Le 10/05/2022 à 01:58, John Wheeler a écrit :
On 5/9/22 17:43, Jean Abou Samra wrote:
Le 10/05/2022 à 00:37, John Wheeler a écrit :
The TAGS file structure is simple enough, and I agree having
only one place to maintain name mangling logic is good.
But, I am not following you on the reference to
'out/bin/lilypond -ddump-tags'. Is -ddump-tags a command line option
to the lilypond executable? Please explain.
You would add it as a new option. See `lilypond -dhelp` and
`lilypond -dhelp-internal` for lists of currently available
-d options (non-"-d" options are for fundamental setup like
--loglevel, all somewhat advanced options are under -d).
See also the Usage manual. You would add that option in
lily.scm around line 382 and implement it at the beginning
of lilypond-main from lily.scm.
Ok, I understand. This would cover the needed references for
functions defined in LY_DEFINE and friends, but would not
address the other standard C++ style references that are
found by the etags program. I am very reluctant to attempt
to re-implement in LilyPond something that available in the
existing utility.
Sure, we'd use the standard solution for those.
And, if I add -ddump-tags to lilypond, the
make system will still need to be adapted to merging the two
sources of TAG references.
Should be easy enough, with etags --append.
I have a merge request prepared that implements
'make TAGS' using the existing lilypond makefile system.
It does rely on a single python script to mangle the C++
identifiers.
Jean, David, Jonas, Unless one of you tell me not to do so,
I will submit it as is for review and consider Jean's
suggestion as a pre-planned improvement.
Fine. We can continue the discussion on the MR.