testrunner: more Tokenizer related refactoring#7440
testrunner: more Tokenizer related refactoring#7440firewave merged 4 commits intocppcheck-opensource:mainfrom
Tokenizer related refactoring#7440Conversation
|
This does not seem like much of a cleanups (especially because of the added noise in the Makefile) but this is about about getting rid of redundant stuff which complicates things as I try to untangle this language mess. As mentioned before the application code is fine but all these custom approaches in the tests complicate things. This stuff really makes my head spin and every time I think I finally cracked this I reach another point where it seems like I have to rip it out all at once - which I want to avoid. |
| ASSERT(startsWith(errout_str(), "[test.cpp:6]: (debug) Failed to parse 'using C = S < S < S < int")); | ||
| TODO_ASSERT(startsWith(errout_str(), "[test.cpp:6]: (debug) Failed to parse 'using C = S < S < S < int")); |
There was a problem hiding this comment.
@chrchr-github This was producing an error because the code contained the preprocessed and non-preprocessed source because it was tokenized twice (I tracked it back to the introduction and it has always been that way). I think this is currently fine because it was about a hang/crash but if we expect to test for the error we might need to add another test.
There was a problem hiding this comment.
What is errout_str() after the change? I agree the output was not important, since the issue was a hang.
There was a problem hiding this comment.
That even seems like progress. But then I don't understand why the refactoring made the debug message go away.
If it's related to what you wrote above, just assert on the empty errout_str.
There was a problem hiding this comment.
But then I don't understand why the refactoring made the debug message go away.
See previous comment.
This was producing an error because the code contained the preprocessed and non-preprocessed source because it was tokenized twice
I was asking because I think we should have a test which actually detected the debug message.
firewave
force-pushed
the
toklist-zzz
branch
2 times, most recently
from
06a5d32 to
55dc79f
Compare
firewave
changed the title
Tokenizer related cleanupsTokenizer related refactoring
This comment was marked as resolved.
This comment was marked as resolved.
3 participants
No description provided.