Fwd: Indexing c/c++ comments and text files

[Michael Sullivan]

This is a message that was originally started on the help list, but it was 
recommended that the discussion move to the bug list as it contains a request 
for two features:

Comment Indexing
----------------
Comments in files parsed by the c/c++ parser (as identified in the langmap) are 
ignored.  I'm wondering if it would be possible to add a command line option to 
not skip the symbols in comments.

Text File Indexing
------------------
Text files that potentially contain useful things to search for (e.g. MIB 
files) don't appear to have a parser (as identified in the langmap).  I'm 
wondering if it would be possible to add a text parser that could be used to 
simply index every word in those files.

Any thoughts?

Begin forwarded message:

> From: Michael Sullivan <address@hidden>
> 
> Did anyone have any thoughts about adding these two features -- I think 
> everyone could really benefit from them.
> 
> On 19-Jul-2010, at 9:55 PM, Michael Sullivan wrote:
> 
>> I only cited id-utils in my first message so that you could see an example 
>> where another tags system had a capability that appears to be missing 
>> natively in global.  However, I'm not actually trying to generate an 
>> id-utils database -- I consider the gtags/global database, tools, and output 
>> to be superior.
>> 
>> Instead, I have two different problems that can be potentially solved in 
>> different ways:
>> 
>> Comments
>> --------
>> ...
>> 
>> Text Files
>> ----------
>> ...
>> 
>> On 19-Jul-2010, at 9:24 PM, Shigio YAMAGUCHI wrote:
>> 
>>> You can use the functions of id-utils by the -I option of gtags and global.
>>> Is it insufficient?
>>> 
>>>> I would like to instruct global/gtags to index c/c++ comments (and 
>>>> potentially other text files).  For example, the following c++ comment 
>>>> text isn't currently indexed by global/gtags, but I would like it to be:
>>>> 
>>>> // SOME_IMPORTANT_ENUM_1
>>>> "Some corresponding text 1",
>>>> 
>>>> // SOME_IMPORTANT_ENUM_2
>>>> "Some corresponding text 2",
>>>> 
>>>> etc.
>>>> 
>>>> The comment text refers to symbols that are indexed elsewhere and I want a 
>>>> cross-reference search to return references to the comment lines above.
>>>> 
>>>> When I used to use id-utils, I was able to achieve this by telling it that 
>>>> c/c++ files were actually text (e.g. set those extensions up to use the 
>>>> text parser in id-utils' langmap).  Although I found a langmap setting in 
>>>> the .globalrc file, it doesn't appear to have a "text" parser.
>>>> 
>>>> Is there a possibility of adding something like this?