CPP_OPTION (pfile, track_macro_expansion)) - arg->expanded_virt_locs = XNEWVEC (location_t, capacity); - -} - -/* If necessary, enlarge ARG->expanded to so that it can contain SIZE - tokens. */ -static void -ensure_expanded_arg_room (cpp_reader *pfile, macro_arg *arg, - size_t size, size_t *expanded_capacity) -{ - if (size <= *expanded_capacity) - return; - - size *= 2; - - arg->expanded = - XRESIZEVEC (const cpp_token *, arg->expanded, size); - *expanded_capacity = size; - - if (CPP_OPTION (pfile, track_macro_expansion)) - { - if (arg->expanded_virt_locs == NULL) - arg->expanded_virt_locs = XNEWVEC (location_t, size); - else - arg->expanded_virt_locs = XRESIZEVEC (location_t, - arg->expanded_virt_locs, - size); - } -} - -/* - Expand an argument ARG before replacing parameters in a - function-like macro. This works by pushing a context with the - argument's tokens, and then expanding that into a temporary buffer - as if it were a normal part of the token stream. collect_args() - has terminated the argument's tokens with a CPP_EOF so that we know - when we have fully expanded the argument. - */ -static void -expand_arg (cpp_reader *pfile, macro_arg *arg) -{ - size_t capacity; - bool saved_warn_trad; - bool track_macro_exp_p = CPP_OPTION (pfile, track_macro_expansion); - bool saved_ignore__Pragma; - - if (arg->count == 0 - || arg->expanded != NULL) - return; - - /* Don't warn about funlike macros when pre-expanding. */ - saved_warn_trad = CPP_WTRADITIONAL (pfile); - CPP_WTRADITIONAL (pfile) = 0; - - /* Loop, reading in the tokens of the argument. */ - capacity = 256; - alloc_expanded_arg_mem (pfile, arg, capacity); - - if (track_macro_exp_p) - push_extended_tokens_context (pfile, NULL, NULL, - arg->virt_locs, - arg->first, - arg->count + 1); - else - push_ptoken_context (pfile, NULL, NULL, - arg->first, arg->count + 1); - - saved_ignore__Pragma = pfile->state.ignore__Pragma; - pfile->state.ignore__Pragma = 1; - - for (;;) - { - const cpp_token *token; - location_t location; - - ensure_expanded_arg_room (pfile, arg, arg->expanded_count + 1, - &capacity); - - token = cpp_get_token_1 (pfile, &location); - - if (token->type == CPP_EOF) - break; - - set_arg_token (arg, token, location, - arg->expanded_count, MACRO_ARG_TOKEN_EXPANDED, - CPP_OPTION (pfile, track_macro_expansion)); - arg->expanded_count++; - } - - _cpp_pop_context (pfile); - - CPP_WTRADITIONAL (pfile) = saved_warn_trad; - pfile->state.ignore__Pragma = saved_ignore__Pragma; -} - -/* Returns the macro associated to the current context if we are in - the context a macro expansion, NULL otherwise. */ -static cpp_hashnode* -macro_of_context (cpp_context *context) -{ - if (context == NULL) - return NULL; - - return (context->tokens_kind == TOKENS_KIND_EXTENDED) - ? context->c.mc->macro_node - : context->c.macro; -} - -/* Return TRUE iff we are expanding a macro or are about to start - expanding one. If we are effectively expanding a macro, the - function macro_of_context returns a pointer to the macro being - expanded. */ -static bool -in_macro_expansion_p (cpp_reader *pfile) -{ - if (pfile == NULL) - return false; - - return (pfile->about_to_expand_macro_p - || macro_of_context (pfile->context)); -} - -/* Pop the current context off the stack, re-enabling the macro if the - context represented a macro's replacement list. Initially the - context structure was not freed so that we can re-use it later, but - now we do free it to reduce peak memory consumption. */ -void -_cpp_pop_context (cpp_reader *pfile) -{ - cpp_context *context = pfile->context; - - /* We should not be popping the base context. */ - gcc_assert (context != &pfile->base_context); - - if (context->c.macro) - { - cpp_hashnode *macro; - if (context->tokens_kind == TOKENS_KIND_EXTENDED) - { - macro_context *mc = context->c.mc; - macro = mc->macro_node; - /* If context->buff is set, it means the life time of tokens - is bound to the life time of this context; so we must - free the tokens; that means we must free the virtual - locations of these tokens too. */ - if (context->buff && mc->virt_locs) - { - free (mc->virt_locs); - mc->virt_locs = NULL; - } - free (mc); - context->c.mc = NULL; - } - else - macro = context->c.macro; - - /* Beware that MACRO can be NULL in cases like when we are - called from expand_arg. In those cases, a dummy context with - tokens is pushed just for the purpose of walking them using - cpp_get_token_1. In that case, no 'macro' field is set into - the dummy context. */ - if (macro != NULL - /* Several contiguous macro expansion contexts can be - associated to the same macro; that means it's the same - macro expansion that spans across all these (sub) - contexts. So we should re-enable an expansion-disabled - macro only when we are sure we are really out of that - macro expansion. */ - && macro_of_context (context->prev) != macro) - macro->flags &= ~NODE_DISABLED; - - if (macro == pfile->top_most_macro_node && context->prev == NULL) - /* We are popping the context of the top-most macro node. */ - pfile->top_most_macro_node = NULL; - } - - if (context->buff) - { - /* Decrease memory peak consumption by freeing the memory used - by the context. */ - _cpp_free_buff (context->buff); - } - - pfile->context = context->prev; - /* decrease peak memory consumption by feeing the context. */ - pfile->context->next = NULL; - free (context); -} - -/* Return TRUE if we reached the end of the set of tokens stored in - CONTEXT, FALSE otherwise. */ -static inline bool -reached_end_of_context (cpp_context *context) -{ - if (context->tokens_kind == TOKENS_KIND_DIRECT) - return FIRST (context).token == LAST (context).token; - else if (context->tokens_kind == TOKENS_KIND_INDIRECT - || context->tokens_kind == TOKENS_KIND_EXTENDED) - return FIRST (context).ptoken == LAST (context).ptoken; - else - abort (); -} - -/* Consume the next token contained in the current context of PFILE, - and return it in *TOKEN. It's "full location" is returned in - *LOCATION. If -ftrack-macro-location is in effeect, fFull location" - means the location encoding the locus of the token across macro - expansion; otherwise it's just is the "normal" location of the - token which (*TOKEN)->src_loc. */ -static inline void -consume_next_token_from_context (cpp_reader *pfile, - const cpp_token ** token, - location_t *location) -{ - cpp_context *c = pfile->context; - - if ((c)->tokens_kind == TOKENS_KIND_DIRECT) - { - *token = FIRST (c).token; - *location = (*token)->src_loc; - FIRST (c).token++; - } - else if ((c)->tokens_kind == TOKENS_KIND_INDIRECT) - { - *token = *FIRST (c).ptoken; - *location = (*token)->src_loc; - FIRST (c).ptoken++; - } - else if ((c)->tokens_kind == TOKENS_KIND_EXTENDED) - { - macro_context *m = c->c.mc; - *token = *FIRST (c).ptoken; - if (m->virt_locs) - { - *location = *m->cur_virt_loc; - m->cur_virt_loc++; - } - else - *location = (*token)->src_loc; - FIRST (c).ptoken++; - } - else - abort (); -} - -/* In the traditional mode of the preprocessor, if we are currently in - a directive, the location of a token must be the location of the - start of the directive line. This function returns the proper - location if we are in the traditional mode, and just returns - LOCATION otherwise. */ - -static inline location_t -maybe_adjust_loc_for_trad_cpp (cpp_reader *pfile, location_t location) -{ - if (CPP_OPTION (pfile, traditional)) - { - if (pfile->state.in_directive) - return pfile->directive_line; - } - return location; -} - -/* Routine to get a token as well as its location. - - Macro expansions and directives are transparently handled, - including entering included files. Thus tokens are post-macro - expansion, and after any intervening directives. External callers - see CPP_EOF only at EOF. Internal callers also see it when meeting - a directive inside a macro call, when at the end of a directive and - state.in_directive is still 1, and at the end of argument - pre-expansion. - - LOC is an out parameter; *LOC is set to the location "as expected - by the user". Please read the comment of - cpp_get_token_with_location to learn more about the meaning of this - location. */ -static const cpp_token* -cpp_get_token_1 (cpp_reader *pfile, location_t *location) -{ - const cpp_token *result; - /* This token is a virtual token that either encodes a location - related to macro expansion or a spelling location. */ - location_t virt_loc = 0; - /* pfile->about_to_expand_macro_p can be overriden by indirect calls - to functions that push macro contexts. So let's save it so that - we can restore it when we are about to leave this routine. */ - bool saved_about_to_expand_macro = pfile->about_to_expand_macro_p; - - for (;;) - { - cpp_hashnode *node; - cpp_context *context = pfile->context; - - /* Context->prev == 0 <=> base context. */ - if (!context->prev) - { - result = _cpp_lex_token (pfile); - virt_loc = result->src_loc; - } - else if (!reached_end_of_context (context)) - { - consume_next_token_from_context (pfile, &result, - &virt_loc); - if (result->flags & PASTE_LEFT) - { - paste_all_tokens (pfile, result); - if (pfile->state.in_directive) - continue; - result = padding_token (pfile, result); - goto out; - } - } - else - { - if (pfile->context->c.macro) - ++num_expanded_macros_counter; - _cpp_pop_context (pfile); - if (pfile->state.in_directive) - continue; - result = &pfile->avoid_paste; - goto out; - } - - if (pfile->state.in_directive && result->type == CPP_COMMENT) - continue; - - if (result->type != CPP_NAME) - break; - - node = result->val.node.node; - - if (node->type == NT_VOID || (result->flags & NO_EXPAND)) - break; - - if (!(node->flags & NODE_USED) - && node->type == NT_USER_MACRO - && !node->value.macro - && !cpp_get_deferred_macro (pfile, node, result->src_loc)) - break; - - if (!(node->flags & NODE_DISABLED)) - { - int ret = 0; - /* If not in a macro context, and we're going to start an - expansion, record the location and the top level macro - about to be expanded. */ - if (!in_macro_expansion_p (pfile)) - { - pfile->invocation_location = result->src_loc; - pfile->top_most_macro_node = node; - } - if (pfile->state.prevent_expansion) - break; - - /* Conditional macros require that a predicate be evaluated - first. */ - if ((node->flags & NODE_CONDITIONAL) != 0) - { - if (pfile->cb.macro_to_expand) - { - bool whitespace_after; - const cpp_token *peek_tok = cpp_peek_token (pfile, 0); - - whitespace_after = (peek_tok->type == CPP_PADDING - || (peek_tok->flags & PREV_WHITE)); - node = pfile->cb.macro_to_expand (pfile, result); - if (node) - ret = enter_macro_context (pfile, node, result, virt_loc); - else if (whitespace_after) - { - /* If macro_to_expand hook returned NULL and it - ate some tokens, see if we don't need to add - a padding token in between this and the - next token. */ - peek_tok = cpp_peek_token (pfile, 0); - if (peek_tok->type != CPP_PADDING - && (peek_tok->flags & PREV_WHITE) == 0) - _cpp_push_token_context (pfile, NULL, - padding_token (pfile, - peek_tok), 1); - } - } - } - else - ret = enter_macro_context (pfile, node, result, virt_loc); - if (ret) - { - if (pfile->state.in_directive || ret == 2) - continue; - result = padding_token (pfile, result); - goto out; - } - } - else - { - /* Flag this token as always unexpandable. FIXME: move this - to collect_args()?. */ - cpp_token *t = _cpp_temp_token (pfile); - t->type = result->type; - t->flags = result->flags | NO_EXPAND; - t->val = result->val; - result = t; - } - - break; - } - - out: - if (location != NULL) - { - if (virt_loc == 0) - virt_loc = result->src_loc; - *location = virt_loc; - - if (!CPP_OPTION (pfile, track_macro_expansion) - && macro_of_context (pfile->context) != NULL) - /* We are in a macro expansion context, are not tracking - virtual location, but were asked to report the location - of the expansion point of the macro being expanded. */ - *location = pfile->invocation_location; - - *location = maybe_adjust_loc_for_trad_cpp (pfile, *location); - } - - pfile->about_to_expand_macro_p = saved_about_to_expand_macro; - - if (pfile->state.directive_file_token - && !pfile->state.parsing_args - && !(result->type == CPP_PADDING || result->type == CPP_COMMENT) - && !(15 & --pfile->state.directive_file_token)) - { - /* Do header-name frobbery. Concatenate < ... > as approprate. - Do header search if needed, and finally drop the outer <> or - "". */ - pfile->state.angled_headers = false; - - /* Do angle-header reconstitution. Then do include searching. - We'll always end up with a ""-quoted header-name in that - case. If searching finds nothing, we emit a diagnostic and - an empty string. */ - size_t len = 0; - char *fname = NULL; - - cpp_token *tmp = _cpp_temp_token (pfile); - *tmp = *result; - - tmp->type = CPP_HEADER_NAME; - bool need_search = !pfile->state.directive_file_token; - pfile->state.directive_file_token = 0; - - bool angle = result->type != CPP_STRING; - if (result->type == CPP_HEADER_NAME - || (result->type == CPP_STRING && result->val.str.text[0] != 'R')) - { - len = result->val.str.len - 2; - fname = XNEWVEC (char, len + 1); - memcpy (fname, result->val.str.text + 1, len); - fname[len] = 0; - } - else if (result->type == CPP_LESS) - fname = _cpp_bracket_include (pfile); - - if (fname) - { - /* We have a header-name. Look it up. This will emit an - unfound diagnostic. Canonicalize the found name. */ - const char *found = fname; - - if (need_search) - { - found = _cpp_find_header_unit (pfile, fname, angle, tmp->src_loc); - if (!found) - found = ""; - len = strlen (found); - } - /* Force a leading './' if it's not absolute. */ - bool dotme = (found[0] == '.' ? !IS_DIR_SEPARATOR (found[1]) - : found[0] && !IS_ABSOLUTE_PATH (found)); - - if (BUFF_ROOM (pfile->u_buff) < len + 1 + dotme * 2) - _cpp_extend_buff (pfile, &pfile->u_buff, len + 1 + dotme * 2); - unsigned char *buf = BUFF_FRONT (pfile->u_buff); - size_t pos = 0; - - if (dotme) - { - buf[pos++] = '.'; - /* Apparently '/' is unconditional. */ - buf[pos++] = '/'; - } - memcpy (&buf[pos], found, len); - pos += len; - buf[pos] = 0; - - tmp->val.str.len = pos; - tmp->val.str.text = buf; - - tmp->type = CPP_HEADER_NAME; - XDELETEVEC (fname); - - result = tmp; - } - } - - return result; -} - -/* External routine to get a token. Also used nearly everywhere - internally, except for places where we know we can safely call - _cpp_lex_token directly, such as lexing a directive name. - - Macro expansions and directives are transparently handled, - including entering included files. Thus tokens are post-macro - expansion, and after any intervening directives. External callers - see CPP_EOF only at EOF. Internal callers also see it when meeting - a directive inside a macro call, when at the end of a directive and - state.in_directive is still 1, and at the end of argument - pre-expansion. */ -const cpp_token * -cpp_get_token (cpp_reader *pfile) -{ - return cpp_get_token_1 (pfile, NULL); -} - -/* Like cpp_get_token, but also returns a virtual token location - separate from the spelling location carried by the returned token. - - LOC is an out parameter; *LOC is set to the location "as expected - by the user". This matters when a token results from macro - expansion; in that case the token's spelling location indicates the - locus of the token in the definition of the macro but *LOC - virtually encodes all the other meaningful locuses associated to - the token. - - What? virtual location? Yes, virtual location. - - If the token results from macro expansion and if macro expansion - location tracking is enabled its virtual location encodes (at the - same time): - - - the spelling location of the token - - - the locus of the macro expansion point - - - the locus of the point where the token got instantiated as part - of the macro expansion process. - - You have to use the linemap API to get the locus you are interested - in from a given virtual location. - - Note however that virtual locations are not necessarily ordered for - relations '<' and '>'. One must use the function - linemap_location_before_p instead of using the relational operator - '<'. - - If macro expansion tracking is off and if the token results from - macro expansion the virtual location is the expansion point of the - macro that got expanded. - - When the token doesn't result from macro expansion, the virtual - location is just the same thing as its spelling location. */ - -const cpp_token * -cpp_get_token_with_location (cpp_reader *pfile, location_t *loc) -{ - return cpp_get_token_1 (pfile, loc); -} - -/* Returns true if we're expanding an object-like macro that was - defined in a system header. Just checks the macro at the top of - the stack. Used for diagnostic suppression. - Also return true for builtin macros. */ -int -cpp_sys_macro_p (cpp_reader *pfile) -{ - cpp_hashnode *node = NULL; - - if (pfile->context->tokens_kind == TOKENS_KIND_EXTENDED) - node = pfile->context->c.mc->macro_node; - else - node = pfile->context->c.macro; - - if (!node) - return false; - if (cpp_builtin_macro_p (node)) - return true; - return node->value.macro && node->value.macro->syshdr; -} - -/* Read each token in, until end of the current file. Directives are - transparently processed. */ -void -cpp_scan_nooutput (cpp_reader *pfile) -{ - /* Request a CPP_EOF token at the end of this file, rather than - transparently continuing with the including file. */ - pfile->buffer->return_at_eof = true; - - pfile->state.discarding_output++; - pfile->state.prevent_expansion++; - - if (CPP_OPTION (pfile, traditional)) - while (_cpp_read_logical_line_trad (pfile)) - ; - else - while (cpp_get_token (pfile)->type != CPP_EOF) - ; - - pfile->state.discarding_output--; - pfile->state.prevent_expansion--; -} - -/* Step back one or more tokens obtained from the lexer. */ -void -_cpp_backup_tokens_direct (cpp_reader *pfile, unsigned int count) -{ - pfile->lookaheads += count; - while (count--) - { - pfile->cur_token--; - if (pfile->cur_token == pfile->cur_run->base - /* Possible with -fpreprocessed and no leading #line. */ - && pfile->cur_run->prev != NULL) - { - pfile->cur_run = pfile->cur_run->prev; - pfile->cur_token = pfile->cur_run->limit; - } - } -} - -/* Step back one (or more) tokens. Can only step back more than 1 if - they are from the lexer, and not from macro expansion. */ -void -_cpp_backup_tokens (cpp_reader *pfile, unsigned int count) -{ - if (pfile->context->prev == NULL) - _cpp_backup_tokens_direct (pfile, count); - else - { - if (count != 1) - abort (); - if (pfile->context->tokens_kind == TOKENS_KIND_DIRECT) - FIRST (pfile->context).token--; - else if (pfile->context->tokens_kind == TOKENS_KIND_INDIRECT) - FIRST (pfile->context).ptoken--; - else if (pfile->context->tokens_kind == TOKENS_KIND_EXTENDED) - { - FIRST (pfile->context).ptoken--; - if (pfile->context->c.macro) - { - macro_context *m = pfile->context->c.mc; - m->cur_virt_loc--; - gcc_checking_assert (m->cur_virt_loc >= m->virt_locs); - } - else - abort (); - } - else - abort (); - } -} - -/* #define directive parsing and handling. */ - -/* Returns true if a macro redefinition warning is required. */ -static bool -warn_of_redefinition (cpp_reader *pfile, cpp_hashnode *node, - const cpp_macro *macro2) -{ - /* Some redefinitions need to be warned about regardless. */ - if (node->flags & NODE_WARN) - return true; - - /* Suppress warnings for builtins that lack the NODE_WARN flag, - unless Wbuiltin-macro-redefined. */ - if (cpp_builtin_macro_p (node)) - return CPP_OPTION (pfile, warn_builtin_macro_redefined); - - /* Redefinitions of conditional (context-sensitive) macros, on - the other hand, must be allowed silently. */ - if (node->flags & NODE_CONDITIONAL) - return false; - - if (cpp_macro *macro1 = get_deferred_or_lazy_macro (pfile, node, macro2->line)) - return cpp_compare_macros (macro1, macro2); - return false; -} - -/* Return TRUE if MACRO1 and MACRO2 differ. */ - -bool -cpp_compare_macros (const cpp_macro *macro1, const cpp_macro *macro2) -{ - /* Redefinition of a macro is allowed if and only if the old and new - definitions are the same. (6.10.3 paragraph 2). */ - - /* Don't check count here as it can be different in valid - traditional redefinitions with just whitespace differences. */ - if (macro1->paramc != macro2->paramc - || macro1->fun_like != macro2->fun_like - || macro1->variadic != macro2->variadic) - return true; - - /* Check parameter spellings. */ - for (unsigned i = macro1->paramc; i--; ) - if (macro1->parm.params[i] != macro2->parm.params[i]) - return true; - - /* Check the replacement text or tokens. */ - if (macro1->kind == cmk_traditional) - return _cpp_expansions_different_trad (macro1, macro2); - - if (macro1->count != macro2->count) - return true; - - for (unsigned i= macro1->count; i--; ) - if (!_cpp_equiv_tokens (¯o1->exp.tokens[i], ¯o2->exp.tokens[i])) - return true; - - return false; -} - -/* Free the definition of hashnode H. */ -void -_cpp_free_definition (cpp_hashnode *h) -{ - /* Macros and assertions no longer have anything to free. */ - h->type = NT_VOID; - h->value.answers = NULL; - h->flags &= ~(NODE_DISABLED | NODE_USED); -} - -/* Save parameter NODE (spelling SPELLING) to the parameter list of - macro MACRO. Returns true on success, false on failure. */ -bool -_cpp_save_parameter (cpp_reader *pfile, unsigned n, cpp_hashnode *node, - cpp_hashnode *spelling) -{ - /* Constraint 6.10.3.6 - duplicate parameter names. */ - if (node->type == NT_MACRO_ARG) - { - cpp_error (pfile, CPP_DL_ERROR, "duplicate macro parameter \"%s\"", - NODE_NAME (node)); - return false; - } - - unsigned len = (n + 1) * sizeof (struct macro_arg_saved_data); - if (len > pfile->macro_buffer_len) - { - pfile->macro_buffer - = XRESIZEVEC (unsigned char, pfile->macro_buffer, len); - pfile->macro_buffer_len = len; - } - - macro_arg_saved_data *saved = (macro_arg_saved_data *)pfile->macro_buffer; - saved[n].canonical_node = node; - saved[n].value = node->value; - saved[n].type = node->type; - - void *base = _cpp_reserve_room (pfile, n * sizeof (cpp_hashnode *), - sizeof (cpp_hashnode *)); - ((cpp_hashnode **)base)[n] = spelling; - - /* Morph into a macro arg. */ - node->type = NT_MACRO_ARG; - /* Index is 1 based. */ - node->value.arg_index = n + 1; - - return true; -} - -/* Restore the parameters to their previous state. */ -void -_cpp_unsave_parameters (cpp_reader *pfile, unsigned n) -{ - /* Clear the fast argument lookup indices. */ - while (n--) - { - struct macro_arg_saved_data *save = - &((struct macro_arg_saved_data *) pfile->macro_buffer)[n]; - - struct cpp_hashnode *node = save->canonical_node; - node->type = save->type; - node->value = save->value; - } -} - -/* Check the syntax of the parameters in a MACRO definition. Return - false on failure. Set *N_PTR and *VARADIC_PTR as appropriate. - '(' ')' - '(' parm-list ',' last-parm ')' - '(' last-parm ')' - parm-list: name - | parm-list, name - last-parm: name - | name '...' - | '...' -*/ - -static bool -parse_params (cpp_reader *pfile, unsigned *n_ptr, bool *varadic_ptr) -{ - unsigned nparms = 0; - bool ok = false; - - for (bool prev_ident = false;;) - { - const cpp_token *token = _cpp_lex_token (pfile); - - switch (token->type) - { - case CPP_COMMENT: - /* Allow/ignore comments in parameter lists if we are - preserving comments in macro expansions. */ - if (!CPP_OPTION (pfile, discard_comments_in_macro_exp)) - break; - - /* FALLTHRU */ - default: - bad: - { - const char *const msgs[5] = - { - N_("expected parameter name, found \"%s\""), - N_("expected ',' or ')', found \"%s\""), - N_("expected parameter name before end of line"), - N_("expected ')' before end of line"), - N_("expected ')' after \"...\"") - }; - unsigned ix = prev_ident; - const unsigned char *as_text = NULL; - if (*varadic_ptr) - ix = 4; - else if (token->type == CPP_EOF) - ix += 2; - else - as_text = cpp_token_as_text (pfile, token); - cpp_error (pfile, CPP_DL_ERROR, msgs[ix], as_text); - } - goto out; - - case CPP_NAME: - if (prev_ident || *varadic_ptr) - goto bad; - prev_ident = true; - - if (!_cpp_save_parameter (pfile, nparms, token->val.node.node, - token->val.node.spelling)) - goto out; - nparms++; - break; - - case CPP_CLOSE_PAREN: - if (prev_ident || !nparms || *varadic_ptr) - { - ok = true; - goto out; - } - - /* FALLTHRU */ - case CPP_COMMA: - if (!prev_ident || *varadic_ptr) - goto bad; - prev_ident = false; - break; - - case CPP_ELLIPSIS: - if (*varadic_ptr) - goto bad; - *varadic_ptr = true; - if (!prev_ident) - { - /* An ISO bare ellipsis. */ - _cpp_save_parameter (pfile, nparms, - pfile->spec_nodes.n__VA_ARGS__, - pfile->spec_nodes.n__VA_ARGS__); - nparms++; - pfile->state.va_args_ok = 1; - if (! CPP_OPTION (pfile, c99) - && CPP_OPTION (pfile, cpp_pedantic) - && CPP_OPTION (pfile, warn_variadic_macros)) - cpp_pedwarning - (pfile, CPP_W_VARIADIC_MACROS, - CPP_OPTION (pfile, cplusplus) - ? N_("anonymous variadic macros were introduced in C++11") - : N_("anonymous variadic macros were introduced in C99")); - else if (CPP_OPTION (pfile, cpp_warn_c90_c99_compat) > 0 - && ! CPP_OPTION (pfile, cplusplus)) - cpp_error (pfile, CPP_DL_WARNING, - "anonymous variadic macros were introduced in C99"); - } - else if (CPP_OPTION (pfile, cpp_pedantic) - && CPP_OPTION (pfile, warn_variadic_macros)) - cpp_pedwarning (pfile, CPP_W_VARIADIC_MACROS, - CPP_OPTION (pfile, cplusplus) - ? N_("ISO C++ does not permit named variadic macros") - : N_("ISO C does not permit named variadic macros")); - break; - } - } - - out: - *n_ptr = nparms; - - return ok; -} - -/* Lex a token from the expansion of MACRO, but mark parameters as we - find them and warn of traditional stringification. */ -static cpp_macro * -lex_expansion_token (cpp_reader *pfile, cpp_macro *macro) -{ - macro = (cpp_macro *)_cpp_reserve_room (pfile, - sizeof (cpp_macro) - sizeof (cpp_token) - + macro->count * sizeof (cpp_token), - sizeof (cpp_token)); - cpp_token *saved_cur_token = pfile->cur_token; - pfile->cur_token = ¯o->exp.tokens[macro->count]; - cpp_token *token = _cpp_lex_direct (pfile); - pfile->cur_token = saved_cur_token; - - /* Is this a parameter? */ - if (token->type == CPP_NAME && token->val.node.node->type == NT_MACRO_ARG) - { - /* Morph into a parameter reference. */ - cpp_hashnode *spelling = token->val.node.spelling; - token->type = CPP_MACRO_ARG; - token->val.macro_arg.arg_no = token->val.node.node->value.arg_index; - token->val.macro_arg.spelling = spelling; - } - else if (CPP_WTRADITIONAL (pfile) && macro->paramc > 0 - && (token->type == CPP_STRING || token->type == CPP_CHAR)) - check_trad_stringification (pfile, macro, &token->val.str); - - return macro; -} - -static cpp_macro * -create_iso_definition (cpp_reader *pfile) -{ - bool following_paste_op = false; - const char *paste_op_error_msg = - N_("'##' cannot appear at either end of a macro expansion"); - unsigned int num_extra_tokens = 0; - unsigned nparms = 0; - cpp_hashnode **params = NULL; - bool varadic = false; - bool ok = false; - cpp_macro *macro = NULL; - - /* Look at the first token, to see if this is a function-like - macro. */ - cpp_token first; - cpp_token *saved_cur_token = pfile->cur_token; - pfile->cur_token = &first; - cpp_token *token = _cpp_lex_direct (pfile); - pfile->cur_token = saved_cur_token; - - if (token->flags & PREV_WHITE) - /* Preceeded by space, must be part of expansion. */; - else if (token->type == CPP_OPEN_PAREN) - { - /* An open-paren, get a parameter list. */ - if (!parse_params (pfile, &nparms, &varadic)) - goto out; - - params = (cpp_hashnode **)_cpp_commit_buff - (pfile, sizeof (cpp_hashnode *) * nparms); - token = NULL; - } - else if (token->type != CPP_EOF - && !(token->type == CPP_COMMENT - && ! CPP_OPTION (pfile, discard_comments_in_macro_exp))) - { - /* While ISO C99 requires whitespace before replacement text - in a macro definition, ISO C90 with TC1 allows characters - from the basic source character set there. */ - if (CPP_OPTION (pfile, c99)) - cpp_error (pfile, CPP_DL_PEDWARN, - CPP_OPTION (pfile, cplusplus) - ? N_("ISO C++11 requires whitespace after the macro name") - : N_("ISO C99 requires whitespace after the macro name")); - else - { - enum cpp_diagnostic_level warntype = CPP_DL_WARNING; - switch (token->type) - { - case CPP_ATSIGN: - case CPP_AT_NAME: - case CPP_OBJC_STRING: - /* '@' is not in basic character set. */ - warntype = CPP_DL_PEDWARN; - break; - case CPP_OTHER: - /* Basic character set sans letters, digits and _. */ - if (strchr ("!\"#%&'()*+,-./:;<=>?[\\]^{|}~", - token->val.str.text[0]) == NULL) - warntype = CPP_DL_PEDWARN; - break; - default: - /* All other tokens start with a character from basic - character set. */ - break; - } - cpp_error (pfile, warntype, - "missing whitespace after the macro name"); - } - } - - macro = _cpp_new_macro (pfile, cmk_macro, - _cpp_reserve_room (pfile, 0, sizeof (cpp_macro))); - - if (!token) - { - macro->variadic = varadic; - macro->paramc = nparms; - macro->parm.params = params; - macro->fun_like = true; - } - else - { - /* Preserve the token we peeked, there is already a single slot for it. */ - macro->exp.tokens[0] = *token; - token = ¯o->exp.tokens[0]; - macro->count = 1; - } - - for (vaopt_state vaopt_tracker (pfile, macro->variadic, NULL);; token = NULL) - { - if (!token) - { - macro = lex_expansion_token (pfile, macro); - token = ¯o->exp.tokens[macro->count++]; - } - - /* Check the stringifying # constraint 6.10.3.2.1 of - function-like macros when lexing the subsequent token. */ - if (macro->count > 1 && token[-1].type == CPP_HASH && macro->fun_like) - { - if (token->type == CPP_MACRO_ARG - || (macro->variadic - && token->type == CPP_NAME - && token->val.node.node == pfile->spec_nodes.n__VA_OPT__)) - { - if (token->flags & PREV_WHITE) - token->flags |= SP_PREV_WHITE; - if (token[-1].flags & DIGRAPH) - token->flags |= SP_DIGRAPH; - token->flags &= ~PREV_WHITE; - token->flags |= STRINGIFY_ARG; - token->flags |= token[-1].flags & PREV_WHITE; - token[-1] = token[0]; - macro->count--; - } - /* Let assembler get away with murder. */ - else if (CPP_OPTION (pfile, lang) != CLK_ASM) - { - cpp_error (pfile, CPP_DL_ERROR, - "'#' is not followed by a macro parameter"); - goto out; - } - } - - if (token->type == CPP_EOF) - { - /* Paste operator constraint 6.10.3.3.1: - Token-paste ##, can appear in both object-like and - function-like macros, but not at the end. */ - if (following_paste_op) - { - cpp_error (pfile, CPP_DL_ERROR, paste_op_error_msg); - goto out; - } - if (!vaopt_tracker.completed ()) - goto out; - break; - } - - /* Paste operator constraint 6.10.3.3.1. */ - if (token->type == CPP_PASTE) - { - /* Token-paste ##, can appear in both object-like and - function-like macros, but not at the beginning. */ - if (macro->count == 1) - { - cpp_error (pfile, CPP_DL_ERROR, paste_op_error_msg); - goto out; - } - - if (following_paste_op) - { - /* Consecutive paste operators. This one will be moved - to the end. */ - num_extra_tokens++; - token->val.token_no = macro->count - 1; - } - else - { - /* Drop the paste operator. */ - --macro->count; - token[-1].flags |= PASTE_LEFT; - if (token->flags & DIGRAPH) - token[-1].flags |= SP_DIGRAPH; - if (token->flags & PREV_WHITE) - token[-1].flags |= SP_PREV_WHITE; - } - following_paste_op = true; - } - else - following_paste_op = false; - - if (vaopt_tracker.update (token) == vaopt_state::ERROR) - goto out; - } - - /* We're committed to winning now. */ - ok = true; - - /* Don't count the CPP_EOF. */ - macro->count--; - - macro = (cpp_macro *)_cpp_commit_buff - (pfile, sizeof (cpp_macro) - sizeof (cpp_token) - + sizeof (cpp_token) * macro->count); - - /* Clear whitespace on first token. */ - if (macro->count) - macro->exp.tokens[0].flags &= ~PREV_WHITE; - - if (num_extra_tokens) - { - /* Place second and subsequent ## or %:%: tokens in sequences of - consecutive such tokens at the end of the list to preserve - information about where they appear, how they are spelt and - whether they are preceded by whitespace without otherwise - interfering with macro expansion. Remember, this is - extremely rare, so efficiency is not a priority. */ - cpp_token *temp = (cpp_token *)_cpp_reserve_room - (pfile, 0, num_extra_tokens * sizeof (cpp_token)); - unsigned extra_ix = 0, norm_ix = 0; - cpp_token *exp = macro->exp.tokens; - for (unsigned ix = 0; ix != macro->count; ix++) - if (exp[ix].type == CPP_PASTE) - temp[extra_ix++] = exp[ix]; - else - exp[norm_ix++] = exp[ix]; - memcpy (&exp[norm_ix], temp, num_extra_tokens * sizeof (cpp_token)); - - /* Record there are extra tokens. */ - macro->extra_tokens = 1; - } - - out: - pfile->state.va_args_ok = 0; - _cpp_unsave_parameters (pfile, nparms); - - return ok ? macro : NULL; -} - -cpp_macro * -_cpp_new_macro (cpp_reader *pfile, cpp_macro_kind kind, void *placement) -{ - cpp_macro *macro = (cpp_macro *) placement; - - /* Zero init all the fields. This'll tell the compiler know all the - following inits are writing a virgin object. */ - memset (macro, 0, offsetof (cpp_macro, exp)); - - macro->line = pfile->directive_line; - macro->parm.params = 0; - macro->lazy = 0; - macro->paramc = 0; - macro->variadic = 0; - macro->used = !CPP_OPTION (pfile, warn_unused_macros); - macro->count = 0; - macro->fun_like = 0; - macro->imported_p = false; - macro->extra_tokens = 0; - /* To suppress some diagnostics. */ - macro->syshdr = pfile->buffer && pfile->buffer->sysp != 0; - - macro->kind = kind; - - return macro; -} - -/* Parse a macro and save its expansion. Returns nonzero on success. */ -bool -_cpp_create_definition (cpp_reader *pfile, cpp_hashnode *node) -{ - cpp_macro *macro; - - if (CPP_OPTION (pfile, traditional)) - macro = _cpp_create_trad_definition (pfile); - else - macro = create_iso_definition (pfile); - - if (!macro) - return false; - - if (cpp_macro_p (node)) - { - if (CPP_OPTION (pfile, warn_unused_macros)) - _cpp_warn_if_unused_macro (pfile, node, NULL); - - if (warn_of_redefinition (pfile, node, macro)) - { - const enum cpp_warning_reason reason - = (cpp_builtin_macro_p (node) && !(node->flags & NODE_WARN)) - ? CPP_W_BUILTIN_MACRO_REDEFINED : CPP_W_NONE; - - bool warned = - cpp_pedwarning_with_line (pfile, reason, - pfile->directive_line, 0, - "\"%s\" redefined", NODE_NAME (node)); - - if (warned && cpp_user_macro_p (node)) - cpp_error_with_line (pfile, CPP_DL_NOTE, - node->value.macro->line, 0, - "this is the location of the previous definition"); - } - _cpp_free_definition (node); - } - - /* Enter definition in hash table. */ - node->type = NT_USER_MACRO; - node->value.macro = macro; - if (! ustrncmp (NODE_NAME (node), DSC ("__STDC_")) - && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_FORMAT_MACROS") - /* __STDC_LIMIT_MACROS and __STDC_CONSTANT_MACROS are mentioned - in the C standard, as something that one must use in C++. - However DR#593 and C++11 indicate that they play no role in C++. - We special-case them anyway. */ - && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_LIMIT_MACROS") - && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_CONSTANT_MACROS")) - node->flags |= NODE_WARN; - - /* If user defines one of the conditional macros, remove the - conditional flag */ - node->flags &= ~NODE_CONDITIONAL; - - return true; -} - -extern void -cpp_define_lazily (cpp_reader *pfile, cpp_hashnode *node, unsigned num) -{ - cpp_macro *macro = node->value.macro; - - gcc_checking_assert (pfile->cb.user_lazy_macro && macro && num < UCHAR_MAX); - - macro->lazy = num + 1; -} - -/* NODE is a deferred macro, resolve it, returning the definition - (which may be NULL). */ -cpp_macro * -cpp_get_deferred_macro (cpp_reader *pfile, cpp_hashnode *node, - location_t loc) -{ - gcc_checking_assert (node->type == NT_USER_MACRO); - - node->value.macro = pfile->cb.user_deferred_macro (pfile, loc, node); - - if (!node->value.macro) - node->type = NT_VOID; - - return node->value.macro; -} - -static cpp_macro * -get_deferred_or_lazy_macro (cpp_reader *pfile, cpp_hashnode *node, - location_t loc) -{ - cpp_macro *macro = node->value.macro; - if (!macro) - { - macro = cpp_get_deferred_macro (pfile, node, loc); - gcc_checking_assert (!macro || !macro->lazy); - } - else if (macro->lazy) - { - pfile->cb.user_lazy_macro (pfile, macro, macro->lazy - 1); - macro->lazy = 0; - } - - return macro; -} - -/* Notify the use of NODE in a macro-aware context (i.e. expanding it, - or testing its existance). Also applies any lazy definition. - Return FALSE if the macro isn't really there. */ - -extern bool -_cpp_notify_macro_use (cpp_reader *pfile, cpp_hashnode *node, - location_t loc) -{ - node->flags |= NODE_USED; - switch (node->type) - { - case NT_USER_MACRO: - if (!get_deferred_or_lazy_macro (pfile, node, loc)) - return false; - /* FALLTHROUGH. */ - - case NT_BUILTIN_MACRO: - if (pfile->cb.used_define) - pfile->cb.used_define (pfile, loc, node); - break; - - case NT_VOID: - if (pfile->cb.used_undef) - pfile->cb.used_undef (pfile, loc, node); - break; - - default: - abort (); - } - - return true; -} - -/* Warn if a token in STRING matches one of a function-like MACRO's - parameters. */ -static void -check_trad_stringification (cpp_reader *pfile, const cpp_macro *macro, - const cpp_string *string) -{ - unsigned int i, len; - const uchar *p, *q, *limit; - - /* Loop over the string. */ - limit = string->text + string->len - 1; - for (p = string->text + 1; p < limit; p = q) - { - /* Find the start of an identifier. */ - while (p < limit && !is_idstart (*p)) - p++; - - /* Find the end of the identifier. */ - q = p; - while (q < limit && is_idchar (*q)) - q++; - - len = q - p; - - /* Loop over the function macro arguments to see if the - identifier inside the string matches one of them. */ - for (i = 0; i < macro->paramc; i++) - { - const cpp_hashnode *node = macro->parm.params[i]; - - if (NODE_LEN (node) == len - && !memcmp (p, NODE_NAME (node), len)) - { - cpp_warning (pfile, CPP_W_TRADITIONAL, - "macro argument \"%s\" would be stringified in traditional C", - NODE_NAME (node)); - break; - } - } - } -} - -/* Returns the name, arguments and expansion of a macro, in a format - suitable to be read back in again, and therefore also for DWARF 2 - debugging info. e.g. "PASTE(X, Y) X ## Y", or "MACNAME EXPANSION". - Caller is expected to generate the "#define" bit if needed. The - returned text is temporary, and automatically freed later. */ -const unsigned char * -cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node) -{ - gcc_checking_assert (cpp_user_macro_p (node)); - - if (const cpp_macro *macro = get_deferred_or_lazy_macro (pfile, node, 0)) - return cpp_macro_definition (pfile, node, macro); - return NULL; -} - -const unsigned char * -cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node, - const cpp_macro *macro) -{ - unsigned int i, len; - unsigned char *buffer; - - /* Calculate length. */ - len = NODE_LEN (node) * 10 + 2; /* ' ' and NUL. */ - if (macro->fun_like) - { - len += 4; /* "()" plus possible final ".." of named - varargs (we have + 1 below). */ - for (i = 0; i < macro->paramc; i++) - len += NODE_LEN (macro->parm.params[i]) + 1; /* "," */ - } - - /* This should match below where we fill in the buffer. */ - if (CPP_OPTION (pfile, traditional)) - len += _cpp_replacement_text_len (macro); - else - { - unsigned int count = macro_real_token_count (macro); - for (i = 0; i < count; i++) - { - const cpp_token *token = ¯o->exp.tokens[i]; - - if (token->type == CPP_MACRO_ARG) - len += NODE_LEN (token->val.macro_arg.spelling); - else - len += cpp_token_len (token); - - if (token->flags & STRINGIFY_ARG) - len++; /* "#" */ - if (token->flags & PASTE_LEFT) - len += 3; /* " ##" */ - if (token->flags & PREV_WHITE) - len++; /* " " */ - } - } - - if (len > pfile->macro_buffer_len) - { - pfile->macro_buffer = XRESIZEVEC (unsigned char, - pfile->macro_buffer, len); - pfile->macro_buffer_len = len; - } - - /* Fill in the buffer. Start with the macro name. */ - buffer = pfile->macro_buffer; - buffer = _cpp_spell_ident_ucns (buffer, node); - - /* Parameter names. */ - if (macro->fun_like) - { - *buffer++ = '('; - for (i = 0; i < macro->paramc; i++) - { - cpp_hashnode *param = macro->parm.params[i]; - - if (param != pfile->spec_nodes.n__VA_ARGS__) - { - memcpy (buffer, NODE_NAME (param), NODE_LEN (param)); - buffer += NODE_LEN (param); - } - - if (i + 1 < macro->paramc) - /* Don't emit a space after the comma here; we're trying - to emit a Dwarf-friendly definition, and the Dwarf spec - forbids spaces in the argument list. */ - *buffer++ = ','; - else if (macro->variadic) - *buffer++ = '.', *buffer++ = '.', *buffer++ = '.'; - } - *buffer++ = ')'; - } - - /* The Dwarf spec requires a space after the macro name, even if the - definition is the empty string. */ - *buffer++ = ' '; - - if (CPP_OPTION (pfile, traditional)) - buffer = _cpp_copy_replacement_text (macro, buffer); - else if (macro->count) - /* Expansion tokens. */ - { - unsigned int count = macro_real_token_count (macro); - for (i = 0; i < count; i++) - { - const cpp_token *token = ¯o->exp.tokens[i]; - - if (token->flags & PREV_WHITE) - *buffer++ = ' '; - if (token->flags & STRINGIFY_ARG) - *buffer++ = '#'; - - if (token->type == CPP_MACRO_ARG) - { - memcpy (buffer, - NODE_NAME (token->val.macro_arg.spelling), - NODE_LEN (token->val.macro_arg.spelling)); - buffer += NODE_LEN (token->val.macro_arg.spelling); - } - else - buffer = cpp_spell_token (pfile, token, buffer, true); - - if (token->flags & PASTE_LEFT) - { - *buffer++ = ' '; - *buffer++ = '#'; - *buffer++ = '#'; - /* Next has PREV_WHITE; see _cpp_create_definition. */ - } - } - } - - *buffer = '\0'; - return pfile->macro_buffer; -} - - -/*-------------------------------------------------------------------------------- - RT extensions ---------------------------------------------------------------------------------*/ - -/*-------------------------------------------------------------------------------- - shared declarations -*/ - - typedef enum parse_clause_status { - PCS_COMPLETE // Clause completely parsed - ,PCS_ERR_EXPECTED_OPEN_DELIM // Failed to find expected opening '(' - ,PCS_ERR_UNEXPECTED_EOF // Hit real EOF before matching ')' - ,PCS_ERR_PASTE_AT_END // Trailing '##' paste operator - ,PCS_ERR_HASH_NOT_FOLLOWED_BY_ARG // '#' not followed by macro parameter - ,PCS_ERR_VAOPT_STATE_INVALID // __VA_OPT__ or variadic tracking error - ,PCS_ERR_EOF_FETCH_FAILED // Failed to fetch next line after EOF - ,PCS_ERR_UNKNOWN // Fallback error (should not occur) - ,PCS_ERR_STATUS_NOT_SET // function did not set the status - } parse_clause_status; - - -/*-------------------------------------------------------------------------------- - debug helpers -*/ - - #define DebugParseClause 1 - #define DebugAssign 1 - #define DebugRTMacro 1 - - #define DebugHelpers 1 - - #if DebugHelpers - - static const char * - ttype_to_text(enum cpp_ttype ttype) - { - switch (ttype) - { - case CPP_EOF: return "EOF"; - case CPP_PADDING: return "PADDING"; - case CPP_COMMENT: return "COMMENT"; - // case CPP_HSPACE: return "HSPACE"; - // case CPP_VSPACE: return "VSPACE"; - case CPP_OTHER: return "OTHER"; - case CPP_OPEN_PAREN: return "OPEN_PAREN"; - case CPP_CLOSE_PAREN: return "CLOSE_PAREN"; - case CPP_OPEN_SQUARE: return "OPEN_SQUARE"; - case CPP_CLOSE_SQUARE: return "CLOSE_SQUARE"; - case CPP_OPEN_BRACE: return "OPEN_BRACE"; - case CPP_CLOSE_BRACE: return "CLOSE_BRACE"; - case CPP_COMMA: return "COMMA"; - case CPP_SEMICOLON: return "SEMICOLON"; - case CPP_ELLIPSIS: return "ELLIPSIS"; - case CPP_NAME: return "NAME"; - case CPP_NUMBER: return "NUMBER"; - case CPP_CHAR: return "CHAR"; - case CPP_STRING: return "STRING"; - case CPP_HEADER_NAME: return "HEADER_NAME"; - case CPP_PLUS: return "PLUS"; - case CPP_MINUS: return "MINUS"; - case CPP_MULT: return "MULT"; - case CPP_DIV: return "DIV"; - case CPP_MOD: return "MOD"; - case CPP_AND: return "AND"; - case CPP_OR: return "OR"; - case CPP_XOR: return "XOR"; - case CPP_NOT: return "NOT"; - case CPP_LSHIFT: return "LSHIFT"; - case CPP_RSHIFT: return "RSHIFT"; - case CPP_EQ: return "EQ"; - // case CPP_NE: return "NE"; - // case CPP_LE: return "LE"; - // case CPP_GE: return "GE"; - // case CPP_LT: return "LT"; - // case CPP_GT: return "GT"; - case CPP_ATSIGN: return "@"; - case CPP_PLUS_EQ: return "PLUS_EQ"; - case CPP_MINUS_EQ: return "MINUS_EQ"; - case CPP_MULT_EQ: return "MULT_EQ"; - case CPP_DIV_EQ: return "DIV_EQ"; - case CPP_MOD_EQ: return "MOD_EQ"; - case CPP_AND_EQ: return "AND_EQ"; - case CPP_OR_EQ: return "OR_EQ"; - case CPP_XOR_EQ: return "XOR_EQ"; - case CPP_LSHIFT_EQ: return "LSHIFT_EQ"; - case CPP_RSHIFT_EQ: return "RSHIFT_EQ"; - // case CPP_CONDITIONAL: return "CONDITIONAL"; - case CPP_COLON: return "COLON"; - case CPP_DEREF: return "DEREF"; - case CPP_DOT: return "DOT"; - case CPP_DEREF_STAR: return "DEREF_STAR"; - case CPP_DOT_STAR: return "DOT_STAR"; - // case CPP_INCREMENT: return "INCREMENT"; - // case CPP_DECREMENT: return "DECREMENT"; - default: return ""; - } - } - - static void - print_ttype(enum cpp_ttype ttype){ - fprintf(stderr, "%s (%d)", ttype_to_text(ttype), ttype); - } - - const char *cpp_token_as_text(const cpp_token *token){ - static char buffer[256]; - - switch (token->type) - { - case CPP_NAME: - snprintf(buffer, sizeof(buffer), "CPP_NAME: '%s'", - NODE_NAME(token->val.node.node)); - break; - - case CPP_NUMBER: - case CPP_STRING: - case CPP_CHAR: - case CPP_HEADER_NAME: - snprintf(buffer, sizeof(buffer), "'%.*s'", - token->val.str.len, - token->val.str.text); - break; - - case CPP_EOF: - return ""; - case CPP_OTHER: - return ""; - case CPP_OPEN_PAREN: - return "'('"; - case CPP_CLOSE_PAREN: - return "')'"; - case CPP_COMMA: - return "','"; - case CPP_SEMICOLON: - return "';'"; - case CPP_PLUS: - return "'+'"; - case CPP_MINUS: - return "'-'"; - case CPP_MULT: - return "'*'"; - case CPP_DIV: - return "'/'"; - case CPP_MOD: - return "'%'"; - case CPP_MACRO_ARG: - snprintf( - buffer - ,sizeof(buffer) - ,"CPP_MACRO_ARG: '%s'" - ,NODE_NAME(token->val.macro_arg.spelling) - ); - break; - case CPP_PADDING: return ""; - case CPP_COMMENT: return ""; - case CPP_HASH: return "'#'"; - case CPP_PASTE: return "'##'"; - case CPP_ELLIPSIS: return "'...'"; - case CPP_COLON: return "':'"; - case CPP_OPEN_SQUARE: return "'['"; - case CPP_CLOSE_SQUARE: return "']'"; - case CPP_OPEN_BRACE: return "'{'"; - case CPP_CLOSE_BRACE: return "'}'"; - case CPP_DOT: return "'.'"; - case CPP_DEREF: return "'->'"; - case CPP_SCOPE: return "'::'"; - case CPP_DOT_STAR: return "'.*'"; - case CPP_DEREF_STAR: return "'->*'"; - case CPP_PRAGMA: return "<_Pragma>"; - case CPP_KEYWORD: return ""; - - default: - snprintf(buffer, sizeof(buffer), "", token->type); - break; - } - - // Append token flags if any are set - if (token->flags & (PREV_WHITE | DIGRAPH | STRINGIFY_ARG | - PASTE_LEFT | NAMED_OP | BOL | PURE_ZERO | - SP_DIGRAPH | SP_PREV_WHITE | NO_EXPAND | PRAGMA_OP)) - { - size_t len = strlen(buffer); - snprintf(buffer + len, sizeof(buffer) - len, " [flags:"); - - if (token->flags & PREV_WHITE) - strncat(buffer, " PREV_WHITE", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & DIGRAPH) - strncat(buffer, " DIGRAPH", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & STRINGIFY_ARG) - strncat(buffer, " STRINGIFY", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & PASTE_LEFT) - strncat(buffer, " ##L", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & NAMED_OP) - strncat(buffer, " NAMED_OP", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & BOL) - strncat(buffer, " BOL", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & PURE_ZERO) - strncat(buffer, " ZERO", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & SP_DIGRAPH) - strncat(buffer, " ##DIGRAPH", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & SP_PREV_WHITE) - strncat(buffer, " SP_WHITE", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & NO_EXPAND) - strncat(buffer, " NO_EXPAND", sizeof(buffer) - strlen(buffer) - 1); - if (token->flags & PRAGMA_OP) - strncat(buffer, " _Pragma", sizeof(buffer) - strlen(buffer) - 1); - - strncat(buffer, " ]", sizeof(buffer) - strlen(buffer) - 1); - } - - return buffer; - } - - void print_token_list(const cpp_token *tokens ,size_t count){ - for (size_t i = 0; i < count; ++i) - fprintf( stderr ,"[%zu] %s\n" ,i , cpp_token_as_text(&tokens[i]) ); - } - - void print_parse_clause_status(enum parse_clause_status status){ - const char *message = NULL; - switch (status) - { - case PCS_COMPLETE: - message = "parse_clause status is OK"; - break; - case PCS_ERR_EXPECTED_OPEN_DELIM: - message = "expected opening delimiter such as '(' but did not find it."; - break; - case PCS_ERR_UNEXPECTED_EOF: - message = "unexpected EOF before closing ')'."; - break; - case PCS_ERR_PASTE_AT_END: - message = "paste operator '##' appeared at the beginning or end of macro body."; - break; - case PCS_ERR_HASH_NOT_FOLLOWED_BY_ARG: - message = "'#' was not followed by a valid macro parameter."; - break; - case PCS_ERR_VAOPT_STATE_INVALID: - message = "invalid __VA_OPT__ tracking state."; - break; - case PCS_ERR_EOF_FETCH_FAILED: - message = "_cpp_get_fresh_line() failed to fetch next line."; - break; - case PCS_ERR_STATUS_NOT_SET: - message = "Internal Error, status was not set"; - break; - case PCS_ERR_UNKNOWN: - default: - message = "unknown or unhandled error."; - break; - } - fprintf(stderr, "%s\n", message); - } - - // a helper function for probing where the parser thinks it is in the source - void debug_peek_token (cpp_reader *pfile){ - const cpp_token *tok = _cpp_lex_token(pfile); - - cpp_error_with_line( - pfile, - CPP_DL_ERROR, - tok->src_loc, - 0, - "DEBUG: next token is: `%s`", - (const char *) cpp_token_as_text(tok) - ); - - _cpp_backup_tokens (pfile, 1); - - } - -#endif - -/*-------------------------------------------------------------------------------- - Parse a clause - - clause ::= "(" literal? ")" - | "[" expr? "]" - | tokens_to_eol ; - - literal ::= ; sequence parsed into tokens, no expansion - expr ::= ; sequence parsed into tokens with recursive expansion of each token - tokens_to_eol ::= ; all tokens until logical end-of-line (including multi-line with `\`) - - Notes: - - The first two forms are explicitly delimited with parentheses or brackets, - and may be empty (e.g., `()` or `[]`). Newlines are taken as white space. - - The third form is implicit: it consumes all remaining tokens on the directive line. - This is typical for simple macro bodies (e.g., in `#define NAME body`). - -*/ - - -/* - Caller sees an open parenthesis or other open delimiter, and calls this. - - This parses tokens until seeing the closing delimiter. - - delimiter_matching == true: balances opening and closing delimiter types while searching for the balanced closing delimiter. - - paren_matching == false; terminating delimiter is CPP_EOF - which to the lexer will be end of the line. (That is how the cpp_reader does it.) - - comma_list: when true, the comma becomes an alias for the final closing delimiter. For - balanced delimiters only a comma at level 0 is a terminating delimiter. - -*/ - - -static enum parse_clause_status -parse_clause_expand(cpp_reader *pfile, - cpp_macro *macro, - bool delimiter_matching, - enum cpp_ttype opening, - enum cpp_ttype closing, - bool comma_list, - location_t *src_loc_pt, - cpp_ttype *terminator_out) -{ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_expand\n"); - fprintf(stderr, " delimiter_matching: %s\n", delimiter_matching ? "true" : "false"); - fprintf(stderr, " opening token: %s (%d)\n", ttype_to_text(opening), opening); - fprintf(stderr, " closing token: %s (%d)\n", ttype_to_text(closing), closing); - fprintf(stderr, " comma_list: %s\n", comma_list ? "true" : "false"); - fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); - #endif - - int nesting_depth = 1; - const cpp_token *token; - - for(;;){ - - /* get a token - */ - // why not use cpp_get_token? or cpp_get_token_no_padding? - token = cpp_get_token_1 (pfile, src_loc_pt); - // this is necessary for the name expr, but does it impact potential other uses of parse_clause? Another flag for this perhaps? - if(token->type == CPP_PADDING) continue; - macro = (cpp_macro *)_cpp_reserve_room( - pfile, - sizeof(cpp_macro) + macro->count * sizeof(cpp_token), - sizeof(cpp_token) - ); - macro->exp.tokens[macro->count] = *token; - #if DebugParseClause - fprintf( stderr, "token: %s\n", cpp_token_as_text(token) ); - #endif - - // lexer supports line macros by inserting CPP_EOF at line ends - if(delimiter_matching && token->type == CPP_EOF){ - #if DebugParseClause - fprintf( stderr, "CPP_EOF during parse with parentheses matching \n"); - #endif - if(!_cpp_get_fresh_line(pfile)){ - return PCS_ERR_EOF_FETCH_FAILED; - } - continue; - } - - /* parentheses matching overhead - */ - if(delimiter_matching){ - - if (token->type == opening) { - nesting_depth++; - } - else if (token->type == closing) { - nesting_depth--; - if (nesting_depth < 0) { - cpp_error(pfile, CPP_DL_ERROR, "unmatched closing delimiter"); - return PCS_ERR_UNEXPECTED_EOF; - } - } - - #if DebugParseClause - if( token->type == opening || token->type == closing){ - fprintf( stderr, "new nesting_depth: %d\n", nesting_depth); - } - #endif - } - - - /* Determine if routine has lexed the final macro body token and should exit. - */ - bool terminted_by_matched_delimiter = - delimiter_matching - && nesting_depth == 0 - && (token->type == closing || comma_list && token->type == CPP_COMMA) - ; - - bool terminated_by_EOL = - !delimiter_matching - && (token->type == CPP_EOF || comma_list && token->type == CPP_COMMA) - ; - - if(terminted_by_matched_delimiter || terminated_by_EOL){ - if(terminator_out) *terminator_out = token->type; - return PCS_COMPLETE; - } - - // commit the new token - macro->count++; - - }// end for next token loop - -} - -bool cgls_flag = false; - -/* - See notes on parse_clause_expand - - This is the same but tokens in the clause are not expanded. - - The end case tests here probably need to badded to parse_clause_expand also. - Perhaps expansion can be another mode bit to be sent int. - - I would have been better perhaps to send in a pointer to a token allocation, - instead of to a src_loc, and and terminal type. -*/ -static enum parse_clause_status parse_clause_literal( - cpp_reader *pfile - ,cpp_macro *macro - ,bool delimiter_matching - ,enum cpp_ttype opening - ,enum cpp_ttype closing - ,bool comma_list - ,location_t *src_loc_pt - ,cpp_ttype *terminator_out - ,unsigned int *num_extra_tokens_out -){ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_literal\n"); - fprintf(stderr, " delimiter_matching: %s\n", delimiter_matching ? "true" : "false"); - fprintf(stderr, " opening token: %s (%d)\n", ttype_to_text(opening), opening); - fprintf(stderr, " closing token: %s (%d)\n", ttype_to_text(closing), closing); - fprintf(stderr, " comma_list: %s\n", comma_list ? "true" : "false"); - fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); - #endif - - bool following_paste_op = false; - unsigned int num_extra_tokens = 0; - int nesting_depth = 1; - cpp_token *lex_token; - const char *paste_op_error_msg = - N_("'##' cannot appear at either end of a macro expansion"); - - - for(vaopt_state vaopt_tracker (pfile, macro->variadic, NULL);;){ - - /* get a token - */ - // first parses lex_token onto `macro->exp.tokens[macro->count]` - // then pulls the token off of `macro->exp.tokens[macro->count]` - // reassigns macro due to possible macro->exp.tokens buffer expansion - macro = lex_expansion_token(pfile, macro); - lex_token = ¯o->exp.tokens[macro->count]; - *src_loc_pt = lex_token->src_loc; - #if DebugParseClause - fprintf( stderr, "lex_token: %s\n", cpp_token_as_text(lex_token) ); - #endif - - // lexer will insert CPP_EOF at the end of each line, because cpp originally only did line macros. - if(delimiter_matching && lex_token->type == CPP_EOF){ - #if DebugParseClause - fprintf( stderr, "CPP_EOF during parse with parentheses matching \n"); - #endif - if(!_cpp_get_fresh_line(pfile)){ - return PCS_ERR_EOF_FETCH_FAILED; - } - continue; - } - - /* tag macro args - */ - if (macro->count > 1 && lex_token[-1].type == CPP_HASH && macro->fun_like) - { - if (lex_token->type == CPP_MACRO_ARG - || (macro->variadic - && lex_token->type == CPP_NAME - && lex_token->val.node.node == pfile->spec_nodes.n__VA_OPT__)) - { - if (lex_token->flags & PREV_WHITE) - lex_token->flags |= SP_PREV_WHITE; - if (lex_token[-1].flags & DIGRAPH) - lex_token->flags |= SP_DIGRAPH; - lex_token->flags &= ~PREV_WHITE; - lex_token->flags |= STRINGIFY_ARG; - lex_token->flags |= lex_token[-1].flags & PREV_WHITE; - lex_token[-1] = lex_token[0]; - macro->count--; - } - else if (CPP_OPTION (pfile, lang) != CLK_ASM) - { - cpp_error(pfile, CPP_DL_ERROR, - "'#' is not followed by a macro parameter"); - return PCS_ERR_HASH_NOT_FOLLOWED_BY_ARG; - } - } - - /* paste end cases - */ - if (lex_token->type == CPP_PASTE) - { - if (macro->count == 0) - { - cpp_error(pfile, CPP_DL_ERROR, paste_op_error_msg); - return PCS_ERR_PASTE_AT_END; // the font end of the buffer - } - - if (following_paste_op) - { - num_extra_tokens++; - lex_token->val.token_no = macro->count - 1; - } - else - { - --macro->count; - lex_token[-1].flags |= PASTE_LEFT; - if (lex_token->flags & DIGRAPH) - lex_token[-1].flags |= SP_DIGRAPH; - if (lex_token->flags & PREV_WHITE) - lex_token[-1].flags |= SP_PREV_WHITE; - } - following_paste_op = true; - } - else{ - following_paste_op = false; - } - - /* parentheses matching overhead - */ - if(delimiter_matching){ - - if (lex_token->type == opening) { - nesting_depth++; - } - else if (lex_token->type == closing) { - nesting_depth--; - if (nesting_depth < 0) { - cpp_error(pfile, CPP_DL_ERROR, "unmatched closing delimiter"); - return PCS_ERR_UNEXPECTED_EOF; - } - } - - #if DebugParseClause - if( lex_token->type == opening || lex_token->type == closing){ - fprintf( stderr, "new nesting_depth: %d\n", nesting_depth); - } - #endif - } - - /* Determine if routine has lexed the final macro body token and should exit. - */ - bool terminated_by_matched_delimiter = - delimiter_matching - && nesting_depth == 0 - && (lex_token->type == closing || comma_list && lex_token->type == CPP_COMMA) - ; - - bool terminated_by_EOL = - !delimiter_matching - && (lex_token->type == CPP_EOF || comma_list && lex_token->type == CPP_COMMA) - ; - - if(terminated_by_matched_delimiter || terminated_by_EOL){ - - if(following_paste_op){ - cpp_error(pfile, CPP_DL_ERROR, paste_op_error_msg); - return PCS_ERR_PASTE_AT_END; - } - if (vaopt_tracker.update(lex_token) == vaopt_state::ERROR){ - return PCS_ERR_VAOPT_STATE_INVALID; - } - if( !vaopt_tracker.completed() ){ - return PCS_ERR_VAOPT_STATE_INVALID; - } - - if(num_extra_tokens_out) *num_extra_tokens_out = num_extra_tokens; - if(terminator_out) *terminator_out = lex_token->type; - return PCS_COMPLETE; - } - - // commit the new token - macro->count++; - - }// end for next token loop - -} - -/* - The opening paren has been parsed, this completes parsing the clause. - - Given a cpp_macro and cpp_reader reference. - Returns the body tokens in `macro->exp.tokens`. - - The macro need not have been committed. -*/ - -static enum parse_clause_status parse_clause_balanced( - cpp_reader *pfile - ,cpp_macro *macro - ,enum cpp_ttype opening - ,enum cpp_ttype closing - ,bool expand - ,bool comma_list - ,location_t *src_loc_pt - ,cpp_ttype *terminator_out - ,unsigned int *num_extra_tokens_out -){ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_balanced\n"); - fprintf(stderr, " opening token: %s (%d)\n", ttype_to_text(opening), opening); - fprintf(stderr, " closing token: %s (%d)\n", ttype_to_text(closing), closing); - fprintf(stderr, " comma_list: %s\n", comma_list ? "true" : "false"); - fprintf(stderr, " expand: %s\n", expand ? "true" : "false"); - fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); - #endif - - /* check for opening paren - */ - const cpp_token *token = _cpp_lex_token(pfile); - *src_loc_pt = token->src_loc; - #if DebugParseClause - fprintf(stderr, "checking if token is opening: %s" ,cpp_token_as_text(token)); - #endif - - if( token->type != opening ){ - _cpp_backup_tokens(pfile, 1); - return PCS_ERR_EXPECTED_OPEN_DELIM; - } - - /* complete the parse - */ - - // make comma an alias for the terminating delimiter - // this gets passed on directly to parse_clause_{literal,expand} - bool delimiter_matching = true; - - int saved_keep_tokens = pfile->keep_tokens; - int saved_in_directive = pfile->state.in_directive; - pfile->keep_tokens = 1; - pfile->state.in_directive = 0; - - parse_clause_status status; - if(expand){ - status = parse_clause_expand - (pfile - ,macro - ,delimiter_matching - ,opening ,closing - ,comma_list - ,src_loc_pt - ,terminator_out - ); - if(num_extra_tokens_out) *num_extra_tokens_out = 0; - #if DebugParseClause - fprintf(stderr, ">> parse_clause_balanced expand status:"); - print_parse_clause_status(status); - #endif - }else{ - status = parse_clause_literal - (pfile - ,macro - ,delimiter_matching - ,opening ,closing - ,comma_list - ,src_loc_pt - ,terminator_out - ,num_extra_tokens_out - ); - #if DebugParseClause - fprintf(stderr, ">> parse_clause_balanced literal status:"); - print_parse_clause_status(status); - #endif - } - - // Restore parser state - pfile->keep_tokens = saved_keep_tokens; - pfile->state.in_directive = saved_in_directive; - - #if DebugParseClause - fprintf(stderr, ">> parse_clause_balanced final returning status:"); - print_parse_clause_status(status); - #endif - return status; -} - -typedef enum PCPSO_choice{ - PCPSO_unknown = 0 - ,PCPSO_paren - ,PCPSO_square -} PCPSO_choice; - -parse_clause_status parse_clause_paren_square_option( - cpp_reader *pfile - ,cpp_macro *macro - ,PCPSO_choice *choice - ,bool comma_list - ,location_t *src_loc_pt - ,cpp_ttype *terminator_out - ,unsigned int *num_extra_tokens_out -){ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_paren_square_option\n"); - fprintf(stderr, " comma_list: %s\n", comma_list ? "true" : "false"); - fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); - #endif - - *choice = PCPSO_unknown; - - parse_clause_status status = parse_clause_balanced - (pfile - ,macro - ,CPP_OPEN_PAREN - ,CPP_CLOSE_PAREN - ,false // no expand - ,comma_list - ,src_loc_pt - ,terminator_out - ,num_extra_tokens_out - ); - - if(status != PCS_ERR_EXPECTED_OPEN_DELIM){ - *choice = PCPSO_paren; - return status; - } - - status = parse_clause_balanced - (pfile - ,macro - ,CPP_OPEN_SQUARE - ,CPP_CLOSE_SQUARE - ,true // expand - ,comma_list - ,src_loc_pt - ,terminator_out - ,num_extra_tokens_out - ); - - if(status == PCS_ERR_EXPECTED_OPEN_DELIM){ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_paren_square_option early returning status:"); - print_parse_clause_status(PCS_ERR_EXPECTED_OPEN_DELIM); - #endif - return PCS_ERR_EXPECTED_OPEN_DELIM; - } - - *choice = PCPSO_square; - - #if DebugParseClause - fprintf(stderr, ">> parse_clause_paren_square_option final returning status:"); - print_parse_clause_status(status); - #endif - return status; -} - -/* -The prior clause parse terminated in a comma. For parsing a comma list of clauses. -*/ -static enum parse_clause_status parse_clause_comma_continue( -){ - return PCS_ERR_UNKNOWN; -} - -/* - Check if a collected macro body reduces to a single identifier token. - - Preconditions: - - macro is non-null - - macro->exp.tokens has been populated (e.g., via parse_clause) - - macro->count is valid - - Returns: - - cpp_hashnode* if valid (i.e., single CPP_NAME token) - - NULL if invalid, and emits error message - - -Note in do_define in directives.cc there is some logic related to callbacks and warning if trying to redefine a built-in macro. That should be integrated here. - - -*/ -static cpp_hashnode * -name_clause_is_name(cpp_reader *pfile, const cpp_macro *macro) -{ - if (!macro || macro->count != 1) - { - cpp_error(pfile, CPP_DL_ERROR, - "expected exactly one token in assign name expression, got %u", - macro ? macro->count : 0); - return NULL; - } - - const cpp_token *tok = ¯o->exp.tokens[0]; - - if (tok->type != CPP_NAME) - { - cpp_error(pfile, CPP_DL_ERROR, - "expected identifier in assign name expression, got: %s", - cpp_token_as_text(tok)); - return NULL; - } - - return tok->val.node.node; -} - - -/*-------------------------------------------------------------------------------- - `#assign` directive RT extension - - called from directives.cc::do_assign() - -*/ - -bool _cpp_create_assign(cpp_reader *pfile){ - - parse_clause_status status; - location_t src_loc; - unsigned int num_extra_tokens = 0; - PCPSO_choice choice; - - /* Parse name clause into a temporary macro. - - This macro will not be committed, so it will be overwritten on the next _cpp_new_macro call. - */ - cpp_macro *name_macro = _cpp_new_macro( - pfile - ,cmk_macro - ,_cpp_reserve_room( pfile, 0, sizeof(cpp_macro) ) - ); - name_macro->variadic = false; - name_macro->paramc = 0; - name_macro->parm.params = NULL; - name_macro->fun_like = false; - - status = parse_clause_paren_square_option( - pfile - ,name_macro - ,&choice - ,false // not a commas list - ,&src_loc - ,NULL // don't need to know the terminator - ,&num_extra_tokens - ); - - - #if DebugAssign - fprintf(stderr,"name_macro->count: %d\n" ,name_macro->count); - fprintf(stderr,"assign directive name tokens:\n"); - print_token_list(name_macro->exp.tokens ,name_macro->count); - #endif - - /* The name clause must be either a literally valid name, or it must expand into - a valid name, depending if the programmer used () or []. - If valid, keep the name node. - */ - cpp_hashnode *name_node = name_clause_is_name(pfile ,name_macro); - if(name_node){ - #if DebugAssign - fprintf( - stderr - ,"assign macro name: '%.*s'\n" - ,(int) NODE_LEN(name_node) - ,NODE_NAME(name_node) - ); - #endif - }else{ - #if DebugAssign - fprintf(stderr, "node is not a name\n"); - #endif - return false; - } - - /* Unpaint name_node - - There are three scenarios where name_node will already exist in the symbol table - before the name clause of `#assign` is evaluated: - - 1. A macro definition already exists for name_node, and the name clause - is not expanded (i.e., it was delineated with '()'). - - 2. A macro definition exists, and the name clause *is* expanded (i.e., it - was delineated with '[]'), but name_node was painted and thus skipped - during expansion. - - 3. A macro definition exists and was not painted initially, but the name - clause expands recursively to itself (e.g., `list -> list`), resulting - in name_node being painted *during* the name clause evaluation. - - After the name clause is parsed, the body clause might be expanded. If so, - name_node must not be painted — this ensures that it will expand at least once. This enables patterns like: - - #assign ()(list)(list second) - - ...to work even if 'list' was painted prior to entering #assign. - - If the macro recurs during evaluation of the body clause, it will be automatically painted by the expansion engine, as usual. - - Note also: upon exit from this routine, the newly created macro will *not* be painted. Its disabled flag will remain clear. - - Consequently, for a recursive macro, assign can be called repeatedly to get 'one more level' of evaluation upon each call. - */ - if (cpp_macro_p(name_node)) { - name_node->flags &= ~NODE_DISABLED; - } - - /* create a new macro and put the #assign body clause in it - */ - cpp_macro *body_macro = _cpp_new_macro( - pfile - ,cmk_macro - ,_cpp_reserve_room( pfile, 0, sizeof(cpp_macro) ) - ); - body_macro->variadic = false; - body_macro->paramc = 0; - body_macro->parm.params = NULL; - body_macro->fun_like = false; - - status = parse_clause_paren_square_option( - pfile - ,body_macro - ,&choice - ,false // not a commas list - ,&src_loc - ,NULL // don't need to know the terminator - ,&num_extra_tokens - ); - #if DebugAssign - fprintf(stderr,"assign directive body tokens:\n"); - print_token_list(body_macro->exp.tokens ,body_macro->count); - #endif - - - cpp_macro *assign_macro = (cpp_macro *)_cpp_commit_buff( - pfile - ,sizeof(cpp_macro) - sizeof(cpp_token) + sizeof(cpp_token) * body_macro->count - ); - assign_macro->count = body_macro->count; - memcpy( - assign_macro->exp.tokens - ,body_macro->exp.tokens - ,sizeof(cpp_token) * body_macro->count - ); - body_macro->variadic = false; - body_macro->paramc = 0; - body_macro->parm.params = NULL; - body_macro->fun_like = false; - - /* Install the assign macro under name_node. - - If name_node previously had a macro definition, discard it. - Then install the new macro, and clear any disabled flag. - - This ensures the assigned macro can be expanded immediately, - even if it appeared in its own body clause and was painted. - */ - name_node->flags &= ~NODE_USED; - - if (cpp_macro_p(name_node)) { - // There is no mechanism in libcpp to free the memory taken by a committed macro, but wec an cast it adrift. - name_node->value.macro = NULL; - } - name_node->type = NT_USER_MACRO; - name_node->value.macro = assign_macro; - name_node->flags &= ~NODE_DISABLED; - - /* all done - */ - #if DebugAssign - fprintf( - stderr - ,"macro '%.*s' assigned successfully.\n\n" - ,(int) NODE_LEN(name_node) - ,NODE_NAME(name_node) - ); - #endif - - return true; - -} - -/*-------------------------------------------------------------------------------- - `#macro` directive RT extension - - Given a pfile, returns a macro definition. - - #macro name (parameter [,parameter] ...) (body_expr) - #macro name () (body_expr) - - Upon entry, the name was already been parsed in directives.cc::do_macro, so the next token will be the opening paren of the parameter list. - - Thi code is similar to `_cpp_create_definition` though uses paren blancing around the body, instead of requiring the macro body be on a single line. - - The cpp_macro struct is defined in cpplib.h: `struct GTY(()) cpp_macro {` it has a flexible array field in a union as a last member: cpp_token tokens[1]; - - This code was derived from create_iso_definition(). The break out portions shared - with create_macro_definition code should be shared with the main code, so that there - is only one place for edits. - -*/ -static cpp_macro *create_rt_macro (cpp_reader *pfile){ - - #if DebugRTMacro - fprintf(stderr,"entering create_rt_macro\n"); - #endif - - unsigned int num_extra_tokens = 0; - unsigned paramc = 0; - cpp_hashnode **params = NULL; - bool varadic = false; - bool ok = false; - cpp_macro *macro = NULL; - parse_clause_status status; - - /* parse parameter list - - after parse_parms runs, the next token returned by pfile will be subsequent to the parameter list, e.g.: - 7 | #macro Q(f ,...) printf(f ,__VA_ARGS__) - | ^~~~~~ - */ - const cpp_token *token = _cpp_lex_token(pfile); - location_t src_loc = token->src_loc; - - if(token->type != CPP_OPEN_PAREN){ - cpp_error_with_line( - pfile - ,CPP_DL_ERROR - ,src_loc - ,0 - ,"expected '(' to open arguments list, but found: %s" - ,cpp_token_as_text(token) - ); - goto out; - } - - if( !parse_params(pfile, ¶mc, &varadic) ) goto out; - - // finalizes the reserved room, otherwise it will be reused on the next reserve room call. - params = (cpp_hashnode **)_cpp_commit_buff( pfile, sizeof (cpp_hashnode *) * paramc ); - token = NULL; - - /* parse body macro - - A macro struct instance is variable size, due to tokens added to the macro.exp.tokens - during parse, and possible reallocations. - - Function like macros will later need space to hold parameter values. - */ - macro = _cpp_new_macro( - pfile - ,cmk_macro - ,_cpp_reserve_room( pfile, 0, sizeof(cpp_macro) ) - ); - // used by parse_clause_literal - macro->variadic = varadic; - macro->paramc = paramc; - macro->parm.params = params; - macro->fun_like = true; - - PCPSO_choice choice; - status = parse_clause_paren_square_option( - pfile - ,macro - ,&choice - ,false // not a commas list - ,&src_loc - ,NULL // don't need to know the terminator - ,&num_extra_tokens - ); - if( status != PCS_COMPLETE ){ - fprintf(stderr, "parse_paren_clause returned: "); - print_parse_clause_status(status); - goto out; - } - - #if DebugRTMacro - fprintf(stderr,"rt_macro directive body tokens:\n"); - print_token_list(macro->exp.tokens ,macro->count); - #endif - - // commit the macro, attach the parameter list - ok = true; - macro = (cpp_macro *)_cpp_commit_buff( - pfile - , - sizeof (cpp_macro) - - sizeof (cpp_token) - + sizeof (cpp_token) * macro->count - + sizeof(cpp_hashnode *) * paramc - ); - macro->variadic = varadic; - macro->paramc = paramc; - macro->parm.params = params; - macro->fun_like = true; - - /* some end cases we must clean up - */ - /* - It might be that the first token of the macro body was preceded by white space, so - the white space flag is set. However, upon expansion, there might not be a white - space before said token, so the following code clears the flag. - */ - if (macro->count) - macro->exp.tokens[0].flags &= ~PREV_WHITE; - - /* - Identifies consecutive ## tokens (a.k.a. CPP_PASTE) that were invalid or ambiguous, - - Removes them from the main macro body, - - Stashes them at the end of the tokens[] array in the same memory, - - Sets macro->extra_tokens = 1 to signal their presence. - */ - if (num_extra_tokens) - { - /* Place second and subsequent ## or %:%: tokens in sequences of - consecutive such tokens at the end of the list to preserve - information about where they appear, how they are spelt and - whether they are preceded by whitespace without otherwise - interfering with macro expansion. Remember, this is - extremely rare, so efficiency is not a priority. */ - cpp_token *temp = (cpp_token *)_cpp_reserve_room - (pfile, 0, num_extra_tokens * sizeof (cpp_token)); - unsigned extra_ix = 0, norm_ix = 0; - cpp_token *exp = macro->exp.tokens; - for (unsigned ix = 0; ix != macro->count; ix++) - if (exp[ix].type == CPP_PASTE) - temp[extra_ix++] = exp[ix]; - else - exp[norm_ix++] = exp[ix]; - memcpy (&exp[norm_ix], temp, num_extra_tokens * sizeof (cpp_token)); - - /* Record there are extra tokens. */ - macro->extra_tokens = 1; - } - - out: - - /* - - This resets a flag in the parser’s state machine, pfile. - - The field `va_args_ok` tracks whether the current macro body is allowed to reference `__VA_ARGS__` (or more precisely, `__VA_OPT__`). - - It's set **while parsing a macro body** that might use variadic logic — particularly in `vaopt_state` tracking. - - Resetting it here ensures that future macros aren't accidentally parsed under the assumption that variadic substitution is valid. - */ - pfile->state.va_args_ok = 0; - - /* - Earlier we did: - if (!parse_params(pfile, ¶mc, &variadic)) goto out; - This cleans up temporary memory used by parse_params. - */ - _cpp_unsave_parameters (pfile, paramc); - - return ok ? macro : NULL; - -} - -/* - called from directives.cc:: do_macro -*/ -bool -_cpp_create_rt_macro(cpp_reader *pfile, cpp_hashnode *node){ - - #if DebugRTMacro - fprintf(stderr,"entering _cpp_create_macro\n"); - #endif - - cpp_macro *macro; - macro = create_rt_macro (pfile); - - if (!macro) - return false; - - if (cpp_macro_p (node)) - { - if (CPP_OPTION (pfile, warn_unused_macros)) - _cpp_warn_if_unused_macro (pfile, node, NULL); - - if (warn_of_redefinition (pfile, node, macro)) - { - const enum cpp_warning_reason reason - = (cpp_builtin_macro_p (node) && !(node->flags & NODE_WARN)) - ? CPP_W_BUILTIN_MACRO_REDEFINED : CPP_W_NONE; - - bool warned = - cpp_pedwarning_with_line (pfile, reason, - pfile->directive_line, 0, - "\"%s\" redefined", NODE_NAME (node)); - - if (warned && cpp_user_macro_p (node)) - cpp_error_with_line (pfile, CPP_DL_NOTE, - node->value.macro->line, 0, - "this is the location of the previous definition"); - } - _cpp_free_definition (node); - } - - /* Enter definition in hash table. */ - node->type = NT_USER_MACRO; - node->value.macro = macro; - if (! ustrncmp (NODE_NAME (node), DSC ("__STDC_")) - && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_FORMAT_MACROS") - /* __STDC_LIMIT_MACROS and __STDC_CONSTANT_MACROS are mentioned - in the C standard, as something that one must use in C++. - However DR#593 and C++11 indicate that they play no role in C++. - We special-case them anyway. */ - && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_LIMIT_MACROS") - && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_CONSTANT_MACROS")) - node->flags |= NODE_WARN; - - /* If user defines one of the conditional macros, remove the - conditional flag */ - node->flags &= ~NODE_CONDITIONAL; - - return true; -} - -/*-------------------------------------------------------------------------------- - builtin RT_CAT macro RT extension - -*/ - -static const uchar *evaluate_RT_CAT(cpp_reader *pfile){ - return UC"XY"; -} diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/README.org" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/README.org" deleted file mode 100644 index 4456c81..0000000 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/README.org" +++ /dev/null @@ -1,26 +0,0 @@ -** directives.cc - -Midway there is a table that lists the directives. - -I put the code for `#assign` and `#macro` at the bottom under the section: "RT Extensions". - - -** macros.cc - -The code in the directives file points into macros to do the work. I put that code for `#assign` and `#macro` at the bottom under the section: "RT Extensions". - - -** include/cpplib.h - -Added entry to `enum cpp_builtin_type` for `BT_CAT` - - -** init.cc - -Has, `static const struct builtin_macro builtin_array[] =`, I added a line for - - B("__CAT", BT_CAT, true), - -Strange the original for this file has a comma after the last entry, so I did the same. - -The `true` means that there will be a warning if __CAT is redefined. diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/include/#cpplib.h#" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/include/#cpplib.h#" new file mode 100644 index 0000000..aea752f --- /dev/null +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/include/#cpplib.h#" @@ -0,0 +1,1585 @@ +/* Definitions for CPP library. + Copyright (C) 1995-2022 Free Software Foundation, Inc. + Written by Per Bothner, 1994-95. + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 3, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING3. If not see +. + + In other words, you are welcome to use, share and improve this program. + You are forbidden to forbid anyone else to use, share and improve + what you give them. Help stamp out software-hoarding! */ +#ifndef LIBCPP_CPPLIB_H +#define LIBCPP_CPPLIB_H + +#include +#include "symtab.h" +#include "line-map.h" + +typedef struct cpp_reader cpp_reader; +typedef struct cpp_buffer cpp_buffer; +typedef struct cpp_options cpp_options; +typedef struct cpp_token cpp_token; +typedef struct cpp_string cpp_string; +typedef struct cpp_hashnode cpp_hashnode; +typedef struct cpp_macro cpp_macro; +typedef struct cpp_callbacks cpp_callbacks; +typedef struct cpp_dir cpp_dir; + +struct _cpp_file; + +/* The first three groups, apart from '=', can appear in preprocessor + expressions (+= and -= are used to indicate unary + and - resp.). + This allows a lookup table to be implemented in _cpp_parse_expr. + + The first group, to CPP_LAST_EQ, can be immediately followed by an + '='. The lexer needs operators ending in '=', like ">>=", to be in + the same order as their counterparts without the '=', like ">>". + + See the cpp_operator table optab in expr.cc if you change the order or + add or remove anything in the first group. */ + +#define TTYPE_TABLE \ + OP(EQ, "=") \ + OP(NOT, "!") \ + OP(GREATER, ">") /* compare */ \ + OP(LESS, "<") \ + OP(PLUS, "+") /* math */ \ + OP(MINUS, "-") \ + OP(MULT, "*") \ + OP(DIV, "/") \ + OP(MOD, "%") \ + OP(AND, "&") /* bit ops */ \ + OP(OR, "|") \ + OP(XOR, "^") \ + OP(RSHIFT, ">>") \ + OP(LSHIFT, "<<") \ + \ + OP(COMPL, "~") \ + OP(AND_AND, "&&") /* logical */ \ + OP(OR_OR, "||") \ + OP(QUERY, "?") \ + OP(COLON, ":") \ + OP(COMMA, ",") /* grouping */ \ + OP(OPEN_PAREN, "(") \ + OP(CLOSE_PAREN, ")") \ + TK(EOF, NONE) \ + OP(EQ_EQ, "==") /* compare */ \ + OP(NOT_EQ, "!=") \ + OP(GREATER_EQ, ">=") \ + OP(LESS_EQ, "<=") \ + OP(SPACESHIP, "<=>") \ + \ + /* These two are unary + / - in preprocessor expressions. */ \ + OP(PLUS_EQ, "+=") /* math */ \ + OP(MINUS_EQ, "-=") \ + \ + OP(MULT_EQ, "*=") \ + OP(DIV_EQ, "/=") \ + OP(MOD_EQ, "%=") \ + OP(AND_EQ, "&=") /* bit ops */ \ + OP(OR_EQ, "|=") \ + OP(XOR_EQ, "^=") \ + OP(RSHIFT_EQ, ">>=") \ + OP(LSHIFT_EQ, "<<=") \ + /* Digraphs together, beginning with CPP_FIRST_DIGRAPH. */ \ + OP(HASH, "#") /* digraphs */ \ + OP(PASTE, "##") \ + OP(OPEN_SQUARE, "[") \ + OP(CLOSE_SQUARE, "]") \ + OP(OPEN_BRACE, "{") \ + OP(CLOSE_BRACE, "}") \ + /* The remainder of the punctuation. Order is not significant. */ \ + OP(SEMICOLON, ";") /* structure */ \ + OP(ELLIPSIS, "...") \ + OP(PLUS_PLUS, "++") /* increment */ \ + OP(MINUS_MINUS, "--") \ + OP(DEREF, "->") /* accessors */ \ + OP(DOT, ".") \ + OP(SCOPE, "::") \ + OP(DEREF_STAR, "->*") \ + OP(DOT_STAR, ".*") \ + OP(ATSIGN, "@") /* used in Objective-C */ \ + \ + TK(NAME, IDENT) /* word */ \ + TK(AT_NAME, IDENT) /* @word - Objective-C */ \ + TK(NUMBER, LITERAL) /* 34_be+ta */ \ + \ + TK(CHAR, LITERAL) /* 'char' */ \ + TK(WCHAR, LITERAL) /* L'char' */ \ + TK(CHAR16, LITERAL) /* u'char' */ \ + TK(CHAR32, LITERAL) /* U'char' */ \ + TK(UTF8CHAR, LITERAL) /* u8'char' */ \ + TK(OTHER, LITERAL) /* stray punctuation */ \ + \ + TK(STRING, LITERAL) /* "string" */ \ + TK(WSTRING, LITERAL) /* L"string" */ \ + TK(STRING16, LITERAL) /* u"string" */ \ + TK(STRING32, LITERAL) /* U"string" */ \ + TK(UTF8STRING, LITERAL) /* u8"string" */ \ + TK(OBJC_STRING, LITERAL) /* @"string" - Objective-C */ \ + TK(HEADER_NAME, LITERAL) /* in #include */ \ + \ + TK(CHAR_USERDEF, LITERAL) /* 'char'_suffix - C++-0x */ \ + TK(WCHAR_USERDEF, LITERAL) /* L'char'_suffix - C++-0x */ \ + TK(CHAR16_USERDEF, LITERAL) /* u'char'_suffix - C++-0x */ \ + TK(CHAR32_USERDEF, LITERAL) /* U'char'_suffix - C++-0x */ \ + TK(UTF8CHAR_USERDEF, LITERAL) /* u8'char'_suffix - C++-0x */ \ + TK(STRING_USERDEF, LITERAL) /* "string"_suffix - C++-0x */ \ + TK(WSTRING_USERDEF, LITERAL) /* L"string"_suffix - C++-0x */ \ + TK(STRING16_USERDEF, LITERAL) /* u"string"_suffix - C++-0x */ \ + TK(STRING32_USERDEF, LITERAL) /* U"string"_suffix - C++-0x */ \ + TK(UTF8STRING_USERDEF,LITERAL) /* u8"string"_suffix - C++-0x */ \ + \ + TK(COMMENT, LITERAL) /* Only if output comments. */ \ + /* SPELL_LITERAL happens to DTRT. */ \ + TK(MACRO_ARG, NONE) /* Macro argument. */ \ + TK(PRAGMA, NONE) /* Only for deferred pragmas. */ \ + TK(PRAGMA_EOL, NONE) /* End-of-line for deferred pragmas. */ \ + TK(PADDING, NONE) /* Whitespace for -E. */ + +#define OP(e, s) CPP_ ## e, +#define TK(e, s) CPP_ ## e, +enum cpp_ttype +{ + TTYPE_TABLE + N_TTYPES, + + /* A token type for keywords, as opposed to ordinary identifiers. */ + CPP_KEYWORD, + + /* Positions in the table. */ + CPP_LAST_EQ = CPP_LSHIFT, + CPP_FIRST_DIGRAPH = CPP_HASH, + CPP_LAST_PUNCTUATOR= CPP_ATSIGN, + CPP_LAST_CPP_OP = CPP_LESS_EQ +}; +#undef OP +#undef TK + +/* C language kind, used when calling cpp_create_reader. */ +enum c_lang {CLK_GNUC89 = 0, CLK_GNUC99, CLK_GNUC11, CLK_GNUC17, CLK_GNUC2X, + CLK_STDC89, CLK_STDC94, CLK_STDC99, CLK_STDC11, CLK_STDC17, + CLK_STDC2X, + CLK_GNUCXX, CLK_CXX98, CLK_GNUCXX11, CLK_CXX11, + CLK_GNUCXX14, CLK_CXX14, CLK_GNUCXX17, CLK_CXX17, + CLK_GNUCXX20, CLK_CXX20, CLK_GNUCXX23, CLK_CXX23, + CLK_ASM}; + +/* Payload of a NUMBER, STRING, CHAR or COMMENT token. */ +struct GTY(()) cpp_string { + unsigned int len; + const unsigned char *text; +}; + +/* Flags for the cpp_token structure. */ +#define PREV_WHITE (1 << 0) /* If whitespace before this token. */ +#define DIGRAPH (1 << 1) /* If it was a digraph. */ +#define STRINGIFY_ARG (1 << 2) /* If macro argument to be stringified. */ +#define PASTE_LEFT (1 << 3) /* If on LHS of a ## operator. */ +#define NAMED_OP (1 << 4) /* C++ named operators. */ +#define PREV_FALLTHROUGH (1 << 5) /* On a token preceeded by FALLTHROUGH + comment. */ +#define BOL (1 << 6) /* Token at beginning of line. */ +#define PURE_ZERO (1 << 7) /* Single 0 digit, used by the C++ frontend, + set in c-lex.cc. */ +#define COLON_SCOPE PURE_ZERO /* Adjacent colons in C < 23. */ +#define SP_DIGRAPH (1 << 8) /* # or ## token was a digraph. */ +#define SP_PREV_WHITE (1 << 9) /* If whitespace before a ## + operator, or before this token + after a # operator. */ +#define NO_EXPAND (1 << 10) /* Do not macro-expand this token. */ +#define PRAGMA_OP (1 << 11) /* _Pragma token. */ + +/* Specify which field, if any, of the cpp_token union is used. */ + +enum cpp_token_fld_kind { + CPP_TOKEN_FLD_NODE, + CPP_TOKEN_FLD_SOURCE, + CPP_TOKEN_FLD_STR, + CPP_TOKEN_FLD_ARG_NO, + CPP_TOKEN_FLD_TOKEN_NO, + CPP_TOKEN_FLD_PRAGMA, + CPP_TOKEN_FLD_NONE +}; + +/* A macro argument in the cpp_token union. */ +struct GTY(()) cpp_macro_arg { + /* Argument number. */ + unsigned int arg_no; + /* The original spelling of the macro argument token. */ + cpp_hashnode * + GTY ((nested_ptr (union tree_node, + "%h ? CPP_HASHNODE (GCC_IDENT_TO_HT_IDENT (%h)) : NULL", + "%h ? HT_IDENT_TO_GCC_IDENT (HT_NODE (%h)) : NULL"))) + spelling; +}; + +/* An identifier in the cpp_token union. */ +struct GTY(()) cpp_identifier { + /* The canonical (UTF-8) spelling of the identifier. */ + cpp_hashnode * + GTY ((nested_ptr (union tree_node, + "%h ? CPP_HASHNODE (GCC_IDENT_TO_HT_IDENT (%h)) : NULL", + "%h ? HT_IDENT_TO_GCC_IDENT (HT_NODE (%h)) : NULL"))) + node; + /* The original spelling of the identifier. */ + cpp_hashnode * + GTY ((nested_ptr (union tree_node, + "%h ? CPP_HASHNODE (GCC_IDENT_TO_HT_IDENT (%h)) : NULL", + "%h ? HT_IDENT_TO_GCC_IDENT (HT_NODE (%h)) : NULL"))) + spelling; +}; + +/* A preprocessing token. This has been carefully packed and should + occupy 16 bytes on 32-bit hosts and 24 bytes on 64-bit hosts. */ +struct GTY(()) cpp_token { + + /* Location of first char of token, together with range of full token. */ + location_t src_loc; + + ENUM_BITFIELD(cpp_ttype) type : CHAR_BIT; /* token type */ + unsigned short flags; /* flags - see above */ + + union cpp_token_u + { + /* An identifier. */ + struct cpp_identifier GTY ((tag ("CPP_TOKEN_FLD_NODE"))) node; + + /* Inherit padding from this token. */ + cpp_token * GTY ((tag ("CPP_TOKEN_FLD_SOURCE"))) source; + + /* A string, or number. */ + struct cpp_string GTY ((tag ("CPP_TOKEN_FLD_STR"))) str; + + /* Argument no. (and original spelling) for a CPP_MACRO_ARG. */ + struct cpp_macro_arg GTY ((tag ("CPP_TOKEN_FLD_ARG_NO"))) macro_arg; + + /* Original token no. for a CPP_PASTE (from a sequence of + consecutive paste tokens in a macro expansion). */ + unsigned int GTY ((tag ("CPP_TOKEN_FLD_TOKEN_NO"))) token_no; + + /* Caller-supplied identifier for a CPP_PRAGMA. */ + unsigned int GTY ((tag ("CPP_TOKEN_FLD_PRAGMA"))) pragma; + } GTY ((desc ("cpp_token_val_index (&%1)"))) val; +}; + +/* Say which field is in use. */ +extern enum cpp_token_fld_kind cpp_token_val_index (const cpp_token *tok); + +/* A type wide enough to hold any multibyte source character. + cpplib's character constant interpreter requires an unsigned type. + Also, a typedef for the signed equivalent. + The width of this type is capped at 32 bits; there do exist targets + where wchar_t is 64 bits, but only in a non-default mode, and there + would be no meaningful interpretation for a wchar_t value greater + than 2^32 anyway -- the widest wide-character encoding around is + ISO 10646, which stops at 2^31. */ +#if CHAR_BIT * SIZEOF_INT >= 32 +# define CPPCHAR_SIGNED_T int +#elif CHAR_BIT * SIZEOF_LONG >= 32 +# define CPPCHAR_SIGNED_T long +#else +# error "Cannot find a least-32-bit signed integer type" +#endif +typedef unsigned CPPCHAR_SIGNED_T cppchar_t; +typedef CPPCHAR_SIGNED_T cppchar_signed_t; + +/* Style of header dependencies to generate. */ +enum cpp_deps_style { DEPS_NONE = 0, DEPS_USER, DEPS_SYSTEM }; + +/* The possible normalization levels, from most restrictive to least. */ +enum cpp_normalize_level { + /* In NFKC. */ + normalized_KC = 0, + /* In NFC. */ + normalized_C, + /* In NFC, except for subsequences where being in NFC would make + the identifier invalid. */ + normalized_identifier_C, + /* Not normalized at all. */ + normalized_none +}; + +enum cpp_main_search +{ + CMS_none, /* A regular source file. */ + CMS_header, /* Is a directly-specified header file (eg PCH or + header-unit). */ + CMS_user, /* Search the user INCLUDE path. */ + CMS_system, /* Search the system INCLUDE path. */ +}; + +/* The possible bidirectional control characters checking levels. */ +enum cpp_bidirectional_level { + /* No checking. */ + bidirectional_none = 0, + /* Only detect unpaired uses of bidirectional control characters. */ + bidirectional_unpaired = 1, + /* Detect any use of bidirectional control characters. */ + bidirectional_any = 2, + /* Also warn about UCNs. */ + bidirectional_ucn = 4 +}; + +/* This structure is nested inside struct cpp_reader, and + carries all the options visible to the command line. */ +struct cpp_options +{ + /* The language we're preprocessing. */ + enum c_lang lang; + + /* Nonzero means use extra default include directories for C++. */ + unsigned char cplusplus; + + /* Nonzero means handle cplusplus style comments. */ + unsigned char cplusplus_comments; + + /* Nonzero means define __OBJC__, treat @ as a special token, use + the OBJC[PLUS]_INCLUDE_PATH environment variable, and allow + "#import". */ + unsigned char objc; + + /* Nonzero means don't copy comments into the output file. */ + unsigned char discard_comments; + + /* Nonzero means don't copy comments into the output file during + macro expansion. */ + unsigned char discard_comments_in_macro_exp; + + /* Nonzero means process the ISO trigraph sequences. */ + unsigned char trigraphs; + + /* Nonzero means process the ISO digraph sequences. */ + unsigned char digraphs; + + /* Nonzero means to allow hexadecimal floats and LL suffixes. */ + unsigned char extended_numbers; + + /* Nonzero means process u/U prefix literals (UTF-16/32). */ + unsigned char uliterals; + + /* Nonzero means process u8 prefixed character literals (UTF-8). */ + unsigned char utf8_char_literals; + + /* Nonzero means process r/R raw strings. If this is set, uliterals + must be set as well. */ + unsigned char rliterals; + + /* Nonzero means print names of header files (-H). */ + unsigned char print_include_names; + + /* Nonzero means complain about deprecated features. */ + unsigned char cpp_warn_deprecated; + + /* Nonzero means warn if slash-star appears in a comment. */ + unsigned char warn_comments; + + /* Nonzero means to warn about __DATA__, __TIME__ and __TIMESTAMP__ usage. */ + unsigned char warn_date_time; + + /* Nonzero means warn if a user-supplied include directory does not + exist. */ + unsigned char warn_missing_include_dirs; + + /* Nonzero means warn if there are any trigraphs. */ + unsigned char warn_trigraphs; + + /* Nonzero means warn about multicharacter charconsts. */ + unsigned char warn_multichar; + + /* Nonzero means warn about various incompatibilities with + traditional C. */ + unsigned char cpp_warn_traditional; + + /* Nonzero means warn about long long numeric constants. */ + unsigned char cpp_warn_long_long; + + /* Nonzero means warn about text after an #endif (or #else). */ + unsigned char warn_endif_labels; + + /* Nonzero means warn about implicit sign changes owing to integer + promotions. */ + unsigned char warn_num_sign_change; + + /* Zero means don't warn about __VA_ARGS__ usage in c89 pedantic mode. + Presumably the usage is protected by the appropriate #ifdef. */ + unsigned char warn_variadic_macros; + + /* Nonzero means warn about builtin macros that are redefined or + explicitly undefined. */ + unsigned char warn_builtin_macro_redefined; + + /* Different -Wimplicit-fallthrough= levels. */ + unsigned char cpp_warn_implicit_fallthrough; + + /* Nonzero means we should look for header.gcc files that remap file + names. */ + unsigned char remap; + + /* Zero means dollar signs are punctuation. */ + unsigned char dollars_in_ident; + + /* Nonzero means UCNs are accepted in identifiers. */ + unsigned char extended_identifiers; + + /* True if we should warn about dollars in identifiers or numbers + for this translation unit. */ + unsigned char warn_dollars; + + /* Nonzero means warn if undefined identifiers are evaluated in an #if. */ + unsigned char warn_undef; + + /* Nonzero means warn if "defined" is encountered in a place other than + an #if. */ + unsigned char warn_expansion_to_defined; + + /* Nonzero means warn of unused macros from the main file. */ + unsigned char warn_unused_macros; + + /* Nonzero for the 1999 C Standard, including corrigenda and amendments. */ + unsigned char c99; + + /* Nonzero if we are conforming to a specific C or C++ standard. */ + unsigned char std; + + /* Nonzero means give all the error messages the ANSI standard requires. */ + unsigned char cpp_pedantic; + + /* Nonzero means we're looking at already preprocessed code, so don't + bother trying to do macro expansion and whatnot. */ + unsigned char preprocessed; + + /* Nonzero means we are going to emit debugging logs during + preprocessing. */ + unsigned char debug; + + /* Nonzero means we are tracking locations of tokens involved in + macro expansion. 1 Means we track the location in degraded mode + where we do not track locations of tokens resulting from the + expansion of arguments of function-like macro. 2 Means we do + track all macro expansions. This last option is the one that + consumes the highest amount of memory. */ + unsigned char track_macro_expansion; + + /* Nonzero means handle C++ alternate operator names. */ + unsigned char operator_names; + + /* Nonzero means warn about use of C++ alternate operator names. */ + unsigned char warn_cxx_operator_names; + + /* True for traditional preprocessing. */ + unsigned char traditional; + + /* Nonzero for C++ 2011 Standard user-defined literals. */ + unsigned char user_literals; + + /* Nonzero means warn when a string or character literal is followed by a + ud-suffix which does not beging with an underscore. */ + unsigned char warn_literal_suffix; + + /* Nonzero means interpret imaginary, fixed-point, or other gnu extension + literal number suffixes as user-defined literal number suffixes. */ + unsigned char ext_numeric_literals; + + /* Nonzero means extended identifiers allow the characters specified + in C11. */ + unsigned char c11_identifiers; + + /* Nonzero for C++ 2014 Standard binary constants. */ + unsigned char binary_constants; + + /* Nonzero for C++ 2014 Standard digit separators. */ + unsigned char digit_separators; + + /* Nonzero for C2X decimal floating-point constants. */ + unsigned char dfp_constants; + + /* Nonzero for C++20 __VA_OPT__ feature. */ + unsigned char va_opt; + + /* Nonzero for the '::' token. */ + unsigned char scope; + + /* Nonzero for the '#elifdef' and '#elifndef' directives. */ + unsigned char elifdef; + + /* Nonzero means tokenize C++20 module directives. */ + unsigned char module_directives; + + /* Nonzero for C++23 size_t literals. */ + unsigned char size_t_literals; + + /* Holds the name of the target (execution) character set. */ + const char *narrow_charset; + + /* Holds the name of the target wide character set. */ + const char *wide_charset; + + /* Holds the name of the input character set. */ + const char *input_charset; + + /* The minimum permitted level of normalization before a warning + is generated. See enum cpp_normalize_level. */ + int warn_normalize; + + /* True to warn about precompiled header files we couldn't use. */ + bool warn_invalid_pch; + + /* True if dependencies should be restored from a precompiled header. */ + bool restore_pch_deps; + + /* True if warn about differences between C90 and C99. */ + signed char cpp_warn_c90_c99_compat; + + /* True if warn about differences between C11 and C2X. */ + signed char cpp_warn_c11_c2x_compat; + + /* True if warn about differences between C++98 and C++11. */ + bool cpp_warn_cxx11_compat; + + /* Nonzero if bidirectional control characters checking is on. See enum + cpp_bidirectional_level. */ + unsigned char cpp_warn_bidirectional; + + /* Dependency generation. */ + struct + { + /* Style of header dependencies to generate. */ + enum cpp_deps_style style; + + /* Assume missing files are generated files. */ + bool missing_files; + + /* Generate phony targets for each dependency apart from the first + one. */ + bool phony_targets; + + /* Generate dependency info for modules. */ + bool modules; + + /* If true, no dependency is generated on the main file. */ + bool ignore_main_file; + + /* If true, intend to use the preprocessor output (e.g., for compilation) + in addition to the dependency info. */ + bool need_preprocessor_output; + } deps; + + /* Target-specific features set by the front end or client. */ + + /* Precision for target CPP arithmetic, target characters, target + ints and target wide characters, respectively. */ + size_t precision, char_precision, int_precision, wchar_precision; + + /* True means chars (wide chars) are unsigned. */ + bool unsigned_char, unsigned_wchar; + + /* True if the most significant byte in a word has the lowest + address in memory. */ + bool bytes_big_endian; + + /* Nonzero means __STDC__ should have the value 0 in system headers. */ + unsigned char stdc_0_in_system_headers; + + /* True disables tokenization outside of preprocessing directives. */ + bool directives_only; + + /* True enables canonicalization of system header file paths. */ + bool canonical_system_headers; + + /* The maximum depth of the nested #include. */ + unsigned int max_include_depth; + + cpp_main_search main_search : 8; +}; + +/* Diagnostic levels. To get a diagnostic without associating a + position in the translation unit with it, use cpp_error_with_line + with a line number of zero. */ + +enum cpp_diagnostic_level { + /* Warning, an error with -Werror. */ + CPP_DL_WARNING = 0, + /* Same as CPP_DL_WARNING, except it is not suppressed in system headers. */ + CPP_DL_WARNING_SYSHDR, + /* Warning, an error with -pedantic-errors or -Werror. */ + CPP_DL_PEDWARN, + /* An error. */ + CPP_DL_ERROR, + /* An internal consistency check failed. Prints "internal error: ", + otherwise the same as CPP_DL_ERROR. */ + CPP_DL_ICE, + /* An informative note following a warning. */ + CPP_DL_NOTE, + /* A fatal error. */ + CPP_DL_FATAL +}; + +/* Warning reason codes. Use a reason code of CPP_W_NONE for unclassified + warnings and diagnostics that are not warnings. */ + +enum cpp_warning_reason { + CPP_W_NONE = 0, + CPP_W_DEPRECATED, + CPP_W_COMMENTS, + CPP_W_MISSING_INCLUDE_DIRS, + CPP_W_TRIGRAPHS, + CPP_W_MULTICHAR, + CPP_W_TRADITIONAL, + CPP_W_LONG_LONG, + CPP_W_ENDIF_LABELS, + CPP_W_NUM_SIGN_CHANGE, + CPP_W_VARIADIC_MACROS, + CPP_W_BUILTIN_MACRO_REDEFINED, + CPP_W_DOLLARS, + CPP_W_UNDEF, + CPP_W_UNUSED_MACROS, + CPP_W_CXX_OPERATOR_NAMES, + CPP_W_NORMALIZE, + CPP_W_INVALID_PCH, + CPP_W_WARNING_DIRECTIVE, + CPP_W_LITERAL_SUFFIX, + CPP_W_SIZE_T_LITERALS, + CPP_W_DATE_TIME, + CPP_W_PEDANTIC, + CPP_W_C90_C99_COMPAT, + CPP_W_C11_C2X_COMPAT, + CPP_W_CXX11_COMPAT, + CPP_W_EXPANSION_TO_DEFINED, + CPP_W_BIDIRECTIONAL +}; + +/* Callback for header lookup for HEADER, which is the name of a + source file. It is used as a method of last resort to find headers + that are not otherwise found during the normal include processing. + The return value is the malloced name of a header to try and open, + if any, or NULL otherwise. This callback is called only if the + header is otherwise unfound. */ +typedef const char *(*missing_header_cb)(cpp_reader *, const char *header, cpp_dir **); + +/* Call backs to cpplib client. */ +struct cpp_callbacks +{ + /* Called when a new line of preprocessed output is started. */ + void (*line_change) (cpp_reader *, const cpp_token *, int); + + /* Called when switching to/from a new file. + The line_map is for the new file. It is NULL if there is no new file. + (In C this happens when done with + and also + when done with a main file.) This can be used for resource cleanup. */ + void (*file_change) (cpp_reader *, const line_map_ordinary *); + + void (*dir_change) (cpp_reader *, const char *); + void (*include) (cpp_reader *, location_t, const unsigned char *, + const char *, int, const cpp_token **); + void (*define) (cpp_reader *, location_t, cpp_hashnode *); + void (*undef) (cpp_reader *, location_t, cpp_hashnode *); + void (*ident) (cpp_reader *, location_t, const cpp_string *); + void (*def_pragma) (cpp_reader *, location_t); + int (*valid_pch) (cpp_reader *, const char *, int); + void (*read_pch) (cpp_reader *, const char *, int, const char *); + missing_header_cb missing_header; + + /* Context-sensitive macro support. Returns macro (if any) that should + be expanded. */ + cpp_hashnode * (*macro_to_expand) (cpp_reader *, const cpp_token *); + + /* Called to emit a diagnostic. This callback receives the + translated message. */ + bool (*diagnostic) (cpp_reader *, + enum cpp_diagnostic_level, + enum cpp_warning_reason, + rich_location *, + const char *, va_list *) + ATTRIBUTE_FPTR_PRINTF(5,0); + + /* Callbacks for when a macro is expanded, or tested (whether + defined or not at the time) in #ifdef, #ifndef or "defined". */ + void (*used_define) (cpp_reader *, location_t, cpp_hashnode *); + void (*used_undef) (cpp_reader *, location_t, cpp_hashnode *); + /* Called before #define and #undef or other macro definition + changes are processed. */ + void (*before_define) (cpp_reader *); + /* Called whenever a macro is expanded or tested. + Second argument is the location of the start of the current expansion. */ + void (*used) (cpp_reader *, location_t, cpp_hashnode *); + + /* Callback to identify whether an attribute exists. */ + int (*has_attribute) (cpp_reader *, bool); + + /* Callback to determine whether a built-in function is recognized. */ + int (*has_builtin) (cpp_reader *); + + /* Callback that can change a user lazy into normal macro. */ + void (*user_lazy_macro) (cpp_reader *, cpp_macro *, unsigned); + + /* Callback to handle deferred cpp_macros. */ + cpp_macro *(*user_deferred_macro) (cpp_reader *, location_t, cpp_hashnode *); + + /* Callback to parse SOURCE_DATE_EPOCH from environment. */ + time_t (*get_source_date_epoch) (cpp_reader *); + + /* Callback for providing suggestions for misspelled directives. */ + const char *(*get_suggestion) (cpp_reader *, const char *, const char *const *); + + /* Callback for when a comment is encountered, giving the location + of the opening slash, a pointer to the content (which is not + necessarily 0-terminated), and the length of the content. + The content contains the opening slash-star (or slash-slash), + and for C-style comments contains the closing star-slash. For + C++-style comments it does not include the terminating newline. */ + void (*comment) (cpp_reader *, location_t, const unsigned char *, + size_t); + + /* Callback for filename remapping in __FILE__ and __BASE_FILE__ macro + expansions. */ + const char *(*remap_filename) (const char*); + + /* Maybe translate a #include into something else. Return a + cpp_buffer containing the translation if translating. */ + char *(*translate_include) (cpp_reader *, line_maps *, location_t, + const char *path); +}; + +#ifdef VMS +#define INO_T_CPP ino_t ino[3] +#elif defined (_AIX) && SIZEOF_INO_T == 4 +#define INO_T_CPP ino64_t ino +#else +#define INO_T_CPP ino_t ino +#endif + +#if defined (_AIX) && SIZEOF_DEV_T == 4 +#define DEV_T_CPP dev64_t dev +#else +#define DEV_T_CPP dev_t dev +#endif + +/* Chain of directories to look for include files in. */ +struct cpp_dir +{ + /* NULL-terminated singly-linked list. */ + struct cpp_dir *next; + + /* NAME of the directory, NUL-terminated. */ + char *name; + unsigned int len; + + /* One if a system header, two if a system header that has extern + "C" guards for C++. */ + unsigned char sysp; + + /* Is this a user-supplied directory? */ + bool user_supplied_p; + + /* The canonicalized NAME as determined by lrealpath. This field + is only used by hosts that lack reliable inode numbers. */ + char *canonical_name; + + /* Mapping of file names for this directory for MS-DOS and related + platforms. A NULL-terminated array of (from, to) pairs. */ + const char **name_map; + + /* Routine to construct pathname, given the search path name and the + HEADER we are trying to find, return a constructed pathname to + try and open. If this is NULL, the constructed pathname is as + constructed by append_file_to_dir. */ + char *(*construct) (const char *header, cpp_dir *dir); + + /* The C front end uses these to recognize duplicated + directories in the search path. */ + INO_T_CPP; + DEV_T_CPP; +}; + +/* The kind of the cpp_macro. */ +enum cpp_macro_kind { + cmk_macro, /* An ISO macro (token expansion). */ + cmk_assert, /* An assertion. */ + cmk_traditional /* A traditional macro (text expansion). */ +}; + +/* Each macro definition is recorded in a cpp_macro structure. + Variadic macros cannot occur with traditional cpp. */ +struct GTY(()) cpp_macro { + union cpp_parm_u + { + /* Parameters, if any. If parameter names use extended identifiers, + the original spelling of those identifiers, not the canonical + UTF-8 spelling, goes here. */ + cpp_hashnode ** GTY ((tag ("false"), + nested_ptr (union tree_node, + "%h ? CPP_HASHNODE (GCC_IDENT_TO_HT_IDENT (%h)) : NULL", + "%h ? HT_IDENT_TO_GCC_IDENT (HT_NODE (%h)) : NULL"), + length ("%1.paramc"))) params; + + /* If this is an assertion, the next one in the chain. */ + cpp_macro *GTY ((tag ("true"))) next; + } GTY ((desc ("%1.kind == cmk_assert"))) parm; + + /* Definition line number. */ + location_t line; + + /* Number of tokens in body, or bytes for traditional macros. */ + /* Do we really need 2^32-1 range here? */ + unsigned int count; + + /* Number of parameters. */ + unsigned short paramc; + + /* Non-zero if this is a user-lazy macro, value provided by user. */ + unsigned char lazy; + + /* The kind of this macro (ISO, trad or assert) */ + unsigned kind : 2; + + /* If a function-like macro. */ + unsigned int fun_like : 1; + + /* If a variadic macro. */ + unsigned int variadic : 1; + + /* If macro defined in system header. */ + unsigned int syshdr : 1; + + /* Nonzero if it has been expanded or had its existence tested. */ + unsigned int used : 1; + + /* Indicate whether the tokens include extra CPP_PASTE tokens at the + end to track invalid redefinitions with consecutive CPP_PASTE + tokens. */ + unsigned int extra_tokens : 1; + + /* Imported C++20 macro (from a header unit). */ + unsigned int imported_p : 1; + + /* 0 bits spare (32-bit). 32 on 64-bit target. */ + + union cpp_exp_u + { + /* Trailing array of replacement tokens (ISO), or assertion body value. */ + cpp_token GTY ((tag ("false"), length ("%1.count"))) tokens[1]; + + /* Pointer to replacement text (traditional). See comment at top + of cpptrad.c for how traditional function-like macros are + encoded. */ + const unsigned char *GTY ((tag ("true"))) text; + } GTY ((desc ("%1.kind == cmk_traditional"))) exp; +}; + +/* Poisoned identifiers are flagged NODE_POISONED. NODE_OPERATOR (C++ + only) indicates an identifier that behaves like an operator such as + "xor". NODE_DIAGNOSTIC is for speed in lex_token: it indicates a + diagnostic may be required for this node. Currently this only + applies to __VA_ARGS__, poisoned identifiers, and -Wc++-compat + warnings about NODE_OPERATOR. */ + +/* Hash node flags. */ +#define NODE_OPERATOR (1 << 0) /* C++ named operator. */ +#define NODE_POISONED (1 << 1) /* Poisoned identifier. */ +#define NODE_DIAGNOSTIC (1 << 2) /* Possible diagnostic when lexed. */ +#define NODE_WARN (1 << 3) /* Warn if redefined or undefined. */ +#define NODE_DISABLED (1 << 4) /* A disabled macro. */ +#define NODE_USED (1 << 5) /* Dumped with -dU. */ +#define NODE_CONDITIONAL (1 << 6) /* Conditional macro */ +#define NODE_WARN_OPERATOR (1 << 7) /* Warn about C++ named operator. */ +#define NODE_MODULE (1 << 8) /* C++-20 module-related name. */ + +/* Different flavors of hash node. */ +enum node_type +{ + NT_VOID = 0, /* Maybe an assert? */ + NT_MACRO_ARG, /* A macro arg. */ + NT_USER_MACRO, /* A user macro. */ + NT_BUILTIN_MACRO, /* A builtin macro. */ + NT_MACRO_MASK = NT_USER_MACRO /* Mask for either macro kind. */ +}; + +/* Different flavors of builtin macro. _Pragma is an operator, but we + handle it with the builtin code for efficiency reasons. */ +enum cpp_builtin_type +{ + BT_SPECLINE = 0, /* `__LINE__' */ + BT_DATE, /* `__DATE__' */ + BT_FILE, /* `__FILE__' */ + BT_FILE_NAME, /* `__FILE_NAME__' */ + BT_BASE_FILE, /* `__BASE_FILE__' */ + BT_INCLUDE_LEVEL, /* `__INCLUDE_LEVEL__' */ + BT_TIME, /* `__TIME__' */ + BT_STDC, /* `__STDC__' */ + BT_PRAGMA, /* `_Pragma' operator */ + BT_TIMESTAMP, /* `__TIMESTAMP__' */ + BT_COUNTER, /* `__COUNTER__' */ + BT_HAS_ATTRIBUTE, /* `__has_attribute(x)' */ + BT_HAS_STD_ATTRIBUTE, /* `__has_c_attribute(x)' */ + BT_HAS_BUILTIN, /* `__has_builtin(x)' */ + BT_HAS_INCLUDE, /* `__has_include(x)' */ + BT_HAS_INCLUDE_NEXT, /* `__has_include_next(x)' */ + + // RT Extension + BT_RT_ASSIGN, + BT_RT_TO_ARG_LIST, + BT_RT_TO_TOKEN_LIST, + BT_RT_FIRST, + BT_RT_REST, + BT_RT_MAP, + BT_RT_AL_MAP, + BT_RT_IF, + BT_RT_NOT, + BT_RT_AND, + BT_RT_OR, + BT_RT_IS_IDENTIFIER, + BT_RT_IS_NAME, + BT_RT_PASTE +}; + +#define CPP_HASHNODE(HNODE) ((cpp_hashnode *) (HNODE)) +#define HT_NODE(NODE) (&(NODE)->ident) +#define NODE_LEN(NODE) HT_LEN (HT_NODE (NODE)) +#define NODE_NAME(NODE) HT_STR (HT_NODE (NODE)) + +/* The common part of an identifier node shared amongst all 3 C front + ends. Also used to store CPP identifiers, which are a superset of + identifiers in the grammatical sense. */ + +union GTY(()) _cpp_hashnode_value { + /* Assert (maybe NULL) */ + cpp_macro * GTY((tag ("NT_VOID"))) answers; + /* Macro (maybe NULL) */ + cpp_macro * GTY((tag ("NT_USER_MACRO"))) macro; + /* Code for a builtin macro. */ + enum cpp_builtin_type GTY ((tag ("NT_BUILTIN_MACRO"))) builtin; + /* Macro argument index. */ + unsigned short GTY ((tag ("NT_MACRO_ARG"))) arg_index; +}; + +struct GTY(()) cpp_hashnode { + struct ht_identifier ident; + unsigned int is_directive : 1; + unsigned int directive_index : 7; /* If is_directive, + then index into directive table. + Otherwise, a NODE_OPERATOR. */ + unsigned int rid_code : 8; /* Rid code - for front ends. */ + unsigned int flags : 9; /* CPP flags. */ + ENUM_BITFIELD(node_type) type : 2; /* CPP node type. */ + + /* 5 bits spare. */ + + /* The deferred cookie is applicable to NT_USER_MACRO or NT_VOID. + The latter for when a macro had a prevailing undef. + On a 64-bit system there would be 32-bits of padding to the value + field. So placing the deferred index here is not costly. */ + unsigned deferred; /* Deferred cookie */ + + union _cpp_hashnode_value GTY ((desc ("%1.type"))) value; +}; + +/* A class for iterating through the source locations within a + string token (before escapes are interpreted, and before + concatenation). */ + +class cpp_string_location_reader { + public: + cpp_string_location_reader (location_t src_loc, + line_maps *line_table); + + source_range get_next (); + + private: + location_t m_loc; + int m_offset_per_column; +}; + +/* A class for storing the source ranges of all of the characters within + a string literal, after escapes are interpreted, and after + concatenation. + + This is not GTY-marked, as instances are intended to be temporary. */ + +class cpp_substring_ranges +{ + public: + cpp_substring_ranges (); + ~cpp_substring_ranges (); + + int get_num_ranges () const { return m_num_ranges; } + source_range get_range (int idx) const + { + linemap_assert (idx < m_num_ranges); + return m_ranges[idx]; + } + + void add_range (source_range range); + void add_n_ranges (int num, cpp_string_location_reader &loc_reader); + + private: + source_range *m_ranges; + int m_num_ranges; + int m_alloc_ranges; +}; + +/* Call this first to get a handle to pass to other functions. + + If you want cpplib to manage its own hashtable, pass in a NULL + pointer. Otherwise you should pass in an initialized hash table + that cpplib will share; this technique is used by the C front + ends. */ +extern cpp_reader *cpp_create_reader (enum c_lang, struct ht *, + class line_maps *); + +/* Reset the cpp_reader's line_map. This is only used after reading a + PCH file. */ +extern void cpp_set_line_map (cpp_reader *, class line_maps *); + +/* Call this to change the selected language standard (e.g. because of + command line options). */ +extern void cpp_set_lang (cpp_reader *, enum c_lang); + +/* Set the include paths. */ +extern void cpp_set_include_chains (cpp_reader *, cpp_dir *, cpp_dir *, int); + +/* Call these to get pointers to the options, callback, and deps + structures for a given reader. These pointers are good until you + call cpp_finish on that reader. You can either edit the callbacks + through the pointer returned from cpp_get_callbacks, or set them + with cpp_set_callbacks. */ +extern cpp_options *cpp_get_options (cpp_reader *) ATTRIBUTE_PURE; +extern cpp_callbacks *cpp_get_callbacks (cpp_reader *) ATTRIBUTE_PURE; +extern void cpp_set_callbacks (cpp_reader *, cpp_callbacks *); +extern class mkdeps *cpp_get_deps (cpp_reader *) ATTRIBUTE_PURE; + +extern const char *cpp_probe_header_unit (cpp_reader *, const char *file, + bool angle_p, location_t); + +/* Call these to get name data about the various compile-time + charsets. */ +extern const char *cpp_get_narrow_charset_name (cpp_reader *) ATTRIBUTE_PURE; +extern const char *cpp_get_wide_charset_name (cpp_reader *) ATTRIBUTE_PURE; + +/* This function reads the file, but does not start preprocessing. It + returns the name of the original file; this is the same as the + input file, except for preprocessed input. This will generate at + least one file change callback, and possibly a line change callback + too. If there was an error opening the file, it returns NULL. */ +extern const char *cpp_read_main_file (cpp_reader *, const char *, + bool injecting = false); +extern location_t cpp_main_loc (const cpp_reader *); + +/* Adjust for the main file to be an include. */ +extern void cpp_retrofit_as_include (cpp_reader *); + +/* Set up built-ins with special behavior. Use cpp_init_builtins() + instead unless your know what you are doing. */ +extern void cpp_init_special_builtins (cpp_reader *); + +/* Set up built-ins like __FILE__. */ +extern void cpp_init_builtins (cpp_reader *, int); + +/* This is called after options have been parsed, and partially + processed. */ +extern void cpp_post_options (cpp_reader *); + +/* Set up translation to the target character set. */ +extern void cpp_init_iconv (cpp_reader *); + +/* Call this to finish preprocessing. If you requested dependency + generation, pass an open stream to write the information to, + otherwise NULL. It is your responsibility to close the stream. */ +extern void cpp_finish (cpp_reader *, FILE *deps_stream); + +/* Call this to release the handle at the end of preprocessing. Any + use of the handle after this function returns is invalid. */ +extern void cpp_destroy (cpp_reader *); + +extern unsigned int cpp_token_len (const cpp_token *); +extern unsigned char *cpp_token_as_text (cpp_reader *, const cpp_token *); +extern unsigned char *cpp_spell_token (cpp_reader *, const cpp_token *, + unsigned char *, bool); +extern void cpp_register_pragma (cpp_reader *, const char *, const char *, + void (*) (cpp_reader *), bool); +extern void cpp_register_deferred_pragma (cpp_reader *, const char *, + const char *, unsigned, bool, bool); +extern int cpp_avoid_paste (cpp_reader *, const cpp_token *, + const cpp_token *); +extern const cpp_token *cpp_get_token (cpp_reader *); +extern const cpp_token *cpp_get_token_with_location (cpp_reader *, + location_t *); +inline bool cpp_user_macro_p (const cpp_hashnode *node) +{ + return node->type == NT_USER_MACRO; +} +inline bool cpp_builtin_macro_p (const cpp_hashnode *node) +{ + return node->type == NT_BUILTIN_MACRO; +} +inline bool cpp_macro_p (const cpp_hashnode *node) +{ + return node->type & NT_MACRO_MASK; +} +inline cpp_macro *cpp_set_deferred_macro (cpp_hashnode *node, + cpp_macro *forced = NULL) +{ + cpp_macro *old = node->value.macro; + + node->value.macro = forced; + node->type = NT_USER_MACRO; + node->flags &= ~NODE_USED; + + return old; +} +cpp_macro *cpp_get_deferred_macro (cpp_reader *, cpp_hashnode *, location_t); + +/* Returns true if NODE is a function-like user macro. */ +inline bool cpp_fun_like_macro_p (cpp_hashnode *node) +{ + return cpp_user_macro_p (node) && node->value.macro->fun_like; +} + +extern const unsigned char *cpp_macro_definition (cpp_reader *, cpp_hashnode *); +extern const unsigned char *cpp_macro_definition (cpp_reader *, cpp_hashnode *, + const cpp_macro *); +inline location_t cpp_macro_definition_location (cpp_hashnode *node) +{ + const cpp_macro *macro = node->value.macro; + return macro ? macro->line : 0; +} +/* Return an idempotent time stamp (possibly from SOURCE_DATE_EPOCH). */ +enum class CPP_time_kind +{ + FIXED = -1, /* Fixed time via source epoch. */ + DYNAMIC = -2, /* Dynamic via time(2). */ + UNKNOWN = -3 /* Wibbly wobbly, timey wimey. */ +}; +extern CPP_time_kind cpp_get_date (cpp_reader *, time_t *); + +extern void _cpp_backup_tokens (cpp_reader *, unsigned int); +extern const cpp_token *cpp_peek_token (cpp_reader *, int); + +/* Evaluate a CPP_*CHAR* token. */ +extern cppchar_t cpp_interpret_charconst (cpp_reader *, const cpp_token *, + unsigned int *, int *); +/* Evaluate a vector of CPP_*STRING* tokens. */ +extern bool cpp_interpret_string (cpp_reader *, + const cpp_string *, size_t, + cpp_string *, enum cpp_ttype); +extern const char *cpp_interpret_string_ranges (cpp_reader *pfile, + const cpp_string *from, + cpp_string_location_reader *, + size_t count, + cpp_substring_ranges *out, + enum cpp_ttype type); +extern bool cpp_interpret_string_notranslate (cpp_reader *, + const cpp_string *, size_t, + cpp_string *, enum cpp_ttype); + +/* Convert a host character constant to the execution character set. */ +extern cppchar_t cpp_host_to_exec_charset (cpp_reader *, cppchar_t); + +/* Used to register macros and assertions, perhaps from the command line. + The text is the same as the command line argument. */ +extern void cpp_define (cpp_reader *, const char *); +extern void cpp_define_unused (cpp_reader *, const char *); +extern void cpp_define_formatted (cpp_reader *pfile, + const char *fmt, ...) ATTRIBUTE_PRINTF_2; +extern void cpp_define_formatted_unused (cpp_reader *pfile, + const char *fmt, + ...) ATTRIBUTE_PRINTF_2; +extern void cpp_assert (cpp_reader *, const char *); +extern void cpp_undef (cpp_reader *, const char *); +extern void cpp_unassert (cpp_reader *, const char *); + +/* Mark a node as a lazily defined macro. */ +extern void cpp_define_lazily (cpp_reader *, cpp_hashnode *node, unsigned N); + +/* Undefine all macros and assertions. */ +extern void cpp_undef_all (cpp_reader *); + +extern cpp_buffer *cpp_push_buffer (cpp_reader *, const unsigned char *, + size_t, int); +extern int cpp_defined (cpp_reader *, const unsigned char *, int); + +/* A preprocessing number. Code assumes that any unused high bits of + the double integer are set to zero. */ + +/* This type has to be equal to unsigned HOST_WIDE_INT, see + gcc/c-family/c-lex.cc. */ +typedef uint64_t cpp_num_part; +typedef struct cpp_num cpp_num; +struct cpp_num +{ + cpp_num_part high; + cpp_num_part low; + bool unsignedp; /* True if value should be treated as unsigned. */ + bool overflow; /* True if the most recent calculation overflowed. */ +}; + +/* cpplib provides two interfaces for interpretation of preprocessing + numbers. + + cpp_classify_number categorizes numeric constants according to + their field (integer, floating point, or invalid), radix (decimal, + octal, hexadecimal), and type suffixes. */ + +#define CPP_N_CATEGORY 0x000F +#define CPP_N_INVALID 0x0000 +#define CPP_N_INTEGER 0x0001 +#define CPP_N_FLOATING 0x0002 + +#define CPP_N_WIDTH 0x00F0 +#define CPP_N_SMALL 0x0010 /* int, float, short _Fract/Accum */ +#define CPP_N_MEDIUM 0x0020 /* long, double, long _Fract/_Accum. */ +#define CPP_N_LARGE 0x0040 /* long long, long double, + long long _Fract/Accum. */ + +#define CPP_N_WIDTH_MD 0xF0000 /* machine defined. */ +#define CPP_N_MD_W 0x10000 +#define CPP_N_MD_Q 0x20000 + +#define CPP_N_RADIX 0x0F00 +#define CPP_N_DECIMAL 0x0100 +#define CPP_N_HEX 0x0200 +#define CPP_N_OCTAL 0x0400 +#define CPP_N_BINARY 0x0800 + +#define CPP_N_UNSIGNED 0x1000 /* Properties. */ +#define CPP_N_IMAGINARY 0x2000 +#define CPP_N_DFLOAT 0x4000 +#define CPP_N_DEFAULT 0x8000 + +#define CPP_N_FRACT 0x100000 /* Fract types. */ +#define CPP_N_ACCUM 0x200000 /* Accum types. */ +#define CPP_N_FLOATN 0x400000 /* _FloatN types. */ +#define CPP_N_FLOATNX 0x800000 /* _FloatNx types. */ + +#define CPP_N_USERDEF 0x1000000 /* C++11 user-defined literal. */ + +#define CPP_N_SIZE_T 0x2000000 /* C++23 size_t literal. */ + +#define CPP_N_WIDTH_FLOATN_NX 0xF0000000 /* _FloatN / _FloatNx value + of N, divided by 16. */ +#define CPP_FLOATN_SHIFT 24 +#define CPP_FLOATN_MAX 0xF0 + +/* Classify a CPP_NUMBER token. The return value is a combination of + the flags from the above sets. */ +extern unsigned cpp_classify_number (cpp_reader *, const cpp_token *, + const char **, location_t); + +/* Return the classification flags for a float suffix. */ +extern unsigned int cpp_interpret_float_suffix (cpp_reader *, const char *, + size_t); + +/* Return the classification flags for an int suffix. */ +extern unsigned int cpp_interpret_int_suffix (cpp_reader *, const char *, + size_t); + +/* Evaluate a token classified as category CPP_N_INTEGER. */ +extern cpp_num cpp_interpret_integer (cpp_reader *, const cpp_token *, + unsigned int); + +/* Sign extend a number, with PRECISION significant bits and all + others assumed clear, to fill out a cpp_num structure. */ +cpp_num cpp_num_sign_extend (cpp_num, size_t); + +/* Output a diagnostic of some kind. */ +extern bool cpp_error (cpp_reader *, enum cpp_diagnostic_level, + const char *msgid, ...) + ATTRIBUTE_PRINTF_3; +extern bool cpp_warning (cpp_reader *, enum cpp_warning_reason, + const char *msgid, ...) + ATTRIBUTE_PRINTF_3; +extern bool cpp_pedwarning (cpp_reader *, enum cpp_warning_reason, + const char *msgid, ...) + ATTRIBUTE_PRINTF_3; +extern bool cpp_warning_syshdr (cpp_reader *, enum cpp_warning_reason reason, + const char *msgid, ...) + ATTRIBUTE_PRINTF_3; + +/* As their counterparts above, but use RICHLOC. */ +extern bool cpp_warning_at (cpp_reader *, enum cpp_warning_reason, + rich_location *richloc, const char *msgid, ...) + ATTRIBUTE_PRINTF_4; +extern bool cpp_pedwarning_at (cpp_reader *, enum cpp_warning_reason, + rich_location *richloc, const char *msgid, ...) + ATTRIBUTE_PRINTF_4; + +/* Output a diagnostic with "MSGID: " preceding the + error string of errno. No location is printed. */ +extern bool cpp_errno (cpp_reader *, enum cpp_diagnostic_level, + const char *msgid); +/* Similarly, but with "FILENAME: " instead of "MSGID: ", where + the filename is not localized. */ +extern bool cpp_errno_filename (cpp_reader *, enum cpp_diagnostic_level, + const char *filename, location_t loc); + +/* Same as cpp_error, except additionally specifies a position as a + (translation unit) physical line and physical column. If the line is + zero, then no location is printed. */ +extern bool cpp_error_with_line (cpp_reader *, enum cpp_diagnostic_level, + location_t, unsigned, + const char *msgid, ...) + ATTRIBUTE_PRINTF_5; +extern bool cpp_warning_with_line (cpp_reader *, enum cpp_warning_reason, + location_t, unsigned, + const char *msgid, ...) + ATTRIBUTE_PRINTF_5; +extern bool cpp_pedwarning_with_line (cpp_reader *, enum cpp_warning_reason, + location_t, unsigned, + const char *msgid, ...) + ATTRIBUTE_PRINTF_5; +extern bool cpp_warning_with_line_syshdr (cpp_reader *, enum cpp_warning_reason, + location_t, unsigned, + const char *msgid, ...) + ATTRIBUTE_PRINTF_5; + +extern bool cpp_error_at (cpp_reader * pfile, enum cpp_diagnostic_level, + location_t src_loc, const char *msgid, ...) + ATTRIBUTE_PRINTF_4; + +extern bool cpp_error_at (cpp_reader * pfile, enum cpp_diagnostic_level, + rich_location *richloc, const char *msgid, ...) + ATTRIBUTE_PRINTF_4; + +/* In lex.cc */ +extern int cpp_ideq (const cpp_token *, const char *); +extern void cpp_output_line (cpp_reader *, FILE *); +extern unsigned char *cpp_output_line_to_string (cpp_reader *, + const unsigned char *); +extern const unsigned char *cpp_alloc_token_string + (cpp_reader *, const unsigned char *, unsigned); +extern void cpp_output_token (const cpp_token *, FILE *); +extern const char *cpp_type2name (enum cpp_ttype, unsigned char flags); +/* Returns the value of an escape sequence, truncated to the correct + target precision. PSTR points to the input pointer, which is just + after the backslash. LIMIT is how much text we have. WIDE is true + if the escape sequence is part of a wide character constant or + string literal. Handles all relevant diagnostics. */ +extern cppchar_t cpp_parse_escape (cpp_reader *, const unsigned char ** pstr, + const unsigned char *limit, int wide); + +/* Structure used to hold a comment block at a given location in the + source code. */ + +typedef struct +{ + /* Text of the comment including the terminators. */ + char *comment; + + /* source location for the given comment. */ + location_t sloc; +} cpp_comment; + +/* Structure holding all comments for a given cpp_reader. */ + +typedef struct +{ + /* table of comment entries. */ + cpp_comment *entries; + + /* number of actual entries entered in the table. */ + int count; + + /* number of entries allocated currently. */ + int allocated; +} cpp_comment_table; + +/* Returns the table of comments encountered by the preprocessor. This + table is only populated when pfile->state.save_comments is true. */ +extern cpp_comment_table *cpp_get_comments (cpp_reader *); + +/* In hash.c */ + +/* Lookup an identifier in the hashtable. Puts the identifier in the + table if it is not already there. */ +extern cpp_hashnode *cpp_lookup (cpp_reader *, const unsigned char *, + unsigned int); + +typedef int (*cpp_cb) (cpp_reader *, cpp_hashnode *, void *); +extern void cpp_forall_identifiers (cpp_reader *, cpp_cb, void *); + +/* In macro.cc */ +extern void cpp_scan_nooutput (cpp_reader *); +extern int cpp_sys_macro_p (cpp_reader *); +extern unsigned char *cpp_quote_string (unsigned char *, const unsigned char *, + unsigned int); +extern bool cpp_compare_macros (const cpp_macro *macro1, + const cpp_macro *macro2); + +/* In files.cc */ +extern bool cpp_included (cpp_reader *, const char *); +extern bool cpp_included_before (cpp_reader *, const char *, location_t); +extern void cpp_make_system_header (cpp_reader *, int, int); +extern bool cpp_push_include (cpp_reader *, const char *); +extern bool cpp_push_default_include (cpp_reader *, const char *); +extern void cpp_change_file (cpp_reader *, enum lc_reason, const char *); +extern const char *cpp_get_path (struct _cpp_file *); +extern cpp_dir *cpp_get_dir (struct _cpp_file *); +extern cpp_buffer *cpp_get_buffer (cpp_reader *); +extern struct _cpp_file *cpp_get_file (cpp_buffer *); +extern cpp_buffer *cpp_get_prev (cpp_buffer *); +extern void cpp_clear_file_cache (cpp_reader *); + +/* cpp_get_converted_source returns the contents of the given file, as it exists + after cpplib has read it and converted it from the input charset to the + source charset. Return struct will be zero-filled if the data could not be + read for any reason. The data starts at the DATA pointer, but the TO_FREE + pointer is what should be passed to free(), as there may be an offset. */ +struct cpp_converted_source +{ + char *to_free; + char *data; + size_t len; +}; +cpp_converted_source cpp_get_converted_source (const char *fname, + const char *input_charset); + +/* In pch.cc */ +struct save_macro_data; +extern int cpp_save_state (cpp_reader *, FILE *); +extern int cpp_write_pch_deps (cpp_reader *, FILE *); +extern int cpp_write_pch_state (cpp_reader *, FILE *); +extern int cpp_valid_state (cpp_reader *, const char *, int); +extern void cpp_prepare_state (cpp_reader *, struct save_macro_data **); +extern int cpp_read_state (cpp_reader *, const char *, FILE *, + struct save_macro_data *); + +/* In lex.cc */ +extern void cpp_force_token_locations (cpp_reader *, location_t); +extern void cpp_stop_forcing_token_locations (cpp_reader *); +enum CPP_DO_task +{ + CPP_DO_print, + CPP_DO_location, + CPP_DO_token +}; + +extern void cpp_directive_only_process (cpp_reader *pfile, + void *data, + void (*cb) (cpp_reader *, + CPP_DO_task, + void *data, ...)); + +/* In expr.cc */ +extern enum cpp_ttype cpp_userdef_string_remove_type + (enum cpp_ttype type); +extern enum cpp_ttype cpp_userdef_string_add_type + (enum cpp_ttype type); +extern enum cpp_ttype cpp_userdef_char_remove_type + (enum cpp_ttype type); +extern enum cpp_ttype cpp_userdef_char_add_type + (enum cpp_ttype type); +extern bool cpp_userdef_string_p + (enum cpp_ttype type); +extern bool cpp_userdef_char_p + (enum cpp_ttype type); +extern const char * cpp_get_userdef_suffix + (const cpp_token *); + +/* In charset.cc */ + +/* The result of attempting to decode a run of UTF-8 bytes. */ + +struct cpp_decoded_char +{ + const char *m_start_byte; + const char *m_next_byte; + + bool m_valid_ch; + cppchar_t m_ch; +}; + +/* Information for mapping between code points and display columns. + + This is a tabstop value, along with a callback for getting the + widths of characters. Normally this callback is cpp_wcwidth, but we + support other schemes for escaping non-ASCII unicode as a series of + ASCII chars when printing the user's source code in diagnostic-show-locus.cc + + For example, consider: + - the Unicode character U+03C0 "GREEK SMALL LETTER PI" (UTF-8: 0xCF 0x80) + - the Unicode character U+1F642 "SLIGHTLY SMILING FACE" + (UTF-8: 0xF0 0x9F 0x99 0x82) + - the byte 0xBF (a stray trailing byte of a UTF-8 character) + Normally U+03C0 would occupy one display column, U+1F642 + would occupy two display columns, and the stray byte would be + printed verbatim as one display column. + + However when escaping them as unicode code points as "" + and "" they occupy 8 and 9 display columns respectively, + and when escaping them as bytes as "<80>" and "<9F><99><82>" + they occupy 8 and 16 display columns respectively. In both cases + the stray byte is escaped to as 4 display columns. */ + +struct cpp_char_column_policy +{ + cpp_char_column_policy (int tabstop, + int (*width_cb) (cppchar_t c)) + : m_tabstop (tabstop), + m_undecoded_byte_width (1), + m_width_cb (width_cb) + {} + + int m_tabstop; + /* Width in display columns of a stray byte that isn't decodable + as UTF-8. */ + int m_undecoded_byte_width; + int (*m_width_cb) (cppchar_t c); +}; + +/* A class to manage the state while converting a UTF-8 sequence to cppchar_t + and computing the display width one character at a time. */ +class cpp_display_width_computation { + public: + cpp_display_width_computation (const char *data, int data_length, + const cpp_char_column_policy &policy); + const char *next_byte () const { return m_next; } + int bytes_processed () const { return m_next - m_begin; } + int bytes_left () const { return m_bytes_left; } + bool done () const { return !bytes_left (); } + int display_cols_processed () const { return m_display_cols; } + + int process_next_codepoint (cpp_decoded_char *out); + int advance_display_cols (int n); + + private: + const char *const m_begin; + const char *m_next; + size_t m_bytes_left; + const cpp_char_column_policy &m_policy; + int m_display_cols; +}; + +/* Convenience functions that are simple use cases for class + cpp_display_width_computation. Tab characters will be expanded to spaces + as determined by POLICY.m_tabstop, and non-printable-ASCII characters + will be escaped as per POLICY. */ + +int cpp_byte_column_to_display_column (const char *data, int data_length, + int column, + const cpp_char_column_policy &policy); +inline int cpp_display_width (const char *data, int data_length, + const cpp_char_column_policy &policy) +{ + return cpp_byte_column_to_display_column (data, data_length, data_length, + policy); +} +int cpp_display_column_to_byte_column (const char *data, int data_length, + int display_col, + const cpp_char_column_policy &policy); +int cpp_wcwidth (cppchar_t c); + +bool cpp_input_conversion_is_trivial (const char *input_charset); +int cpp_check_utf8_bom (const char *data, size_t data_length); + +#endif /* ! LIBCPP_CPPLIB_H */ diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/include/cpplib.h" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/include/cpplib.h" index 6bf316a..aea752f 100644 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/include/cpplib.h" +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/include/cpplib.h" @@ -928,7 +928,20 @@ enum cpp_builtin_type BT_HAS_INCLUDE_NEXT, /* `__has_include_next(x)' */ // RT Extension - BT_RT_CAT /* `RT_CAT(SEP ,...)' */ + BT_RT_ASSIGN, + BT_RT_TO_ARG_LIST, + BT_RT_TO_TOKEN_LIST, + BT_RT_FIRST, + BT_RT_REST, + BT_RT_MAP, + BT_RT_AL_MAP, + BT_RT_IF, + BT_RT_NOT, + BT_RT_AND, + BT_RT_OR, + BT_RT_IS_IDENTIFIER, + BT_RT_IS_NAME, + BT_RT_PASTE }; #define CPP_HASHNODE(HNODE) ((cpp_hashnode *) (HNODE)) diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/init.cc" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/init.cc" index 2d5b0a7..36cdc6a 100644 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/init.cc" +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/init.cc" @@ -402,7 +402,22 @@ struct builtin_macro #define B(n, t, f) { DSC(n), t, f } static const struct builtin_macro builtin_array[] = { - B("RT_CAT", BT_RT_CAT, true), /* RT Extension */ + + B("_ASSIGN", BT_RT_ASSIGN, true), + B("_TO_ARG_LIST", BT_RT_TO_ARG_LIST, true), + B("_TO_TOKEN_LIST", BT_RT_TO_TOKEN_LIST, true), + B("_FIRST", BT_RT_FIRST, true), + B("_REST", BT_RT_REST, true), + B("_MAP", BT_RT_MAP, true), + B("_AL_MAP", BT_RT_AL_MAP, true), + B("_IF", BT_RT_IF, true), + B("_NOT", BT_RT_NOT, true), + B("_AND", BT_RT_AND, true), + B("_OR", BT_RT_OR, true), + B("_IS_IDENTIFIER", BT_RT_IS_IDENTIFIER, true), + B("_IS_NAME", BT_RT_IS_NAME, true), + B("_PASTE", BT_RT_PASTE, true), + B("__TIMESTAMP__", BT_TIMESTAMP, false), B("__TIME__", BT_TIME, false), diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/macro.cc" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/macro.cc" index 94801e2..ae539f9 100644 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/macro.cc" +++ "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/library/macro.cc" @@ -31,8 +31,20 @@ along with this program; see the file COPYING3. If not see #include "internal.h" // RT extension -static const uchar *evaluate_TO_ARG_LIST(cpp_reader *pfile); - + static const uchar *evaluate_RT_ASSIGN(cpp_reader *pfile); + static const uchar *evaluate_RT_TO_ARG_LIST(cpp_reader *pfile); + static const uchar *evaluate_RT_TO_TOKEN_LIST(cpp_reader *pfile); + static const uchar *evaluate_RT_FIRST(cpp_reader *pfile); + static const uchar *evaluate_RT_REST(cpp_reader *pfile); + static const uchar *evaluate_RT_MAP(cpp_reader *pfile); + static const uchar *evaluate_RT_AL_MAP(cpp_reader *pfile); + static const uchar *evaluate_RT_IF(cpp_reader *pfile); + static const uchar *evaluate_RT_NOT(cpp_reader *pfile); + static const uchar *evaluate_RT_AND(cpp_reader *pfile); + static const uchar *evaluate_RT_OR(cpp_reader *pfile); + static const uchar *evaluate_RT_IS_IDENTIFIER(cpp_reader *pfile); + static const uchar *evaluate_RT_IS_NAME(cpp_reader *pfile); + static const uchar *evaluate_RT_PASTE(cpp_reader *pfile); typedef struct macro_arg macro_arg; /* This structure represents the tokens of a macro argument. These @@ -685,10 +697,61 @@ _cpp_builtin_macro_text (cpp_reader *pfile, cpp_hashnode *node, node->value.builtin == BT_HAS_INCLUDE_NEXT); break; - case BT_RT_CAT: - result = evaluate_RT_CAT(pfile); + case BT_RT_ASSIGN: + result = evaluate_RT_ASSIGN(pfile); + break; + + case BT_RT_TO_ARG_LIST: + result = evaluate_RT_TO_ARG_LIST(pfile); + break; + + case BT_RT_TO_TOKEN_LIST: + result = evaluate_RT_TO_TOKEN_LIST(pfile); + break; + + case BT_RT_FIRST: + result = evaluate_RT_FIRST(pfile); + break; + + case BT_RT_REST: + result = evaluate_RT_REST(pfile); + break; + + case BT_RT_MAP: + result = evaluate_RT_MAP(pfile); + break; + + case BT_RT_AL_MAP: + result = evaluate_RT_AL_MAP(pfile); + break; + + case BT_RT_IF: + result = evaluate_RT_IF(pfile); break; + case BT_RT_NOT: + result = evaluate_RT_NOT(pfile); + break; + + case BT_RT_AND: + result = evaluate_RT_AND(pfile); + break; + + case BT_RT_OR: + result = evaluate_RT_OR(pfile); + break; + + case BT_RT_IS_IDENTIFIER: + result = evaluate_RT_IS_IDENTIFIER(pfile); + break; + + case BT_RT_IS_NAME: + result = evaluate_RT_IS_NAME(pfile); + break; + + case BT_RT_PASTE: + result = evaluate_RT_PASTE(pfile); + break; } @@ -3553,15 +3616,38 @@ parse_params (cpp_reader *pfile, unsigned *n_ptr, bool *varadic_ptr) return ok; } -/* Lex a token from the expansion of MACRO, but mark parameters as we - find them and warn of traditional stringification. */ +/* + "Lex a token from the expansion of MACRO, but mark parameters as we + find them and warn of traditional stringification." -original comment. + + This routine, despite its name, does no expansion. It redirects the token + pointer inside the lexer so that it writes the next token raw from the + input file into the 'expansion array' of the macro. + + The expansion array in the macro is an expandable (sort of) array of tokens, + used for holding the body of the macro. I.e. 'expansion' here refers to + the token array getting longer, not to the expansion of macros. + + This routine requires that the macro passed in is on the bump pointer buffer (the buffer returned by _cpp_reserve_room). Given this requirement, there was no need to pass the original macro as they could instead recovered it from the buffer: + + cpp_macro *macro = (cpp_macro *)BUFF_FRONT(pfile->a_buff); + + The return value is that of the buffer holding the macro with one more token on its expansion array than it had when it was passed in. +*/ static cpp_macro * lex_expansion_token (cpp_reader *pfile, cpp_macro *macro) { - macro = (cpp_macro *)_cpp_reserve_room (pfile, - sizeof (cpp_macro) - sizeof (cpp_token) - + macro->count * sizeof (cpp_token), - sizeof (cpp_token)); + macro = (cpp_macro *)_cpp_reserve_room ( + pfile + ,sizeof (cpp_macro) + - sizeof (cpp_token) + + macro->count * sizeof (cpp_token) + ,sizeof (cpp_token) + ); + + // Tells the lexer to lex the next token into ¯o->exp.tokens[macro->count]. + // Perhaps the lexer was already set to lex the next token to a different buffer? + // So just in case, its value original value is saved then restored. cpp_token *saved_cur_token = pfile->cur_token; pfile->cur_token = ¯o->exp.tokens[macro->count]; cpp_token *token = _cpp_lex_direct (pfile); @@ -4154,9 +4240,37 @@ cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node, shared declarations */ - typedef enum parse_clause_status { + typedef struct{ + unsigned int count; + cpp_token *token_array; // _cpp_reserve_room buffer set by clause parser + } token_list; + + typedef struct{ + unsigned int count; + token_list token_list[1]; + } argument_list; + + typedef enum clause_parse_delimiting { + CPD_EOF + ,CPD_BALANCED + } clause_parse_delimiting; + + typedef enum clause_parse_comma { + CPC_ERROR + ,CPC_TERMINATOR + ,CPD_IS_TOKEN + } clause_parse_comma; + + typedef enum clause_parse_expand { + CPE_NOEXPAND + ,CPE_EXPAND_OPTION + ,CPE_EXPAND + } clause_parse_expand; + + typedef enum clause_parse_status { PCS_COMPLETE // Clause completely parsed ,PCS_ERR_EXPECTED_OPEN_DELIM // Failed to find expected opening '(' + ,PCS_ERR_UNEXPECTED_COMMA // probably has too many arguments in list ,PCS_ERR_UNEXPECTED_EOF // Hit real EOF before matching ')' ,PCS_ERR_PASTE_AT_END // Trailing '##' paste operator ,PCS_ERR_HASH_NOT_FOLLOWED_BY_ARG // '#' not followed by macro parameter @@ -4164,17 +4278,23 @@ cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node, ,PCS_ERR_EOF_FETCH_FAILED // Failed to fetch next line after EOF ,PCS_ERR_UNKNOWN // Fallback error (should not occur) ,PCS_ERR_STATUS_NOT_SET // function did not set the status - } parse_clause_status; + } clause_parse_status; /*-------------------------------------------------------------------------------- debug helpers */ + // debug info for clause parsing #define DebugParseClause 1 - #define DebugAssign 1 + + // debug info for the macro directive #define DebugRTMacro 1 + // debug info for the assign directive and built in macro + #define DebugAssign 1 + + // gates compilation of functions that were defined specifically to assist with debug #define DebugHelpers 1 #if DebugHelpers @@ -4359,11 +4479,11 @@ cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node, } void print_token_list(const cpp_token *tokens ,size_t count){ - for (size_t i = 0; i < count; ++i) + for (size_t i = 0; i < countus; ++i) fprintf( stderr ,"[%zu] %s\n" ,i , cpp_token_as_text(&tokens[i]) ); } - void print_parse_clause_status(enum parse_clause_status status){ + void print_clause_parse_status(enum clause_parse_status status){ const char *message = NULL; switch (status) { @@ -4419,342 +4539,223 @@ cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node, #endif /*-------------------------------------------------------------------------------- - Parse a clause + Parse clauses - clause ::= "(" literal? ")" - | "[" expr? "]" - | tokens_to_eol ; + Clause parsers are intended to grammatical, with CPP semantics added by functions that are given a token list as an argument. We will see if this paints us into a corner soon enough I guess. - literal ::= ; sequence parsed into tokens, no expansion - expr ::= ; sequence parsed into tokens with recursive expansion of each token - tokens_to_eol ::= ; all tokens until logical end-of-line (including multi-line with `\`) + A clause is a delimited token list. Deliminators include either: + - balanced delimiters, current parens or brackets + - end of line (CPP quirk: the lexer replaces newline with EOF) + - comma - Notes: - - The first two forms are explicitly delimited with parentheses or brackets, - and may be empty (e.g., `()` or `[]`). Newlines are taken as white space. - - The third form is implicit: it consumes all remaining tokens on the directive line. - This is typical for simple macro bodies (e.g., in `#define NAME body`). + A comma in a clause is optionally: + - an error + - an alternative terminator + - merely another token -*/ + Optionally, the lexer can be told to expand tokens before they arrive at the clause parsing routine, or to not expand them. +*/ /* - Caller sees an open parenthesis or other open delimiter, and calls this. + Similar to `macro.cc::lex_expansion_token`, but returns the token rather than writing directly into an macro structure. - This parses tokens until seeing the closing delimiter. + Would be used for standard behavior when parsing a function body, but we instead elected + to use _cpp_lex_token. Note the doc `fetching_a_token.org`. - delimiter_matching == true: balances opening and closing delimiter types while searching for the balanced closing delimiter. - - paren_matching == false; terminating delimiter is CPP_EOF - which to the lexer will be end of the line. (That is how the cpp_reader does it.) +*/ +#if 0 // no longer used + cpp_token get_token_noexpand(cpp_reader *pfile){ + cpp_token result; + + // Tells the lexer to lex the next token into result. + // Perhaps the lexer was already set to lex the next token to a different buffer? + // So just in case, its value original value is saved then restored. + cpp_token *saved_cur_token = pfile->cur_token; + pfile->cur_token = &result; + _cpp_lex_direct(pfile); + pfile->cur_token = saved_cur_token; + + return result; + } +#endif - comma_list: when true, the comma becomes an alias for the final closing delimiter. For - balanced delimiters only a comma at level 0 is a terminating delimiter. +/* + Given the pfile and mode of clause parsing, returns a token list and the terminating delimiter. The token_array for the token list is allocated on the bump buffer, `pfile->a`. -*/ + When parsing a balanced paren clause, the opening paren has already been parsed, perhaps by `clause_parse`. This function then completes the parse. + When parsing a line clause, this parses the clause. -static enum parse_clause_status -parse_clause_expand(cpp_reader *pfile, - cpp_macro *macro, - bool delimiter_matching, - enum cpp_ttype opening, - enum cpp_ttype closing, - bool comma_list, - location_t *src_loc_pt, - cpp_ttype *terminator_out) -{ + Optionally stops parsing at a comma. +*/ +static enum clause_parse_status clause_parse_1( + // inputs + cpp_reader *pfile + ,clause_parse_delimiting cpd + ,enum cpp_ttype opening // needed for counting in balanced delimiters mode + ,enum cpp_ttype closing // " + ,clause_parse_comma cpc + ,bool expand // whether tokens should be expanded when lexed + + // outputs + ,token_list *tl // caller allocates the pointed to token_list + ,cpp_token *terminator // caller allocates the pointed to token, or sets terminator to null +){ #if DebugParseClause - fprintf(stderr, ">> parse_clause_expand\n"); + fprintf(stderr, ">> parse_clause\n"); fprintf(stderr, " delimiter_matching: %s\n", delimiter_matching ? "true" : "false"); fprintf(stderr, " opening token: %s (%d)\n", ttype_to_text(opening), opening); fprintf(stderr, " closing token: %s (%d)\n", ttype_to_text(closing), closing); fprintf(stderr, " comma_list: %s\n", comma_list ? "true" : "false"); - fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); + // ./include/line-map.h:typedef unsigned int location_t; + fprintf(stderr, " src_loc: %u\n", src_loc); #endif int nesting_depth = 1; - const cpp_token *token; + cpp_token token; + tl->count = 0; + tl->token_array = NULL; + location_t src_loc; + bool first = true; for(;;){ /* get a token - */ - // why not use cpp_get_token? or cpp_get_token_no_padding? - token = cpp_get_token_1 (pfile, src_loc_pt); - // this is necessary for the name expr, but does it impact potential other uses of parse_clause? Another flag for this perhaps? - if(token->type == CPP_PADDING) continue; - macro = (cpp_macro *)_cpp_reserve_room( - pfile, - sizeof(cpp_macro) + macro->count * sizeof(cpp_token), - sizeof(cpp_token) - ); - macro->exp.tokens[macro->count] = *token; - #if DebugParseClause - fprintf( stderr, "token: %s\n", cpp_token_as_text(token) ); - #endif - // lexer supports line macros by inserting CPP_EOF at line ends - if(delimiter_matching && token->type == CPP_EOF){ - #if DebugParseClause - fprintf( stderr, "CPP_EOF during parse with parentheses matching \n"); - #endif - if(!_cpp_get_fresh_line(pfile)){ - return PCS_ERR_EOF_FETCH_FAILED; - } - continue; - } - - /* parentheses matching overhead + For cpp_get_token_1 src_loc is an out parameter, = the location user would expect. + cpp_get_token_1 is defined in this file (macro.cc). */ - if(delimiter_matching){ - - if (token->type == opening) { - nesting_depth++; - } - else if (token->type == closing) { - nesting_depth--; - if (nesting_depth < 0) { - cpp_error(pfile, CPP_DL_ERROR, "unmatched closing delimiter"); - return PCS_ERR_UNEXPECTED_EOF; - } - } + if(expand){ + token = *cpp_get_token_1(pfile, &src_loc); + }else{ + token = *_cpp_lex_token(pfile); + src_loc = token.src_loc; + } - #if DebugParseClause - if( token->type == opening || token->type == closing){ - fprintf( stderr, "new nesting_depth: %d\n", nesting_depth); - } - #endif - } + /* skip padding + This is necessary for the name expr, but does it impact potential other uses of parse_clause? Another flag needed for this perhaps? - /* Determine if routine has lexed the final macro body token and should exit. + Didn't use `macro.cc::cpp_get_token_no_padding` due to the two lex options above, also because that routine does not return location. */ - bool terminted_by_matched_delimiter = - delimiter_matching - && nesting_depth == 0 - && (token->type == closing || comma_list && token->type == CPP_COMMA) - ; - - bool terminated_by_EOL = - !delimiter_matching - && (token->type == CPP_EOF || comma_list && token->type == CPP_COMMA) - ; - - if(terminted_by_matched_delimiter || terminated_by_EOL){ - if(terminator_out) *terminator_out = token->type; - return PCS_COMPLETE; - } - - // commit the new token - macro->count++; - - }// end for next token loop - -} - -bool cgls_flag = false; - -/* - See notes on parse_clause_expand + if(token.type == CPP_PADDING) continue; - This is the same but tokens in the clause are not expanded. - The end case tests here probably need to badded to parse_clause_expand also. - Perhaps expansion can be another mode bit to be sent int. - - I would have been better perhaps to send in a pointer to a token allocation, - instead of to a src_loc, and and terminal type. -*/ -static enum parse_clause_status parse_clause_literal( - cpp_reader *pfile - ,cpp_macro *macro - ,bool delimiter_matching - ,enum cpp_ttype opening - ,enum cpp_ttype closing - ,bool comma_list - ,location_t *src_loc_pt - ,cpp_ttype *terminator_out - ,unsigned int *num_extra_tokens_out -){ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_literal\n"); - fprintf(stderr, " delimiter_matching: %s\n", delimiter_matching ? "true" : "false"); - fprintf(stderr, " opening token: %s (%d)\n", ttype_to_text(opening), opening); - fprintf(stderr, " closing token: %s (%d)\n", ttype_to_text(closing), closing); - fprintf(stderr, " comma_list: %s\n", comma_list ? "true" : "false"); - fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); - #endif - - bool following_paste_op = false; - unsigned int num_extra_tokens = 0; - int nesting_depth = 1; - cpp_token *lex_token; - const char *paste_op_error_msg = - N_("'##' cannot appear at either end of a macro expansion"); - - - for(vaopt_state vaopt_tracker (pfile, macro->variadic, NULL);;){ - - /* get a token + /* Note that the lexer replaces newline with EOF when parsing a directive, but we + allow for multiple line clauses in directives. */ - // first parses lex_token onto `macro->exp.tokens[macro->count]` - // then pulls the token off of `macro->exp.tokens[macro->count]` - // reassigns macro due to possible macro->exp.tokens buffer expansion - macro = lex_expansion_token(pfile, macro); - lex_token = ¯o->exp.tokens[macro->count]; - *src_loc_pt = lex_token->src_loc; - #if DebugParseClause - fprintf( stderr, "lex_token: %s\n", cpp_token_as_text(lex_token) ); - #endif - - // lexer will insert CPP_EOF at the end of each line, because cpp originally only did line macros. - if(delimiter_matching && lex_token->type == CPP_EOF){ + if(cpd != CPD_EOF && token.type == CPP_EOF){ #if DebugParseClause fprintf( stderr, "CPP_EOF during parse with parentheses matching \n"); #endif if(!_cpp_get_fresh_line(pfile)){ + if(terminator) *terminator = token; return PCS_ERR_EOF_FETCH_FAILED; } continue; } - /* tag macro args - */ - if (macro->count > 1 && lex_token[-1].type == CPP_HASH && macro->fun_like) - { - if (lex_token->type == CPP_MACRO_ARG - || (macro->variadic - && lex_token->type == CPP_NAME - && lex_token->val.node.node == pfile->spec_nodes.n__VA_OPT__)) - { - if (lex_token->flags & PREV_WHITE) - lex_token->flags |= SP_PREV_WHITE; - if (lex_token[-1].flags & DIGRAPH) - lex_token->flags |= SP_DIGRAPH; - lex_token->flags &= ~PREV_WHITE; - lex_token->flags |= STRINGIFY_ARG; - lex_token->flags |= lex_token[-1].flags & PREV_WHITE; - lex_token[-1] = lex_token[0]; - macro->count--; - } - else if (CPP_OPTION (pfile, lang) != CLK_ASM) - { - cpp_error(pfile, CPP_DL_ERROR, - "'#' is not followed by a macro parameter"); - return PCS_ERR_HASH_NOT_FOLLOWED_BY_ARG; - } - } - - /* paste end cases + /* if we shouldn't see a comma */ - if (lex_token->type == CPP_PASTE) - { - if (macro->count == 0) - { - cpp_error(pfile, CPP_DL_ERROR, paste_op_error_msg); - return PCS_ERR_PASTE_AT_END; // the font end of the buffer - } - - if (following_paste_op) - { - num_extra_tokens++; - lex_token->val.token_no = macro->count - 1; - } - else - { - --macro->count; - lex_token[-1].flags |= PASTE_LEFT; - if (lex_token->flags & DIGRAPH) - lex_token[-1].flags |= SP_DIGRAPH; - if (lex_token->flags & PREV_WHITE) - lex_token[-1].flags |= SP_PREV_WHITE; - } - following_paste_op = true; - } - else{ - following_paste_op = false; + if( cpc == CPC_ERROR && token.type == CPP_COMMA ){ + if(terminator) *terminator = token; + return PCS_ERR_UNEXPECTED_COMMA; } /* parentheses matching overhead */ - if(delimiter_matching){ - - if (lex_token->type == opening) { + if(cpd == CPD_BALANCED){ + if (token.type == opening) { nesting_depth++; } - else if (lex_token->type == closing) { + else if (token.type == closing) { nesting_depth--; if (nesting_depth < 0) { cpp_error(pfile, CPP_DL_ERROR, "unmatched closing delimiter"); + if(terminator) *terminator = token; return PCS_ERR_UNEXPECTED_EOF; } } - #if DebugParseClause - if( lex_token->type == opening || lex_token->type == closing){ + if( token.type == opening || token.type == closing){ fprintf( stderr, "new nesting_depth: %d\n", nesting_depth); } #endif } - /* Determine if routine has lexed the final macro body token and should exit. + /* Determine if clause has reached a terminator */ bool terminated_by_matched_delimiter = - delimiter_matching - && nesting_depth == 0 - && (lex_token->type == closing || comma_list && lex_token->type == CPP_COMMA) + cpd == CPD_BALANCED + && nesting_depth == 0 + && token.type == closing ; - bool terminated_by_EOL = - !delimiter_matching - && (lex_token->type == CPP_EOF || comma_list && lex_token->type == CPP_COMMA) + bool terminated_by_comma = + cpc == CPC_TERMINATOR + && token.type == CPP_COMMA ; - if(terminated_by_matched_delimiter || terminated_by_EOL){ - - if(following_paste_op){ - cpp_error(pfile, CPP_DL_ERROR, paste_op_error_msg); - return PCS_ERR_PASTE_AT_END; - } - if (vaopt_tracker.update(lex_token) == vaopt_state::ERROR){ - return PCS_ERR_VAOPT_STATE_INVALID; - } - if( !vaopt_tracker.completed() ){ - return PCS_ERR_VAOPT_STATE_INVALID; - } + bool terminated_by_EOL = + cpd == CPD_EOF + && token.type == CPP_EOF + ; - if(num_extra_tokens_out) *num_extra_tokens_out = num_extra_tokens; - if(terminator_out) *terminator_out = lex_token->type; + if( + terminated_by_matched_delimiter + || terminted_by_comma + || terminated_by_EOL + ){ + if(terminator) *terminator = token; return PCS_COMPLETE; } - // commit the new token - macro->count++; + // store the token in the token list + tl->token_array = (cpp_token *)_cpp_reserve_room( + pfile + ,tl->count * sizeof(cpp_token) + ,sizeof(cpp_token) + ); + tl->token_array[tl->count] = token; + tl->count++; + #if DebugParseClause + fprintf( stderr, "token: %s\n", cpp_token_as_text(&token) ); + #endif }// end for next token loop } + /* - The opening paren has been parsed, this completes parsing the clause. + Request to parse a clause - Given a cpp_macro and cpp_reader reference. - Returns the body tokens in `macro->exp.tokens`. + If given expand_option, square taken as expand, and paren taken as no expand. - The macro need not have been committed. -*/ + Check the returned terminator token to determine which option occurred. -static enum parse_clause_status parse_clause_balanced( - cpp_reader *pfile - ,cpp_macro *macro - ,enum cpp_ttype opening - ,enum cpp_ttype closing - ,bool expand - ,bool comma_list - ,location_t *src_loc_pt - ,cpp_ttype *terminator_out - ,unsigned int *num_extra_tokens_out + When there is an optional comma for termination, check the terminator to see + if the clause parse ran into a comma. + + For the continuing, end of line, and expand option clause types, the input opening and closing arguments are ignored. + +*/ +static enum clause_parse_status clause_parse( + // inputs + cpp_reader *pfile + ,clause_parse_delimiting cpd + ,enum cpp_ttype opening // needed for counting in balanced delimiters mode + ,enum cpp_ttype closing // " + ,bool continuing // typically true when continuing after bumping into a comma terminator + ,clause_parse_comma cpc + ,clause_parse_expand cpe + + // outputs + ,token_list *tl // caller allocates the pointed to token_list + ,cpp_token *terminator // caller allocates the pointed to token, or sets terminator to null ){ + #if DebugParseClause fprintf(stderr, ">> parse_clause_balanced\n"); fprintf(stderr, " opening token: %s (%d)\n", ttype_to_text(opening), opening); @@ -4764,167 +4765,75 @@ static enum parse_clause_status parse_clause_balanced( fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); #endif - /* check for opening paren - */ - const cpp_token *token = _cpp_lex_token(pfile); - *src_loc_pt = token->src_loc; - #if DebugParseClause - fprintf(stderr, "checking if token is opening: %s" ,cpp_token_as_text(token)); - #endif - - if( token->type != opening ){ - _cpp_backup_tokens(pfile, 1); - return PCS_ERR_EXPECTED_OPEN_DELIM; - } - - /* complete the parse - */ - - // make comma an alias for the terminating delimiter - // this gets passed on directly to parse_clause_{literal,expand} - bool delimiter_matching = true; - - int saved_keep_tokens = pfile->keep_tokens; - int saved_in_directive = pfile->state.in_directive; - pfile->keep_tokens = 1; - pfile->state.in_directive = 0; - - parse_clause_status status; - if(expand){ - status = parse_clause_expand - (pfile - ,macro - ,delimiter_matching - ,opening ,closing - ,comma_list - ,src_loc_pt - ,terminator_out - ); - if(num_extra_tokens_out) *num_extra_tokens_out = 0; - #if DebugParseClause - fprintf(stderr, ">> parse_clause_balanced expand status:"); - print_parse_clause_status(status); - #endif - }else{ - status = parse_clause_literal - (pfile - ,macro - ,delimiter_matching - ,opening ,closing - ,comma_list - ,src_loc_pt - ,terminator_out - ,num_extra_tokens_out - ); - #if DebugParseClause - fprintf(stderr, ">> parse_clause_balanced literal status:"); - print_parse_clause_status(status); - #endif - } - - // Restore parser state - pfile->keep_tokens = saved_keep_tokens; - pfile->state.in_directive = saved_in_directive; + if( + continuing + || cpd == CPD_EOF + ){ + return clause_parse_1( + pfile + ,cpd + ,opening + ,closing + ,cpc + ,cpe + ,tl + ,terminator + ); + } + const cpp_token *token = _cpp_lex_token(pfile); #if DebugParseClause - fprintf(stderr, ">> parse_clause_balanced final returning status:"); - print_parse_clause_status(status); + fprintf(stderr, "opening token: %s" ,cpp_token_as_text(token)); #endif - return status; -} -typedef enum PCPSO_choice{ - PCPSO_unknown = 0 - ,PCPSO_paren - ,PCPSO_square -} PCPSO_choice; - -parse_clause_status parse_clause_paren_square_option( - cpp_reader *pfile - ,cpp_macro *macro - ,PCPSO_choice *choice - ,bool comma_list - ,location_t *src_loc_pt - ,cpp_ttype *terminator_out - ,unsigned int *num_extra_tokens_out -){ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_paren_square_option\n"); - fprintf(stderr, " comma_list: %s\n", comma_list ? "true" : "false"); - fprintf(stderr, " src_loc_pt: %p\n", (void *)src_loc_pt); - #endif + if(cpe == CPE_EXPAND_OPTION){ + switch(token.type){ + CPP_OPEN_SQUARE: + return clause_parse_1( + pfile + ,CPD_BALANCED + ,CPP_OPEN_SQUARE + ,CPP_CLOSE_SQUARE + ,cpc + ,CPE_EXPAND + ,tl + ,terminator + ); - *choice = PCPSO_unknown; - - parse_clause_status status = parse_clause_balanced - (pfile - ,macro - ,CPP_OPEN_PAREN - ,CPP_CLOSE_PAREN - ,false // no expand - ,comma_list - ,src_loc_pt - ,terminator_out - ,num_extra_tokens_out - ); - - if(status != PCS_ERR_EXPECTED_OPEN_DELIM){ - *choice = PCPSO_paren; - return status; - } + CPP_OPEN_SQUARE: + return clause_parse_1( + pfile + ,CPD_BALANCED + ,CPP_OPEN_PAREN + ,CPP_CLOSE_PAREN + ,cpc + ,CPE_NOEXPAND + ,tl + ,terminator + ); - status = parse_clause_balanced - (pfile - ,macro - ,CPP_OPEN_SQUARE - ,CPP_CLOSE_SQUARE - ,true // expand - ,comma_list - ,src_loc_pt - ,terminator_out - ,num_extra_tokens_out - ); - - if(status == PCS_ERR_EXPECTED_OPEN_DELIM){ - #if DebugParseClause - fprintf(stderr, ">> parse_clause_paren_square_option early returning status:"); - print_parse_clause_status(PCS_ERR_EXPECTED_OPEN_DELIM); - #endif - return PCS_ERR_EXPECTED_OPEN_DELIM; + default: + return PCS_ERR_EXPECTED_OPEN_DELIM; + } } + + // at this point we must be doing CPD_BALANCED with no square/paren option - *choice = PCPSO_square; + if(token.type != opening) return PCS_ERR_EXPECTED_OPEN_DELIM; - #if DebugParseClause - fprintf(stderr, ">> parse_clause_paren_square_option final returning status:"); - print_parse_clause_status(status); - #endif - return status; -} + return clause_parse_1( + pfile + ,cpd + ,opening + ,closing + ,cpc + ,cpe + ,tl + ,terminator + ); -/* -The prior clause parse terminated in a comma. For parsing a comma list of clauses. -*/ -static enum parse_clause_status parse_clause_comma_continue( -){ - return PCS_ERR_UNKNOWN; } -/* - Check if a collected macro body reduces to a single identifier token. - - Preconditions: - - macro is non-null - - macro->exp.tokens has been populated (e.g., via parse_clause) - - macro->count is valid - - Returns: - - cpp_hashnode* if valid (i.e., single CPP_NAME token) - - NULL if invalid, and emits error message - - -Note in do_define in directives.cc there is some logic related to callbacks and warning if trying to redefine a built-in macro. That should be integrated here. - */ static cpp_hashnode * @@ -4961,7 +4870,7 @@ name_clause_is_name(cpp_reader *pfile, const cpp_macro *macro) bool _cpp_create_assign(cpp_reader *pfile){ - parse_clause_status status; + clause_parse_status status; location_t src_loc; unsigned int num_extra_tokens = 0; PCPSO_choice choice; @@ -4983,7 +4892,7 @@ bool _cpp_create_assign(cpp_reader *pfile){ status = parse_clause_paren_square_option( pfile ,name_macro - ,&choice + ,&choice // square brackets, or round parents? ,false // not a commas list ,&src_loc ,NULL // don't need to know the terminator @@ -4997,7 +4906,7 @@ bool _cpp_create_assign(cpp_reader *pfile){ print_token_list(name_macro->exp.tokens ,name_macro->count); #endif - /* The name clause must be either a literally valid name, or it must expand into + /* The name clause must be either a noexpandly valid name, or it must expand into a valid name, depending if the programmer used () or []. If valid, keep the name node. */ @@ -5066,7 +4975,7 @@ bool _cpp_create_assign(cpp_reader *pfile){ status = parse_clause_paren_square_option( pfile ,body_macro - ,&choice + ,&choice // square brackets, or round parents? ,false // not a commas list ,&src_loc ,NULL // don't need to know the terminator @@ -5157,7 +5066,7 @@ static cpp_macro *create_rt_macro (cpp_reader *pfile){ bool varadic = false; bool ok = false; cpp_macro *macro = NULL; - parse_clause_status status; + clause_parse_status status; /* parse parameter list @@ -5198,7 +5107,7 @@ static cpp_macro *create_rt_macro (cpp_reader *pfile){ ,cmk_macro ,_cpp_reserve_room( pfile, 0, sizeof(cpp_macro) ) ); - // used by parse_clause_literal + // used by parse_clause_noexpand macro->variadic = varadic; macro->paramc = paramc; macro->parm.params = params; @@ -5216,7 +5125,7 @@ static cpp_macro *create_rt_macro (cpp_reader *pfile){ ); if( status != PCS_COMPLETE ){ fprintf(stderr, "parse_paren_clause returned: "); - print_parse_clause_status(status); + print_clause_parse_status(status); goto out; } @@ -5365,10 +5274,265 @@ _cpp_create_rt_macro(cpp_reader *pfile, cpp_hashnode *node){ } /*-------------------------------------------------------------------------------- - builtin RT_CAT macro RT extension + RT builtin macro extensions */ -static const uchar *evaluate_RT_CAT(cpp_reader *pfile){ - return UC"XY"; + static const uchar *evaluate_RT_ASSIGN(cpp_reader *pfile){ + if( ! _cpp_create_assign(pfile) ){ + cpp_error( + pfile + ,CPP_DL_ERROR + ,"#assign macro failed" + ); + } + // return UC""; // returning null string gave an internal compiler error + return NULL; // expands as `1` dunno why, but the code is there to make it happen + // return UC" "; // another internal compiler error + } + + static const uchar *evaluate_RT_TO_ARG_LIST(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#to_arg_list macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_TO_TOKEN_LIST(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#to_token_list macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_FIRST(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#first macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_REST(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#rest macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_MAP(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#map macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_AL_MAP(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#al_map macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_IF(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#if macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_NOT(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#not macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_AND(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#and macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_OR(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#or macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_IS_IDENTIFIER(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#is_identifier macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_IS_NAME(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#is_name macro evaluated" + ); + return NULL; + } + + static const uchar *evaluate_RT_PASTE(cpp_reader *pfile){ + cpp_error( + pfile + ,CPP_DL_NOTE + ,"#paste macro evaluated" + ); + return NULL; + } + + +#if 0 + static const uchar *evaluate_RT_ASSIGN(cpp_reader *pfile){ + return UC"_ASSIGN"; + } + + static const uchar *evaluate_RT_TO_ARG_LIST(cpp_reader *pfile){ + return UC"_TO_ARG_LIST"; + } + + static const uchar *evaluate_RT_TO_TOKEN_LIST(cpp_reader *pfile){ + return UC"_TO_TOKEN_LIST"; + } + + static const uchar *evaluate_RT_FIRST(cpp_reader *pfile){ + return UC"_FIRST"; + } + + static const uchar *evaluate_RT_REST(cpp_reader *pfile){ + return UC"_REST"; + } + + static const uchar *evaluate_RT_MAP(cpp_reader *pfile){ + return UC"_MAP"; + } + + static const uchar *evaluate_RT_AL_MAP(cpp_reader *pfile){ + return UC"_AL_MAP"; + } + + static const uchar *evaluate_RT_IF(cpp_reader *pfile){ + return UC"_IF"; + } + + static const uchar *evaluate_RT_NOT(cpp_reader *pfile){ + return UC"_NOT"; + } + + static const uchar *evaluate_RT_AND(cpp_reader *pfile){ + return UC"_AND"; + } + + static const uchar *evaluate_RT_OR(cpp_reader *pfile){ + return UC"_OR"; + } + + static const uchar *evaluate_RT_IS_IDENTIFIER(cpp_reader *pfile){ + return UC"_IS_IDENTIFIER"; + } + + static const uchar *evaluate_RT_IS_NAME(cpp_reader *pfile){ + return UC"_IS_NAME"; + } + + static const uchar *evaluate_RT_PASTE(cpp_reader *pfile){ + return UC"_PASTE"; + } +#endif +#if 0 +/*───────────────────────── RT helper utilities ─────────────────────────*/ + +/* rt_read_paren_argument + Consume exactly one parenthesised argument list, collecting every token + between the outer ‘( … )’ into OUT_LIST. Returns true on success and + emits its own diagnostic on failure. */ +static bool rt_read_paren_argument(cpp_reader *pfile ,vec &out_list){ + const cpp_token *tok = cpp_get_token_no_padding(pfile); + if(tok->type != CPP_OPEN_PAREN){ + cpp_error(pfile ,CPP_DL_ERROR ,"missing '(' after built-in macro"); + return false; + } + unsigned depth = 1; + while(depth){ + tok = cpp_get_token_1(pfile ,nullptr); + if(tok->type == CPP_EOF){ + cpp_error(pfile ,CPP_DL_ERROR ,"unterminated argument list in built-in macro"); + return false; + } + if(tok->type == CPP_OPEN_PAREN) depth++; + else if(tok->type == CPP_CLOSE_PAREN){ + if(--depth == 0) break; + } + if(depth) out_list.safe_push(*tok); /* omit the final ‘)’ */ + } + return true; } + +/* rt_tokens_as_text + Spell a vector of tokens back into a single space-separated byte string + allocated from the preprocessor’s permanent pool. */ +static const uchar *rt_tokens_as_text(cpp_reader *pfile ,const vec &src){ + size_t reserve = src.length()*20 + 1; /* generous bound */ + uchar *buf = _cpp_unaligned_alloc(pfile ,reserve); + uchar *dst = buf; + for(unsigned i = 0 ;i < src.length() ;++i){ + if(i) *dst++ = ' '; + dst = cpp_spell_token(pfile ,&src[i] ,dst ,true); + } + *dst = '\0'; + return buf; +} + +/*──────────────────── _FIRST(token_list) implementation ─────────────────*/ + +static const uchar *evaluate_RT_FIRST(cpp_reader *pfile){ + vec list; + if(!rt_read_paren_argument(pfile ,list)) return UC""; + unsigned idx = 0; + while(idx < list.length() && list[idx].type == CPP_PADDING) ++idx; + if(idx == list.length()) return UC""; /* empty list */ + vec one; + one.safe_push(list[idx]); + return rt_tokens_as_text(pfile ,one); +} + +/*──────────────────── _REST(token_list) implementation ──────────────────*/ + +static const uchar *evaluate_RT_REST(cpp_reader *pfile){ + vec list; + if(!rt_read_paren_argument(pfile ,list)) return UC""; + unsigned first = 0; + while(first < list.length() && list[first].type == CPP_PADDING) ++first; + if(first >= list.length() - 1) return UC""; /* one-token list */ + vec rest; + for(unsigned i = first + 1 ;i < list.length() ;++i) rest.safe_push(list[i]); + return rt_tokens_as_text(pfile ,rest); +} +#endif diff --git "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/todo.org" "b/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/todo.org" deleted file mode 100644 index 0736eda..0000000 --- "a/developer/script_Deb-12.10_gcc-12.4.1\360\237\226\211/todo.org" +++ /dev/null @@ -1,14 +0,0 @@ -2025-05-00 - -- Add the call back and warn logic for #assign in the macro.cc::name_clause_is_name function. - -- The name is currently () or [], probably should allow a single name ID or [] instead. - or perhaps in general, allow for evaluated, not evaluated, or single identifier options. - -- When this matures, should replace the capture/install with diff and patch. - - -2025-05-17 in maco.cc, seems the end cases in `parse_clause_literal()` should be included in `parse_clause_expand()`. - -2025-05-18 It would have been better perhaps to send in a pointer to a token allocation, - instead of to a src_loc, and and terminal type. diff --git a/env_tester b/env_tester index 1a3f89e..ff0809c 100644 --- a/env_tester +++ b/env_tester @@ -6,8 +6,8 @@ if [[ "${BASH_SOURCE[0]}" == "$0" ]]; then fi export REPO_HOME=$(dirname "$script_afp") -export DEVELOPER="$REPO_HOME/developer" -export SCRIPT_DIR="$DEVELOPER"/script_Deb-12.10_gcc-12.4.1🖉 +export TESTER="$REPO_HOME/tester" export PATH="$REPO_HOME/release/amd64_Deb-12.10_gcc-12.4.1/bin:$PATH" +cd "$TESTER" diff --git "a/tester/experiment\360\237\226\211/.gitignore" "b/tester/experiment\360\237\226\211/.gitignore" new file mode 100644 index 0000000..cba7efc --- /dev/null +++ "b/tester/experiment\360\237\226\211/.gitignore" @@ -0,0 +1 @@ +a.out diff --git "a/tester/experiment\360\237\226\211/RT_CAT_1.c" "b/tester/experiment\360\237\226\211/RT_CAT_1.c" new file mode 100644 index 0000000..17663b1 --- /dev/null +++ "b/tester/experiment\360\237\226\211/RT_CAT_1.c" @@ -0,0 +1,6 @@ +#include + +int main(void){ + printf( "The answer is: %s\n", RT_CAT ); + return 0; +} diff --git "a/tester/experiment\360\237\226\211/UC_list.c" "b/tester/experiment\360\237\226\211/UC_list.c" new file mode 100644 index 0000000..952f231 --- /dev/null +++ "b/tester/experiment\360\237\226\211/UC_list.c" @@ -0,0 +1,28 @@ +#include +/* +tests that built in macro stubs are present, this will not work when the macros have implementations. +*/ + +#if 0 + _ASSIGN + _TO_ARG_LIST + _TO_TOKEN_LIST + _FIRST + _REST + _MAP + _AL_MAP + _IF + _NOT + _AND + _OR + _IS_IDENTIFIER + _IS_NAME + _PASTE +#endif + +int main(void){ + _ASSIGN(X)(5); + printf("X: %d" ,X); + + return 0; +} diff --git "a/tester/experiment\360\237\226\211/assign_1.c" "b/tester/experiment\360\237\226\211/assign_1.c" new file mode 100644 index 0000000..f58320c --- /dev/null +++ "b/tester/experiment\360\237\226\211/assign_1.c" @@ -0,0 +1,8 @@ +#include + +#assign (ANSWER) (42) + +int main(void){ + printf( "The answer is: %d\n", ANSWER ); + return 0; +} diff --git "a/tester/experiment\360\237\226\211/assign_2.c" "b/tester/experiment\360\237\226\211/assign_2.c" new file mode 100644 index 0000000..3a41a63 --- /dev/null +++ "b/tester/experiment\360\237\226\211/assign_2.c" @@ -0,0 +1,22 @@ +#include + +#if 0 +#define STRINGIFY(x) #x +#define TOSTRING(x) STRINGIFY(x) +#define SHOW_MACRO(x) _Pragma(TOSTRING(message(#x " → " TOSTRING(x)))) + +SHOW_MACRO(a) +SHOW_MACRO(b) + +SHOW_MACRO($a) +SHOW_MACRO($b) +#endif + +#define a 2 +#define b 3 +#assign (ADD) [a + b] + +int main(void){ + printf("2 + 3 = %d\n", ADD); + return 0; +} diff --git "a/tester/experiment\360\237\226\211/macro_1.c" "b/tester/experiment\360\237\226\211/macro_1.c" new file mode 100644 index 0000000..2ba3e31 --- /dev/null +++ "b/tester/experiment\360\237\226\211/macro_1.c" @@ -0,0 +1,8 @@ +#include + +#rt_macro I(x) (x) + +int main(void){ + printf("5: %x" ,I(5)); + return 0; +} diff --git "a/tester/experiment\360\237\226\211/macro_2.c" "b/tester/experiment\360\237\226\211/macro_2.c" new file mode 100644 index 0000000..421f176 --- /dev/null +++ "b/tester/experiment\360\237\226\211/macro_2.c" @@ -0,0 +1,11 @@ +#include + +#rt_macro Q(f ,...)( + printf(f ,__VA_ARGS__) +) + +int main(void){ + Q("%x %x %x" ,1 ,2 ,3); + putchar('\n'); + return 0; +} diff --git "a/tester/experiment\360\237\226\211/paste_1.c" "b/tester/experiment\360\237\226\211/paste_1.c" new file mode 100644 index 0000000..d438c6b --- /dev/null +++ "b/tester/experiment\360\237\226\211/paste_1.c" @@ -0,0 +1,10 @@ +#include + +#define CAT2(a ,b) a##b +#define STRING_1(a) #a +#define STRING(a) STRING_1(a) + +int main(){ + printf("ab: %s." ,STRING(CAT2(a ,b))); + printf("gcc -E this one: %s." ,STRING(CAT2(a ,+))); +} diff --git "a/tester/experiment\360\237\226\211/recursive_define_0.c" "b/tester/experiment\360\237\226\211/recursive_define_0.c" new file mode 100644 index 0000000..da2919a --- /dev/null +++ "b/tester/experiment\360\237\226\211/recursive_define_0.c" @@ -0,0 +1,52 @@ +#include + + +#define X Q()B()(C) +#define Q() A +#define A(x) 21 +#define B() + +int main(){ + printf("X %d" ,X); +} + +/* + +So cpp truly does walk the token list with `cpp_get_token_1(pfile, &loc);` expanding each token one by one left to right. We should be able to get the desired expansion by adding another layer of macros, so that the argument expansion catches the final A(c). + +2025-05-12T02:53:59Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> gcc test_recursive_define_0.c +test_recursive_define_0.c: In function ‘main’: +test_recursive_define_0.c:5:13: warning: implicit declaration of function ‘A’ [-Wimplicit-function-declaration] + 5 | #define Q() A + | ^ +test_recursive_define_0.c:4:12: note: in expansion of macro ‘Q’ + 4 | #define X Q()B()(C) + | ^ +test_recursive_define_0.c:10:18: note: in expansion of macro ‘X’ + 10 | printf("X %d" ,X); + | ^ +test_recursive_define_0.c:4:19: error: ‘C’ undeclared (first use in this function) + 4 | #define X Q()B()(C) + | ^ +test_recursive_define_0.c:10:18: note: in expansion of macro ‘X’ + 10 | printf("X %d" ,X); + | ^ +test_recursive_define_0.c:4:19: note: each undeclared identifier is reported only once for each function it appears in + 4 | #define X Q()B()(C) + | ^ +test_recursive_define_0.c:10:18: note: in expansion of macro ‘X’ + 10 | printf("X %d" ,X); + | ^ + +2025-05-12T02:54:17Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> gcc -E -P test_recursive_define_0.c | tail -n 5 +extern int __overflow (FILE *, int); + +int main(){ + printf("X %d" ,A(C)); +} + +*? diff --git "a/tester/experiment\360\237\226\211/recursive_define_1.c" "b/tester/experiment\360\237\226\211/recursive_define_1.c" new file mode 100644 index 0000000..a7c6335 --- /dev/null +++ "b/tester/experiment\360\237\226\211/recursive_define_1.c" @@ -0,0 +1,44 @@ +#include + + +#define X Q()B()(C) +#define Q() A +#define A(x) 21 +#define B() + +#define P(x) x + +int main(){ + printf("X %d" ,P(X)); +} + +/* +2025-05-12T03:00:19Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> /bin/gcc test_recursive_define_1.c + +2025-05-12T03:00:51Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> ./a.out +X 21 +2025-05-12T03:00:55Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> gcc test_recursive_define_1.c + +2025-05-12T03:01:03Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> ./a.out +X 21 +2025-05-12T03:01:05Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> gcc -E -P test_recursive_define_1.c | tail -n 5 +extern int __overflow (FILE *, int); + +int main(){ + printf("X %d" ,21); +} + +2025-05-12T03:01:11Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> +*/ diff --git "a/tester/experiment\360\237\226\211/recursive_define_2.c" "b/tester/experiment\360\237\226\211/recursive_define_2.c" new file mode 100644 index 0000000..ecd87cc --- /dev/null +++ "b/tester/experiment\360\237\226\211/recursive_define_2.c" @@ -0,0 +1,58 @@ +#include + + +#define X Q()B()(C) +#define Q() A +#define A(x) 21 +#define B() + +#define P X + +int main(){ + printf("X %d" ,P); +} + +/* + This does not work because it does not cause an argument evaluation, as did test_recursive_define_1.c + +2025-05-12T03:01:11Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> gcc test_recursive_define_2.c +test_recursive_define_2.c: In function ‘main’: +test_recursive_define_2.c:5:13: warning: implicit declaration of function ‘A’ [-Wimplicit-function-declaration] + 5 | #define Q() A + | ^ +test_recursive_define_2.c:4:12: note: in expansion of macro ‘Q’ + 4 | #define X Q()B()(C) + | ^ +test_recursive_define_2.c:9:11: note: in expansion of macro ‘X’ + 9 | #define P X + | ^ +test_recursive_define_2.c:12:18: note: in expansion of macro ‘P’ + 12 | printf("X %d" ,P); + | ^ +test_recursive_define_2.c:4:19: error: ‘C’ undeclared (first use in this function) + 4 | #define X Q()B()(C) + | ^ +test_recursive_define_2.c:9:11: note: in expansion of macro ‘X’ + 9 | #define P X + | ^ +test_recursive_define_2.c:12:18: note: in expansion of macro ‘P’ + 12 | printf("X %d" ,P); + | ^ +test_recursive_define_2.c:4:19: note: each undeclared identifier is reported only once for each function it appears in + 4 | #define X Q()B()(C) + | ^ +test_recursive_define_2.c:9:11: note: in expansion of macro ‘X’ + 9 | #define P X + | ^ +test_recursive_define_2.c:12:18: note: in expansion of macro ‘P’ + 12 | printf("X %d" ,P); + | ^ + +2025-05-12T03:02:36Z[developer] +Thomas-developer@StanleyPark§/home/Thomas/subu_data/developer/N/developer/experiment§ +> +*/ + + diff --git "a/tester/experiment\360\237\226\211/temp" "b/tester/experiment\360\237\226\211/temp" new file mode 100644 index 0000000..32b29d9 --- /dev/null +++ "b/tester/experiment\360\237\226\211/temp" @@ -0,0 +1,9 @@ +git mv assign_test_1.c assign_1.c +git mv assign_test_2.c assign_2.c +git mv macro_test_1.c macro_1.c +git mv macro_test_2.c macro_2.c +git mv recursive_define_0.c recursive_define_0.c +git mv recursive_define_1.c recursive_define_1.c +git mv recursive_define_2.c recursive_define_2.c +git mv RT_CAT_test_1.c RT_CAT_1.c +git mv va_arg_test_1.c va_arg_1.c diff --git "a/tester/experiment\360\237\226\211/va_arg_test_1.c" "b/tester/experiment\360\237\226\211/va_arg_test_1.c" new file mode 100644 index 0000000..8870054 --- /dev/null +++ "b/tester/experiment\360\237\226\211/va_arg_test_1.c" @@ -0,0 +1,9 @@ +#include + +#define A(...) (__VA_ARGS__) + + +int main(void){ + printf( "The answer is: %d\n", A(1,2,3) ); + return 0; +}