Re: Failed to Parse Template and Corrupted Tags

[Shigio YAMAGUCHI]

Hi,

You are right.

After checking about that, it was understood that there are two problems.

1. Treatment of NULL file

A NULL file (file size == 0) is treated as a correct db file.

This is the cause of the inappropriate message ('...seems older format').

It should be treated as a corrupted file (EFTYPE error).

2. Another inappropriate message

Currently, when a tag file is corrupted, you will see the following message.

global: GTAGS not found.

This isn't also appropriate. I think it should be like the following.

global: <full path>/GTAGS seems to be corrupted.

What do you think?

This is an example:

-------------------------------------------------------------------------------

Current GLOBAL (6.5.4)

-------------------------------------------------------------------------------

$ echo 'main(){}' >main.c

$ gtags

$ global -x main

main                1 main.c           main(){}

$ cp /dev/null GTAGS

$ global -x main

global: /tmp/test/GTAGS seems older format. Please remake tag files.

$ echo aaa >GTAGS

$ global -x main

global: GTAGS not found.

-------------------------------------------------------------------------------

|

v

-------------------------------------------------------------------------------

New GLOBAL (6.5.5?)

-------------------------------------------------------------------------------

$ echo 'main(){}' >main.c

$ gtags

$ global -x main

main                1 main.c           main(){}

$ cp /dev/null GTAGS

$ global -x main

global: /tmp/test/GTAGS seems to be corrupted.

$ echo aaa >GTAGS

$ global -x main

global: /tmp/test/GTAGS seems to be corrupted.

-------------------------------------------------------------------------------

Thank you for making me notice the basic bug.

Regards,

Shigio

2016-09-05 16:16 GMT+09:00 Gustaf Waldemarson <address@hidden>:

Hello,
Thanks for fixing it so quickly! Keep up the good work! :)

Perhaps it might be an Idea to update the error message as well? The
'GTAGS seems older format' feels like an odd response to a possibly
corrupted tags file. Perhaps 'Could not read GTAGS file properly. Were
tags created successfully with a non-zero exit code?' might be a bit
more appropriate?

Regards,
Gustaf

2016-09-03 11:27 GMT+02:00 Shigio YAMAGUCHI <address@hidden>:
> Hi,
> You are right. I have rewritten the source code slightly
> using 'goto finish' method.
>
> RCS file: /sources/global/global/libparser/Cpp.c,v
> retrieving revision 1.22
> diff -r1.22 Cpp.c
> 239,240c239,243
> < } else if (c == EOF)
> < die("failed to parse template [+%d %s].", savelineno, curfile);
> ---
>> } else if (c == EOF) {
>> if (param->flags & PARSER_WARNING)
>> warning("failed to parse template [+%d %s].", savelineno, curfile);
>> goto finish;
>> }
> 583a587
>> finish:
>
>
> By the way, in the current GLOBAL, each parser can stop
> the program itself. In that case, the tag files get corrupted.
> When gtags ends with status != 0, you should not use them.
>
> Regards
> Shigio
>
>
> 2016-09-02 16:47 GMT+09:00 Gustaf Waldemarson
> <address@hidden>:
>>
>> Hello Global Maintainers,
>>
>> I seem to have run into one or possibly two bugs while trying out
>> Global:
>>
>> Recipe:
>>
>> 1. Create a new directory for testing:
>>    - mkdir debug
>>
>> 2. Create a new file in the directory e.g. `main.cpp' with the contents:
>>
>>     #include <iostream>
>>
>>     #define TEST_MACRO struct test<
>>
>>     int main(void)
>>     {
>>         std::cout << "hello, world!" << std::endl;
>>         return 0;
>>     }
>>
>> 3. Generate tags:
>>    - `gtags'
>>    - Bug 1: Parser error: `gtags: failed to parse template [+3
>>      ./main.cpp].'
>>
>> 4. Trying to use the generated tags then leads to possibly different
>>    bug:
>>    - `global -u'
>>    - Bug 2: `gtags: /home/guswal01/git/debug/GTAGS seems older
>>      format. Please remake tag files.'
>>
>> This occured on `Global 6.5'.4 when built on Ubuntu 15.10 using the
>> default `sh reconf.sh && ./configure' settings.
>>
>> In my opinion, failing to parse some odd C++ file is usually fine, but
>> in in this it seems like the tag files get corrupted somehow.
>>
>> _______________________________________________
>> Bug-global mailing list
>> address@hidden
>> https://lists.gnu.org/mailman/listinfo/bug-global
>
>
>
>
> --
> Shigio YAMAGUCHI <address@hidden>
> PGP fingerprint: D1CB 0B89 B346 4AB6 5663  C4B6 3CA5 BBB3 57BE DDA3

--

Shigio YAMAGUCHI <address@hidden>

PGP fingerprint: D1CB 0B89 B346 4AB6 5663  C4B6 3CA5 BBB3 57BE DDA3