noelniles / docket Goto Github PK

Because we're using iterparse every time one of the Explorer methods goes through the tree the self.context iterator is exhausted. This means that each method has to reset the iterator by calling self.__init__(). There should be a cleaner way.

some ideas:

• Don't use an iterator.
  • pros: simpler, no iterator resetting
  • cons: eats up memory, impossible at certain sizes
• Keep it like it is.
  • pros: don't have to do anything
  • cons: inefficient
• Some kind of cache that only get created once.
  • pros: only happens once, quick access
  • cons: complicated, when to put what in the cache?
• Anything else...

It will probably be useful to add the license boilerplate at this point.

considirations:

• What license?
  • GPL3?
• Does each file need a license header?
• What format to use for the header?

Make sure filter() has the correct behavior. Add a test to docket_test.py and make sure it passes when running nosetests.

Make sure that count tags has the correct behavior.

Pick a better name for filter().

suggestions

• text_filter,
• filter_by_tag,
• get_all_the_text_from_this_tag...