Details
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
llvm/include/llvm/Support/CommandLine.h | ||
---|---|---|
1972 | This seems confusing. Relative=true, CurrentDir=none: we read ./x/y/b Seems like it would be clearer to just say relative paths are relative to this directory, so the last one would be foo/y/b. Actually looking at the code, I think that's what you implemented? | |
llvm/lib/Support/CommandLine.cpp | ||
1105 | This is @/BasePath I think (args are segments). (I think this means you were running the wrong tests) | |
1183 | this is subtle, consider a comment: CurrentDir is only relevant for "top-level" expansions || !RelativeNames, but nested ones always have absolute paths if RelativeNames so CurrentDir is ignored. |
- Address comments
- Change helper to take absolute paths, instead of taking a currentdir
- Update tests to make use of FS
- Add a new test for container-relative filenames
Build result: pass - 60441 tests passed, 0 failed and 726 were skipped.
Log files: console-log.txt, CMakeCache.txt
llvm/lib/Support/CommandLine.cpp | ||
---|---|---|
1183 | did better, changed ExpandResponseFile to take FName as an absolute path, PTAL |
This seems confusing.
Suppose we have @x/a, which includes @y/b
Relative=true, CurrentDir=none: we read ./x/y/b
Relative=true, CurrentDir=foo: we read foo/x/y/b
Relative=false, CurrentDir=none: we read ./y/b
Relative=false, CurrentDir=foo: we read... ./y/b?
Seems like it would be clearer to just say relative paths are relative to this directory, so the last one would be foo/y/b.
Actually looking at the code, I think that's what you implemented?