Write simpletests

I've started working on this. I have a simple test that creates a feed node, configures some mapping, and consumes an XML (really more of an exercise than a test).

Next, I will create some CCK field mappings, to test individual cases. Sample feed data welcome!

FYI, while writing these tests, I came across a strange behaviour, which turned out to be a bug: #451564: Using timestamp fields in the Date mapper empties created node.

First draft of the tests for base API, CCK mapper, and start of date mapper.

"279 passes, 0 fails, and 0 exceptions"

I've done 1,3,4 per your suggestion, and removed the <type> items from the comments. Did you have specific concerns with the comments?

Added tests for Link and Taxonomy
353 passes, 1 fail, and 0 exceptions
(the failure is due to #451926: FeedAPI reports "1 existing item(s) were updated" when consuming a brand-new feed)

1) Not fixed yet - ask me tomorrow :(
2) Done - including writeFile() calls, except for the two test-cases I'm actively working on
3) Done - added prefix to filenames
4) Done
5) Done

Added start of emimage testcase - WIP. Added an ical sample file.

438 passes, 1 fail, and 0 exceptions (excluding date mapper; error as described above)

I get lots of fails now, maybe because of FeedAPI changes, further testing is needed. For example createFeed() does not work.

Re-rolled.
Mostly i needed to touch initialization parts, surely it's because of changes in feedapi and other modules.

I get repeatable HTTP 500 error executing the Date mapper test, but the other 5 pass fine:

Overall results: 450 passes, 0 fails, and 0 exceptions
FeedAPI Mapper Link: 94 passes, 0 fails, and 0 exceptions
FeedAPI Mapper Embedded Image Field: 82 passes, 0 fails, and 0 exceptions
FeedAPI Mapper Content: 88 passes, 0 fails, and 0 exceptions
FeedAPI Mapper: 114 passes, 0 fails, and 0 exceptions
FeedAPI Mapper Taxonomy: 72 passes, 0 fails, and 0 exceptions

This is with SimpleTest 6.x-2.8. Looking deeper now.

The problem with the Date test is the same as #472684: Install non-default modules one at a time to ensure proper module list is maintained during installation hooks, which isn't fixed in Core D7 yet.

You can work around the issue by declaring the order of your module requirements, without relying on the dependency system. Ie: instead of:

    @parent::setUp(
      'feedapi',
      'feedapi_node',
      'parser_simplepie',
      'parser_common_syndication',
      'feedapi_mapper',
      'content',
      'date',
      'date_api',
      'date_timezone'
    );

use

    @parent::setUp(
      'feedapi',
      'feedapi_node',
      'parser_simplepie',
      'parser_common_syndication',
      'feedapi_mapper',
      'content',
      'date_api',
      'date_timezone',
      'date'
    );

(Note the order of the last three modules). Reroll attached - this is the only change

#13, really strange that this problem did not appear at my dev environment...
All in all, i tested it, it seems that now it's both fine in my and your environment, so now it's RTBC.