6 Commits

Author SHA1 Message Date
Dmitry Polukhin
0b7e8c25ec
[C++20][Modules] Quote header unit name in preprocessor output (-E) (#112883)
Summary:
Before this change clang produced output with header unit names that may
conaint path separators, dots and other non-identifier characters. This
diff prints header unit name in quotes and -E output can be compiled
again. Also remove unnecessary space between header unit name and semi.

Test Plan: check-clang
2024-10-24 08:20:43 +01:00
Fangrui Song
7e59223ac4 [test] %clang_cc1: remove redundant actions
ParseFrontendArgs takes the last OPT_Action_Group option. The other
actions are overridden.
2024-05-05 10:46:06 -07:00
Jan Svoboda
5523fefb01 [clang][lex] Use preferred path separator in includer-relative lookup
There is a long-standing FIXME in `HeaderSearch.cpp` to use the path separator preferred by the platform instead of forward slash. There was an attempt to fix that (1cf6c28a) which got reverted (cf385dc8). I couldn't find an explanation, but my guess is that some tests assuming forward slash started failing.

This commit fixes tests with that assumption.

This is intended to be NFC, but there are two exceptions to that:
* Some diagnostic messages might now contain backslash instead of forward slash.
* Arguments to the "-remap-file" option that use forward slash might stop kicking in. Separators between potential includer path and header name need to be replaced by backslash in that case.
2023-09-08 16:38:08 -07:00
Chuanqi Xu
fc89e9044d [C++20] [Modules] Emit an warning for experimental header units
Currently, the header units are rarely used and it is not well tested.
To avoid further misunderstandings, let's mark it as experimental and
emit a warning when users wants to import it.

This is discussed in modules developers meeting.
2023-05-18 16:12:41 +08:00
Volodymyr Sapsai
381fcaa136 [modules] Replace -Wauto-import with -Rmodule-include-translation.
Diagnostic for `-Wauto-import` shouldn't be a warning because it doesn't
represent a potential problem in code that should be fixed. And the
emitted fix-it is likely to trigger `-Watimport-in-framework-header`
which makes it challenging to have a warning-free codebase. But it is
still useful to see how include directives are translated into modular
imports and which module a header belongs to, that's why keep it as a remark.

Keep `-Wauto-import` for now to allow a gradual migration for codebases
using `-Wno-auto-import`, e.g., `-Weverything -Wno-auto-import`.

rdar://79594287

Differential Revision: https://reviews.llvm.org/D130138
2022-07-21 17:42:04 -07:00
Iain Sandoe
af2d11b1d5 [C++20][Modules] Implement include translation.
This addresses [cpp.include]/7

(when encountering #include header-name)

If the header identified by the header-name denotes an importable header, it
is implementation-defined whether the #include preprocessing directive is
instead replaced by an import directive.

In this implementation, include translation is performed _only_ for headers
in the Global Module fragment, so:
```
module;
 #include "will-be-translated.h" // IFF the header unit is available.

export module M;
 #include "will-not-be-translated.h" // even if the header unit is available
```
The reasoning is that, in general, includes in the module purview would not
be validly translatable (they would have to immediately follow the module
decl and without any other intervening decls).  Otherwise that would violate
the rules on contiguous import directives.

This would be quite complex to track in the preprocessor, and for relatively
little gain (the user can 'import "will-not-be-translated.h";' instead.)

TODO: This is one area where it becomes increasingly difficult to disambiguate
clang modules in C++ from C++ standard modules.  That needs to be addressed in
both the driver and the FE.

Differential Revision: https://reviews.llvm.org/D128981
2022-07-10 11:06:51 +01:00