[all-commits] [llvm/llvm-project] 427c1d: [ASTMatchers] Matchers that take enumerations args...
Nathan James via All-commits
all-commits at lists.llvm.org
Mon Apr 6 12:00:35 PDT 2020
Branch: refs/heads/master
Home: https://github.com/llvm/llvm-project
Commit: 427c1dc4f4248ec71e32a8f3832a4e6258446672
https://github.com/llvm/llvm-project/commit/427c1dc4f4248ec71e32a8f3832a4e6258446672
Author: Nathan James <n.james93 at hotmail.co.uk>
Date: 2020-04-06 (Mon, 06 Apr 2020)
Changed paths:
M clang/include/clang/ASTMatchers/Dynamic/Diagnostics.h
M clang/lib/ASTMatchers/Dynamic/CMakeLists.txt
M clang/lib/ASTMatchers/Dynamic/Diagnostics.cpp
A clang/lib/ASTMatchers/Dynamic/Marshallers.cpp
M clang/lib/ASTMatchers/Dynamic/Marshallers.h
M clang/unittests/ASTMatchers/Dynamic/ParserTest.cpp
Log Message:
-----------
[ASTMatchers] Matchers that take enumerations args provide hints with invalid arguments
Summary:
This adds support for giving hints when using dynamic matchers with enum args. Take this query, I couldn't figure out why the matcher wasn't working:
(Turns out it needed to be "IntegralToBoolean", but thats another bug itself)
```
clang-query> match implicitCastExpr(hasCastKind("CK_IntegralToBoolean"))
1:1: Error parsing argument 1 for matcher implicitCastExpr.
1:18: Error building matcher hasCastKind.
1:30: Incorrect type for arg 1. (Expected = string) != (Actual = String)
```
With this patch the new behaviour looks like this:
```
clang-query> match implicitCastExpr(hasCastKind("CK_IntegralToBoolean"))
1:1: Error parsing argument 1 for matcher implicitCastExpr.
1:18: Error building matcher hasCastKind.
1:30: Unknown value 'CK_IntegralToBoolean' for arg 1; did you mean 'IntegralToBoolean'
```
There are no test cases for this yet as there simply isn't any infrastructure for testing errors reported when parsing args that I can see.
Reviewers: klimek, jdoerfert, aaron.ballman
Reviewed By: aaron.ballman
Subscribers: aaron.ballman, dexonsmith, mgorny, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D77499
More information about the All-commits
mailing list