c-based: fix to handle edge case

[jafl]

where "character" returned is a symbol (> 0xff)

[codecov]

Codecov Report

Patch coverage: 90.32% and project coverage change: -0.01% ⚠️

Comparison is base (cea67b9) 85.07% compared to head (f3a2e44) 85.06%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #3796      +/-   ##
==========================================
- Coverage   85.07%   85.06%   -0.01%     
==========================================
  Files         226      226              
  Lines       53851    53857       +6     
==========================================
+ Hits        45814    45816       +2     
- Misses       8037     8041       +4     

Files Changed	Coverage Δ
parsers/c-based.c	81.88% <83.33%> (-0.35%)	⬇️
parsers/cpreprocessor.c	95.64% <90.90%> (-0.10%)	⬇️
parsers/cxx/cxx_parser_tokenizer.c	93.02% <100.00%> (-0.10%)	⬇️

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[masatake]

Thank you.
Related to #3771. I'm working on this topic very slowly.

[jafl]

@masatake I can tackle this issue and try to finish it. Clearly, anywhere that calls cppGetc() needs to be audited to handle c > 0xff. What other pieces are there that need to be fixed?

[jafl]

I simplified my original changes, because I realized that cppUngetc is just fine with c > 0xff

[jafl]

Failed checks are system issues, not code.

[leleliu008]

Github Actions may fail for network issues, If you encounter this situation next time, just rerun it. It usually would become green after rerunning.

[jafl]

@masatake Here is my proposal for fixing the issue. I created cStringPut as a smart wrapper around vStringPut and made cppUngetc smarter. If you are ok with this, I will convert all vStringPut to cStringPut in ./cxx/cxx_parser_tokenizer.c, cpreprocessor.c, vera.c, and c-based.c. (This is the list of files from #3771)

I only changed the minimal number of calls so far, to minimize the size of the commit and make it easier to review.

[masatake]

Give me some time.

[masatake]

@jafl Looks fine.
Before merging this, I would like to merge #3804.
Could you consider updating cStringPut in your pull request after I merge #3804?

[jafl]

@masatake Sure. Let me know when you merge #3804, and I'll rebase.

[jafl]

@masatake Is it ok if I convert all vStringPut to cStringPut in all the files listed in #3771 ?

[masatake]

I merge #3804. Could you rebase this?
You must update the definition of cStringPut.

@masatake Is it ok if I convert all vStringPut to cStringPut in all the files listed in #3771 ?

At this time, could you update only vStringPut in c-based.c?
I don't have enough time to review the changes for other files at once.
I deleted c.c. Even you updated c-based.c, users get the benefit of the change.

For fixing #3771, I have been thinking about the way of handling EOF.
When I read this pull request, I recognized that thinking about EOF is insufficient. Thank you for the pull request.

I have been finding my mistakes in cpreprocess.c because of my limited understanding of the file.

[masatake]

Thank you.