Re: symbolic links, unreadable files, skip

[KAICHO]

Hi, 

Please consider regarding device files too if possible.
Current global could stop to handle these files in case
the directory-tree has block/character device files in it.

Thanks,
KAICHO




----- Original Message -----
>From: Shigio YAMAGUCHI <address@hidden>
>To: Marcus Harnisch <address@hidden> 
>Cc: Jason Hood <address@hidden>; "address@hidden" <address@hidden>
>Date: 2015/8/29, Sat 08:18
>Subject: Re: symbolic links, unreadable files, skip
> 
>
>OK. I accept your suggestion.
>
>
>Is --skip-unreadable OK for the option name?
>Please tell me if you have better candidates.
>
>
>If the --skip-unreadable option is specified, gtags does readable check
>of source files before passing them to the parsers. The check is done using
>access(path, R_OK).
>
>
>If a file is unreadable, then gtags ignores it.
>Ignoring means the followings:
>
>
>(1) Print warning message.
>(2) The file is not added to GPATH.
>(3) The file is not passed to the parser.
>
>
>Parsers are as before. Procedure die() and warning() are granted to
>plug-in parsers. So, gtags cannot force them to continue the jobs.
>But they will not stop their job because of permission error, because
>such files are not passed.
>
>
>What do you think?
>
>
>Regards
>Shigio
>
>
>
>
>2015-08-28 16:14 GMT+09:00 Marcus Harnisch <address@hidden>:
>
>On Fri, Aug 28, 2015 at 4:22 AM, Jason Hood <address@hidden> wrote:
>>
>>On 28/08/2015 10:26, Shigio YAMAGUCHI wrote:
>>>> Unlike that, an unreadable file means that something critical has
>>>> occurred. In that case, gtags stop its job, and give the user a chance
>>>> to check the disk, permission and etc.
>>>
>>
>>
>>Fair enough. That's why I am not asking to make this the default behavior, 
>>but either a config value or a command line option, which demotes the 
>>error&die to a warning.
>>
>>
>>(and something's gone wrong with skip, although it sounds like
>>>that's impractical in this case, anyway)
>>>
>>
>>
>>It is impractical indeed. The skip rules would be huge and tedious to create 
>>in a trial and error approach.
>>
>>In case of symbolic links, skip seems to look at the name of the link only, 
>>not at the name of the link target. In my scenario, if only one directory 
>>tree is unreadable, but is pointed to by a large number of links, I'd have to 
>>create skip rules for all these links.
>>
>>
>>Cheers
>>Marcus
>>
>>
>>
>
>
>
>-- 
>
>Shigio YAMAGUCHI <address@hidden>
>PGP fingerprint: D1CB 0B89 B346 4AB6 5663  C4B6 3CA5 BBB3 57BE DDA3
>
>_______________________________________________
>Bug-global mailing list
>address@hidden
>https://lists.gnu.org/mailman/listinfo/bug-global
>
>
>