[PATCH] paginate.py SA support, better sorting and bug fix for compund widgets

[randall@…]

Included is a patch for paginate.py which does the following thing:

1. Fix problem with form submission from compound widget.
2. Add support for SQLAlchemy using SelectResults? in sqlalchemy.ext.selectresults
3. Add multi-field sorting so that all fields are sorted in reverse of the order chosen.

This is based on this discussion:

 http://groups.google.com/group/turbogears/browse_thread/thread/834a0e04ac9425a5

[randall@…]

patch for paginate.py

[randall@…]

fixes sorting bug and clean up log.info entries

[randall@…]

API fix for reverse_order

[randall@…]

adds compatibility for tg.paginate.reversed attribute

[randall@…]

fix issue with no default order specified

[sbr77]

better form data fix

[sbr77]

patch to fix form data only (no SA or sorting changes)

[sbr77]

paginate.py.diff.1a is a very simple patch that just fixes the form data submission issues. No SA support or multi-field ordering. If you're just having trouble with forms, this is the patch I recommend because it is so simple. Otherwise use patch 6 (as of today).

[sanjay]

Feedback - after applying paginate.py.diff.6 to my 1.0b2 installation, @paginate with sqlalchemy works nicely for me.

[alberto]

A short unittest for this would be nice so we make sure we don't get regressions when fixing other things. I'll try to write one as soon as I have a chance but if someone who is using this can contribute one before I do it will probably get comitted sooner :)

Alberto

[guest]

I would gladly write tests, but have had difficulty doing so. See this discussion:

 http://groups.google.at/group/turbogears/browse_thread/thread/4380df8d5b5823ec/861d3f88d917d245?lnk=gst&q=paginate+tests&rnum=3&hl=de#861d3f88d917d245

Randall

[alberto]

paginate.py.diff.g no longer applies cleanly. I get:

(Turbogears)jal:~/src/python/checkouts/turbogears/turbogears alberto$ patch -p0 < ~/Desktop/paginate.py.diff.6 
patching file paginate.py
Hunk #2 FAILED at 4.
Hunk #3 succeeded at 30 (offset 3 lines).
Hunk #4 succeeded at 55 (offset 3 lines).
Hunk #5 succeeded at 99 (offset 3 lines).
Hunk #6 succeeded at 134 (offset 3 lines).
Hunk #7 succeeded at 183 (offset 3 lines).
Hunk #8 succeeded at 218 (offset 3 lines).
1 out of 8 hunks FAILED -- saving rejects to file paginate.py.rej

Can someone please update it? Removing [PATCH] for the meantime.

Alberto

BTW: please run svn diff > ma_patch.diff from the *root* of your svn checkout to get an unified diff that's easier to apply.

TODO: Explain this somewhere to ease the job of future contributors...

[guest]

Kevin made this change:

r2326 | kevin | 2007-01-02 09:02:15 -0600 (Tue, 02 Jan 2007) | 2 lines

makes SQLObject optional

It's very minor. I'll create a new patch.

-Randall

[guest]

new patch to work with current 1.0 branch

[alberto]

Comitted at [2404]. Thanks! :)

Alberto

[vhata]

There's a slight bug in sorting: the tg_paginate_order parameter does not get taken into account. The fix is as follows:

elif default_order and not ordering:

ordering = {default_order:[0, not default_reversed]}

• elif ordering and order:

+ if ordering and order:

sort_ordering(ordering, order)

log.debug('ordering %s' % ordering)

Patch attached.

[vhata]

In addition, the tg_paginate_reversed parameter (documented at  http://docs.turbogears.org/1.0/PaginateDecorator ) does not appear to be used at all.

I've attached a patch which adds this functionality, although it could probably be done more elegantly?

[Chris Arndt]

Cruft ticket. Paginate was seriously refactored after the patches here were created. Closing ticket as invalid.