Feature #8725R: support for maximum scale at which the layer should be simplified

[ahuarte47]

This patch adds a new setting for the simplification of vector layers:

The simplification uses a maximum scale at which generalisation should be carried out to the layer rendering options.

So e.g. if I choose 1:50000, no simplification should be carried out in scales [1:1, 1:50000].

[ahuarte47]

Hi @timlinux, this capabilty was suggested by you
Best Regards

[timlinux]

Hi

Awesome thanks! Just one note on coding practice - it is nice if you can annotate your docstrings for new methods and classes etc. with

@added in 2.2

So someone browsing the API can understand what versions their code will be supported in.

[ahuarte47]

Ok, from now on I will do it, or do you want that I write it in current code ?

[timlinux]

Hi

If you are able to do both that will be great! By the way I get may log messages like this:

When using the generalisation. The debug messages themselves make QGIS crawl to a snailspace. It would be good if you could trap for and either silently ignore or log an aggregated message.

[nyalldawson]

@timlinux that looks like http://hub.qgis.org/issues/9375

[ahuarte47]

Hi @timlinux, I am conversing with @jef-n, the commit (3305a6c) has the ogr simplification broken. I am waiting that pull #1087 to be merged (I think that it fixes these errors).

[ahuarte47]

You were faster than me :-), thanks @nyalldawson

[timlinux]

Hi

Ah ok thanks. I will merge this PR in the mean time. In the bug fix period it would be nice if you can annotate any new api additions and provide a new PR.

Thanks!

Tim

[ahuarte47]

Thanks @timlinux, the pull #1087 was merged, can you test your issue please ?

[timlinux]

Hi

I still get these errors using QGIS built against 5d02020

Regards

Tim

[ahuarte47]

Hi @timlinux, can I test your layer ?

[ahuarte47]

Hi @timlinux, If you are testing a postgis layer, then these error comes from a bug in ST_Simplify function of postgis ( http://trac.osgeo.org/postgis/ticket/1698 ). This bug collapses the geometries when the tolerance is too much big.

If you are using postgis, this patch fixes it:
#1099

[timlinux]

@ahuarte47 Yes I am using PostGIS and I can confirm that with 1099 applied the problem is resolved (and the UI is much faster to update now).

Thanks

Tim

[ahuarte47]

Thanks @timlinux, but I have one error, can you merge this pull ?
#1102

It avoids check topology state for point lasyers
Thanks!