OpenAPI: add several features

[segoon]

[codecov[bot]]

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (f3da80e) 85.30% compared to head (9470b79) 85.30%. Report is 1354 commits behind head on master.

Files	Patch %	Lines
parsers/openapi.c	88.88%	1 Missing :warning:

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #3258   +/-   ##
=======================================
  Coverage   85.30%   85.30%           
=======================================
  Files         208      208           
  Lines       49507    49510    +3     
=======================================
+ Hits        42230    42233    +3     
  Misses       7277     7277           

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[masatake]

Thank you. You used yaml: as the prefix of the commit header. Could you use the name of parser instead?

OpenAPI: add several features

[segoon]

@masatake , renamed also removed 'info' parsing as it is not very informative

[masatake]

I'm reading https://spec.openapis.org/oas/latest.html slowly. I use this comment area for taking note about "What should be tagged" in an IDEAL parser.

• 4.8.2 Info Object
  • title should be tagged.
• 4.8.5 Server Object
  • url should be tagged.
  • variable should be tagged. A tag for a variable must have the url in its scope.
• 4.8.6 Server Variable Object
  • enum elements should be tagged. A tag for them must have the variable in its scope.
• 4.8.8.1 Patterned Fields
  • path should be tagged.
• 4.8.10 Operation Object
  • operationId must be tagged. It may have operation: parser specific field; its value is one of "get", "put", ...
• 4.8.11 External Documentation Object
  • url can be extracted as a reference tag.
• 4.8.12 Parameter Object
  • name should be tagged. It may have in: parser specific field. scope: field should be filled. schema may be stored to typeref: field. style is interesting. However, I'm not sure how to store it in a field of a tag entry for name.
• 4.8.13 Request Body Object This is an interesting item. However, I cannot find a good way to handle this in ctags.

(to be continued...)

[segoon]

@masatake hi! What should we do with this PR? Do you expect me to fix the conflict, or you want to e.g. split it into parts and merge the parts by yourself or...?

[masatake]

Sorry for making you wait. But could you wait for a while? I would like to finish the 'Basic' parser related task first.

[segoon]

Yeah, no problem! Thx

[masatake]

Do you expect me to fix the conflict, or do you want to e.g. split it into parts and merge the parts by yourself or...?

I would like to do merge some of the parts (and kinds). I would like you to know how I do; I will add kinds incrementally.

I will make a pull request for merging. I would like you to review the pull request. After that, could you try doing the same as I do for the rest of the parts?

[segoon]

@masatake OK, seems fair. I'll try to split it following the necessary patch "granulation".

[segoon]

@masatake friendly ping. Should I update this PR and show you for code review, or do you want to do it by yourself?

[masatake]

I'm sorry to be late. I'm thinking about how to make the Yaml parser more extensible. Can I ask you to update your PR? In that case, I would like you to do:

(1) split the commit into smaller commits; each commit deals one kind, and (2) prioritize the commits.

If you want to have a kind X in your parser strongly, please put the commit for the kind in front of the commits of the pull request.

I stopped my review when thinking about the "example" kind (See X). I cannot convince myself whether we should make tags for the examples.

If the pull request is separated into commits per kind, I can merge ones obviously useful.

[segoon]

OK, I'll try it the next week.