[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-03 Thread Jiří Techet via Github-comments
Merged #4013 into master.

-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#event-15095959409
You are receiving this because you are subscribed to this thread.

Message ID: 

[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-03 Thread Enrico Tröger via Github-comments
@eht16 approved this pull request.





-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#pullrequestreview-2411854733
You are receiving this because you are subscribed to this thread.

Message ID: 

[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-02 Thread Jiří Techet via Github-comments

> The warning disappears when we use tm_tag_local_var_t for 
"Variables" member in group_MESON.
Goto still works this way but I'm not sure if this is a valid 
"fix" or this might have other side effects?

Yes, this is a valid fix, it has to appear in some group (in fact, it could be 
added to any of the groups, since local variables aren't displayed, it 
doesn't matter).

I added this warning to Geany so we don't forget to update groups when new 
kinds are introduced in which case symbols wouldn't be displayed in the 
sidebar.

-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#issuecomment-2453061915
You are receiving this because you are subscribed to this thread.

Message ID: 


[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-02 Thread Jiří Techet via Github-comments
@techee pushed 1 commit.

7b8fa68b582989e1c3ee3435f27f1db65b4aebf8  Add tm_tag_local_var_t into the 
"Variables" group

-- 
View it on GitHub:
https://github.com/geany/geany/pull/4013/files/e9a73c411ed3122570ebc9226411d0733c1a312e..7b8fa68b582989e1c3ee3435f27f1db65b4aebf8
You are receiving this because you are subscribed to this thread.

Message ID: 



[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-02 Thread Enrico Tröger via Github-comments

Cool.

Now I get:
```
process:641255): Tagmanager-WARNING **: 11:35:25.312: Not all tag types mapped 
to symbol tree groups for Meson
```

The warning disappears when we use `tm_tag_local_var_t` for 
"Variables" member in `group_MESON`.
Goto still works this way but I'm not sure if this is a valid 
"fix" or this might have other side effects?

-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#issuecomment-2452949221
You are receiving this because you are subscribed to this thread.

Message ID: 


[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-02 Thread Enrico Tröger via Github-comments

If this is not valid and causes more troubles, I think, it is ok to just show 
variables in the symbol tree as before.

-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#issuecomment-2452949396
You are receiving this because you are subscribed to this thread.

Message ID: 


[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-01 Thread Jiří Techet via Github-comments

> I'm not sure either, maybe...

Thanks, I've just used that.

> the others are just to be filled and "Tests" left empty 
(assuming most Meson files won't have tests :D)

Yes, we should do our best to discourage users from writing those :-).

> or mapped to local variables which aren't shown in the sidebar so 
symbol goto still works for them

I did just that, but there was a minor problem I had to fix - see the second 
commit. Goto works fine now and variables aren't shown in the sidebar.


-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#issuecomment-2452577766
You are receiving this because you are subscribed to this thread.

Message ID: 


[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-01 Thread Jiří Techet via Github-comments
@techee pushed 2 commits.

d7bc7960355a7dc06cb46a6df4c728146c5160a8  Use Forth and Meson optlib-based 
ctags parsers
e9a73c411ed3122570ebc9226411d0733c1a312e  Fix current_scope value passed to 
tm_workspace_is_autocomplete_tag()

-- 
View it on GitHub:
https://github.com/geany/geany/pull/4013/files/7e02e1d30582c7506469fced1c8e024585322cfd..e9a73c411ed3122570ebc9226411d0733c1a312e
You are receiving this because you are subscribed to this thread.

Message ID: 



[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-01 Thread Enrico Tröger via Github-comments

> > why did you choose against icons? Maybe we can try to match some of 
the Makefile mapping, e.g. "targets" seem to have a similar meaning.
> 
> It's not because of the icons themselves but because there's no 
clear semantic mapping of the kinds provided by meson and icon names (which are 
tailored to programming languages) - so I wasn't sure what to use. If you 
have some suggestions which icons to use for which kinds, no problem to add 
them.

I'm not sure either, maybe:
```c
static TMParserMapGroup group_MESON[] = {
{N_("Projects"), TM_ICON_CLASS, tm_tag_namespace_t},
{N_("Modules"), TM_ICON_NAMESPACE, tm_tag_package_t},
{N_("Variables"), TM_ICON_NONE, tm_tag_variable_t},
{N_("Subdirs"), TM_ICON_OTHER, tm_tag_macro_t},
{N_("Build Targets"), TM_ICON_METHOD, tm_tag_other_t},
{N_("Custom Targets"), TM_ICON_METHOD, tm_tag_enumerator_t},
{N_("Benchmark Targets"), TM_ICON_METHOD, tm_tag_function_t},
{N_("Run Targets"), TM_ICON_METHOD, tm_tag_member_t},
{N_("Tests"), TM_ICON_NONE, tm_tag_field_t},
};
```
- Modules -> TM_ICON_NAMESPACE: as modules here are somewhat similar to 
imports in Python, so use the same icon as in Python
- *Targets -> TM_ICON_METHOD: same semantic as in Makefile, so use the same 
icon as there
- the others are just to be filled and "Tests" left empty (assuming 
most Meson files won't have tests :D)

But as you see, this is also more or less arbitrary. We don't need to add 
them at all.

> > is it useful to list the variables as symbols? This can get many 
quickly, I tested with GLib's meson.build and there the list is pretty long.
> 
> OK, these can be removed (or mapped to local variables which aren't 
shown in the sidebar so symbol goto still works for them).

:+1: 

-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#issuecomment-2452284179
You are receiving this because you are subscribed to this thread.

Message ID: 


[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-01 Thread Jiří Techet via Github-comments

> why did you choose against icons? Maybe we can try to match some of the 
Makefile mapping, e.g. "targets" seem to have a similar meaning.

It's not because of the icons themselves but because there's no clear 
semantic mapping of the kinds provided by meson and icon names (which are 
tailored to programming languages) - so I wasn't sure what to use. If you 
have some suggestions which icons to use for which kinds, no problem to add 
them.

> is it useful to list the variables as symbols? This can get many quickly, 
I tested with GLib's meson.build and there the list is pretty long.

OK, these can be removed (or mapped to local variables which aren't shown 
in the sidebar so symbol goto still works for them).

-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#issuecomment-2452257566
You are receiving this because you are subscribed to this thread.

Message ID: 


[Github-comments] Re: [geany/geany] Use Forth and Meson optlib-based ctags parsers (PR #4013)

2024-11-01 Thread Enrico Tröger via Github-comments

Cool!
Tested with Meson and looks great.

Two minor questions:
- why did you choose against icons? Maybe we can try to match some of the 
Makefile mapping, e.g. "targets" seem to have a similar meaning.
- is it useful to list the variables as symbols? This can get many quickly, I 
tested with GLib's meson.build and there the list is pretty long.

-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/4013#issuecomment-2452230383
You are receiving this because you are subscribed to this thread.

Message ID: