Are you planning to check your changes in?

On Tue, May 27, 2014 at 7:55 PM, Andreas Kupries
<[email protected]>wrote:

> Using the fossil head, i.e.
>     This is fossil version 1.29 [5d4400400a] 2014-05-27 20:36:33 UTC
>
> for a git export on my clone of
>     http://core.tcl.tk/akupries/fx/home
>
> I.e.
>     fossil export --git > DUMP
>
> the resulting DUMP contains lines of the form:
>
> % grep -n ^committer DUMP| head
> 229901:committer Andreas Kupries < <[email protected]> 1367645581 +0000
> [...]
>
> Note the lone "<" between the user name and the email address in its own
> <...>.
> The git fast_import chokes on that.
>
> Looking at my 'info' the data looks ok, i.e
>
> % fx user list
> Checkout @ /home/aku/Projects/Tools/Fossil2Git/source.fossil
>
> +-----------------+-------------+------------------------------------+------------------------------+----------------+
> | Name            | Permissions | Contact                            |
> Changed                      | Notes          |
>
> +-----------------+-------------+------------------------------------+------------------------------+----------------+
> | aku             | s           | Andreas Kupries <[email protected]> |
> Wed Apr 23 22:59:44 PDT 2014 |                |
> [...]
>
>
> A quick look at
>   static void print_person(const char *zUser)
> in
>   fossil/src/export.c
>
> which I believe is most likely responsible for the broken output, did
> not show anything obvious.
>
> However the code looks to be a bit complex with searching for "<", ">",
> etc.
> I suspect I have to single-step it to understand what is going on.
>
> ... Looking a bit more I now suspect the line
>
>      zName = mprintf("%.*s", i, zContact);
>
> because if I read the previous search right the "i" points to the
> opening "<" of the mail at that point, making it part of the extracted
> zName.
>
> Yes, changing this to i-1 gets rid of the bogus "<" in the output and
> "git fast_import" accepts the DUMP file.
>
> Notes:
>
> * I have no desire to bisect the repository to see where the issue came in
> ...
>
>   ... Oh, ok, it seems to have been part of the initial implementation
> of the git export in 2011.
>
>
> http://www.fossil-scm.org/index.html/info/0f024d3e30741af3403c1c87ee5d689741e5a63c
>      It seems to trigger only for users which have email addresses in
> their contact info.
>
> * The code should possibly trim trailing whitespace from the extracted
> name.
>
>   With my small fix there is a trailing space in the name which is
> placed into the output as well, leading to two spaces between name and
> email. The fast_import is ok with that, it just offends my tastes re
> data formatting.
>
>
>
> --
> Andreas Kupries
> Senior Tcl Developer
> Code to Cloud: Smarter, Safer, Fasterâ„¢
> F: 778.786.1133
> [email protected]
> http://www.activestate.com
> Learn about Stackato for Private PaaS: http://www.activestate.com/stackato
>
> EuroTcl'2014, July 12-13, Munich, GER -- http://www.eurotcl.tcl3d.org/
> 21'st Tcl/Tk Conference: Nov 10-14, Portland, OR, USA --
> http://www.tcl.tk/community/tcl2014/
> Send mail to [email protected], by Sep 8
> Registration is open.
> _______________________________________________
> fossil-users mailing list
> [email protected]
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>



-- 
D. Richard Hipp
[email protected]
_______________________________________________
fossil-users mailing list
[email protected]
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Reply via email to