[Github-comments] [geany/geany] Question: adding a kind to exising parser of u-ctags (#2227)

[Masatake YAMATO]

...I think it is nothing about symbol. (the output of nm doesn't help us, e.g.)

We have to solve the issue by designing work-flows.

I would like to explain a situation causing a trouble.
 
I will receive a pull request from a volunteer extending a parser which adds a kind to the parser.
If I have to get an agreement Geany people about adding the kind, it will make timer longer for merging. It can be a bottle neck of development of u-ctags. This is MY postion.
In the other hand, Geany people expects Geany works well with newly pulled u-ctags code always.

u-ctags has TLIB test harness which runs mini-geany. TLIB checks u-ctags doesn't break what Geany wants. Actually, adding a kind in a commit is detected by TLIB. A TLIB test case was failed. In that case, what I can do? I contacted @techee, I have not got an answer yet. (I NEVER COMPLAIN about the delay.) In that case, I cannot merge the change. So I opened this issue.

I would like Geany people to allow me to add kinds to existing parsers without "ack from Geany".
The "ack from Geany" can be bottle neck of development. (I'm only talking about adding a kind.
If u-ctags breaks API, I must get  "ack from Geany" and as I will write, such breakage must be detected by TLIB test harness.


Intead, I would like Geany people to introduce acceptance test in Geany side.
The acceptance test detects the important changes in u-ctags.
The acceptance test should be run when you updates the u-ctags source tree in Geany.

The acceptance test reports something like this:
```
[result]
1 new kind `D/maroparm` is added to C parser. 
You must update the SWITCH converting to parser neutral kind.
```

u-ctags tries not breaking Geany by TLIB test cases. However, it is not enough.
I think Geany side also must have a test accepting the latest u-ctags code.

Writing such test case may be easy. ctags --list-kinds=<PARSER> or ctags --list-kinds-full=<PARSER> may help us. If you are o.k., I will write a prototype and make a pull request.

Thank you for reading.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/issues/2227#issuecomment-514477574
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.geany.org/pipermail/github-comments/attachments/20190723/e2fc401b/attachment.html>