[Qgis-developer] making QgsMapToolIdentify available to plugins

[Radim Blazek]

Hi Denis,

On Thu, Feb 7, 2013 at 11:30 AM, Denis Rouzaud <denis.rouzaud at gmail.com> wrote:
> Hi Radim,
>
> I just finished the work thanks to Jef's debugging ability!
>
> Here is the pull request
> https://github.com/qgis/Quantum-GIS/pull/421

pushed to master in 4e64d72.

Thanks
Radim

> The identify method now returns directly the IdentifyResults instead of
> using a private attribute.
> Also, IdentifyResult is now a single struct (no RasterResult and
> VectorResult)
>
> Everything compiles fine and I tested it.
> However, could you check that everything is ok regarding to the
> rasterChanged?
>
> Also, could you take care of accepting the pull request?
>
> Thanks,
>
> Denis
>
>
>
> On 02/07/2013 11:24 AM, Radim Blazek wrote:
>
> Hi,
> sorry, I do not use IRC usually.
>
> I would like to return to work in identify tool areas next week:
>   - copy feature from context menu
>   - better tree hierarchy for rasters
>   - propagate errors to user - probably just error messages as texts in tree
>   - instead of all that signals on format change just listen in action
> and then call QgsMapToolIdentify and dialog functions
>
> Please let me know when you finish your work in identify tool so that
> we avoid conflicts.
>
> Radim
>
> On Thu, Feb 7, 2013 at 10:04 AM, Denis Rouzaud <denis.rouzaud at gmail.com>
> wrote:
>
> Hi Radim,
>
> I have already merged the changes to return directly the results.
>
> I am currently making a single struct for the results. I have a problem with
> python files, an error in compilation I am trying to solve for some time now
> without any success...
> Would you have some time to give me hand?
>
> Are you on IRC?
>
> Thanks
>
> Denis
>
>
>
> On 02/06/2013 05:07 PM, Radim Blazek wrote:
>
> Also, your changes to the raster result structure made me think that we
> could use a single struct for both vector and raster result.
> It might be easier to look into the results.
> Any opinion on this? Nathan?
>
> I believe, that It would be better to pass results to the dialog in
> the original hierarchical structure instead of splitting them to
> individual results and then reconstruct the tree searching for the
> layer it should go into. For rasters the logical structure may vary
> form one level for raster values (I would prefer to put band values
> directly into layer item instead of adding additional level "feature")
> to four levels for WMS with layer-sublayer-featuretype-feature.
>
> Maybe some parts of QgsMapToolIdentify::identifyVectorLayer and
> QgsMapToolIdentify::identifyRasterLayer, those related to creation the
> tree in dialog should better be moved to QgsIdentifyResultsDialog?
> I'll possibly add QgsRasterIdentifyResult and the dialog could work
> directly with that or to use current QList<QVariant> returned from
> raster identify().
>
> yes,
>
>
>
>