[PATCH] D154014: [SpecialCaseList] Use Globs instead of Regex
Fangrui Song via Phabricator via cfe-commits
cfe-commits at lists.llvm.org
Fri Jun 30 00:13:08 PDT 2023
MaskRay added a comment.
In D154014#4461076 <https://reviews.llvm.org/D154014#4461076>, @rupprecht wrote:
> In D154014#4457883 <https://reviews.llvm.org/D154014#4457883>, @MaskRay wrote:
>
>>> This is a breaking change since some SCLs might use .* or (abc|def) which are supported regexes but not valid globs. Since we have just cut clang 16.x this is a good time to make this change.
>>
>> My user has some ignore lists, but there is no `^[a-z]+:.*\(` or `^[a-z]+:\.` occurrence, so this change is likely safe for us.
>
> I think I'm looking at the same lists, and I see plenty of `.*` in the sanitizer exclusion lists. Also a few cases of `(abc|def|...)`. IIUC, those would both be broken by this -- instead of `.*` meaning "any character any number of times" (regex) it would mean "dot followed by any number of characters" (glob), right? And the `(abc|...)` would just be that literal text, not matching the individual parts?
Sorry that my regex use was incorrect. We do have some `src:x/a.pb.*` style rules, which are technically misused. `src:x/a.pb.*` can match possibly unintended files like `x/axpbx`.
To match just glob `x/a.pb.*` files, the rule should be written as `src:x/a\.pb\.*` (`*` instead of `.*`).
> If my understanding of that is correct, I don't think this is a good change -- there's possibly plenty of configs out there that assume this is regex, and there doesn't seem to be sufficient motivation to just break those. But I can see that globs are useful for the examples posted in the patch description. Is it possible to have some middle ground, e.g. default to regex but allow a config at the top of sanitizer lists to interpret future patterns as globs instead?
I think the main breakages are:
- patterns using `(` and `)`. Our users don't have such patterns.
- patterns using `\.` to mean a literal `.`, e.g. `src:a\.c`. Our users don't use `\`. Such a pattern needs to changed to `src:a.c` (which used to match other files like `axc`)
- patterns like `src:a.pb.*` that match other files like `src:axpbx.c`. This pattern can be changed to glob `src:a?pb?*`.
I think all of these are likely uncommon. While `\.` is technically the correct way to match a literal `.` before this patch, most people likely just use `.`.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D154014/new/
https://reviews.llvm.org/D154014
More information about the cfe-commits
mailing list