Em 06-03-2010 01:16, Gustavo Sverzut Barbieri escreveu:
> On Fri, Mar 5, 2010 at 6:45 PM, Enlightenment SVN
> <no-re...@enlightenment.org> wrote:
>> #include <Elementary.h>
>> #include <Ecore_X.h>
>> @@ -45,7 +44,6 @@
>> extern char * url_post;
>> extern char * url_friends;
>> extern int debug;
>> -extern GConfClient *conf_client;
>
> try to remove these variables shared amongst modules... keep the
> variables as "static" in their own modules and provide higher level
> ways to access them from others... you'll avoid lots of programming
> bugs by doing that.
>
>> void toggle_fullscreen(Eina_Bool new_fullscreen) {
>> - char * fs_icon;
>> + char * fs_value=NULL;
>> + int res=0;
>>
>> - if(new_fullscreen)
>> - fs_icon="arrow_down";
>> - else
>> - fs_icon="arrow_up";
>> -
>> elm_win_fullscreen_set(win, new_fullscreen);
>>
>> + res = asprintf(&fs_value, "%d", new_fullscreen);
>> + if(res != -1)
>> + eet_write(conf, "/options/fullscreen", fs_value,
>> strlen(fs_value), 0);
>> + if(fs_value) free(fs_value);
>
> no need to init fs_value to NULL, just do the free(fs_value from inside the
> if.
>
> Also, I'd rather use eet data descriptors and avoid such things...
> this would be much easier and you'd not have to do the
> parsing/serializing yourself. Just see the monster usage of it in
> e_config.h/e_config.c or some simpler application such as:
>
> http://bazaar.launchpad.net/~launch-lite/launch-lite-proj/netbook-launcher-efl/annotate/head:/src/launcher/conf.c
>
> See example at
> http://trac.enlightenment.org/e/browser/trunk/eet/src/lib/Eet.h#L866
> (no idea why this is not getting into generated doxygen :-( )
>
> You'll see that you can use this for caching last tweets as well.
> ...
>
>> + data = eet_read(conf, "/options/max_messages", &size);
>> + if(data) {
>> + mm = strtol(data, NULL, 10);
>> + if(mm > INT_MIN && mm < INT_MAX)
>> + MAX_MESSAGES = (int)mm;
>> + free(data);
>> + data = NULL;
>> + }
>
> same... just' stick everything in your "config" structure, all the
> options. Then you create your data descriptor for it and just do a
> simple eet_data_read()... get everything there! Then you can go and
> validate if the ranges are valid, but that's super easy, no need and
> go strtol() and friends.
>
> Well, it's a pitty that documentation is somehow broken and not
> listing all the doxygen text (as in Eet.h)... and we really need a
> simple eet example. If you could help there, just create
> src/examples/ and add 2 files there basic.c (with things like you did
> here), and descriptors.c (with thing you can learn from Eet.h). You
> can learn and test things there, people can review it and others can
> benefit from it in future.
>
> BR,
Sure, will check it out, thanks for the tip. Documentation is really a
barrier entry to EFL, I wish it was a little better.
I don't mind looking at .h files as documentation if they're adequately
documented :)
Rui
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
