SORT precedence when specifying sort order

[GoogleCodeExporter]

The documentation says you can use:

doFind(..., SORT=["field":"<"]) which is invalid syntax.

Using a dict as the SORT argument works sometimes.

e.g. using SORT={"AAAA":">", "BBBB":">"} could generate a URL like:

http://...&-sortfield.1=AAAA&-sortorder.1=descend&-sortfield.2=BBBB&-sortorder.2
=descend&...&-find

but it could just as well generate the URL with the sort precedences switched 
like this:

http://...&-sortfield.1=BBBB&-sortorder.1=descend&-sortfield.2=AAAA&-sortorder.2
=descend&...&-find

The code defaults SORT to an empty list, but if you were to pass in 
SORT=["AAAA", "BBBB"] or something along those lines, then the following code 
would raise a TypeError:

        for key in SORT:
            self._addSortParam(key, SORT[key])

Original issue reported on code.google.com by esiot...@gmail.com on 15 Dec 2010 at 7:49

[GoogleCodeExporter]

Sorry, title should not mention sort order.

Original comment by esiot...@gmail.com on 15 Dec 2010 at 7:50

[GoogleCodeExporter]

The problem is in this two lines. Key-value pairs in a python dictionary are 
unordered by definition (a dictionary is a set, not a sequential container). So 
precedence in order criterias is not guarantied. To solve the bug, simply 
change this lines in FMServer.py by:

    for i in range(0, len(SORT), 2):
        self._addSortParam(SORT[i], SORT[i+1])

and then in your code, write the find as a list of pairs field operation:

    doFind(..., SORT['field1', '<', 'field2', '<'])

I hope this can help.

Original comment by e.marti....@gmail.com on 10 Dec 2013 at 9:18

[GoogleCodeExporter]

Thanks for the reply, but it is almost 3 years too late ;)  Maybe it will help 
someone else.  I decided I could not rely on pyfilemaker in production and made 
other arrangements.

Original comment by esiot...@gmail.com on 10 Dec 2013 at 3:56

[GoogleCodeExporter]

There is a way of solving this problem without modifying FMServer.py. 

All you need is the OrderedDict class that is in the collections module. In 
your code, you must write:

#from collections import OdreredDict as odict
...
doFind(..., SORT=odict([('field1', '<'), ('field2', '>'), ('field3', '>')]))

This works very fine now.

Original comment by e.marti....@gmail.com on 19 Aug 2014 at 9:26