[GRASS-dev] g.mlist list multiple types from all mapsets

[Huidae Cho]

Well, we have a similar problem with the original g.list when one type has
way less maps than others. g.list vect,rast wouldn't show vector maps when
there are a couple of them while there are many raster maps. No body
complained about it so far, so it should be fine, I guess.
On Jun 8, 2014 9:05 AM, "Huidae Cho" <grass4u at gmail.com> wrote:

> For our records, I'm attaching a new g.mlist main.c, which supports
> multiple mapsets. I removed -p/-f and mapset=* mutual exclusiveness, but it
> seems to have a problem with M_do_list when there are a small number of
> maps in a previous mapset, but there are many in the current one.
>
> For example, let's say there are only 3 maps in mapset a, but 100 in
> mapset b. M_do_list(.., a) would be scrolled up by M_do_list(.., b) because
> the pager doesn't pause for mapset a, but does for mapset b. Effectively,
> g.mlist mapset=a,b looks as if it only shows maps in mapset b, but
> actually, maps in a were displayed, but scrolled up.
>
> The current trunk version doesn't have this problem because M_do_list is
> called once and it takes care of all the maps in the current search path at
> once.
>
> I think it's little more work to fix it. or just leave g.mlist as is.
>
>
>
> On Sun, Jun 8, 2014 at 8:30 AM, Vaclav Petras <wenzeslaus at gmail.com>
> wrote:
>
>>
>>
>>
>> On Sun, Jun 8, 2014 at 7:10 AM, Huidae Cho <grass4u at gmail.com> wrote:
>>
>>> Non-existent mapset fixed in r60749.
>>>
>>> Thanks. I can confirm that it works. Now we need the test for it.
>>
>> On Fri, Jun 6, 2014 at 12:18 PM, Vaclav Petras <wenzeslaus at gmail.com>
>> wrote:
>>
>>>
>>> By the way, is multiple for mapset a planned feature? Or it is too much?
>>> Pattern for mapset seems too much for sure.
>>>
>>
>>  On Sun, Jun 8, 2014 at 7:10 AM, Huidae Cho <grass4u at gmail.com> wrote:
>>
>>> Multiple mapsets is not too much, but do we need it?
>>>
>>
>> I don't know. I asked because I just automatically tried that syntax
>> because I used `pattern="something,some_*"` syntax recently.
>>
>>
>>> Also, if we implement multiple mapsets, it would be better semantically
>>> to change "mapset=.." to "mapset=*" for all mapsets in the current
>>> location. Martin, do you mind if I change mapset=.. to mapset=*?
>>>
>>> I would say that in any case, * fits more with the other options,
>> although * is also used also for null values (besides nv and others).
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-dev/attachments/20140608/b2968c95/attachment.html>