== To do ==

This is a collection of things that need to be done. The order doesn't imply
any priority. These tasks may be turned into tickets and assigned to milestones.

=== Tasks in Progress ===

 * write unit tests
 * define API --> partially done
 * write documentation, tutorial etc.
 * ~~develop plugin system for custom readers/writes
   (do thorough research on existing plugin systems)~~
   registry that stores functions in a dictionary
   implemented
 * write example applications that use !PyModelData
   (there are two, with an older version though --> update)

=== Next Tasks ===

 * {{{open}}} function for format modules
 * exception handling for {{{open}}} function
 * exception handling for {{{!INCLUDE}}}
 * allow specifying of data types for non-yaml formats, i.e. each column may have a data type

=== Shorter Term Tasks ===

 * check for entries in data that are not in default
 * add property {{{require}}} for keywords that should not work with default values
 * implement more readers/writers for other formats ([http://www.pytables.org PyTables], 
   [http://en.wikipedia.org/wiki/Microsoft_Excel Excel], [http://en.wikipedia.org/wiki/DBASE dBase])
 * add checks and conversions for physical units

=== Longer Term Tasks ===

 * use ordered dict as default in YAML (makes writing back in the order of reading possible --> round-trip)
   --> patch to YAML
 * maybe allow unlimited nesting levels in data (there are only two levels now)
 * make dump aware of numpy arrays
 * allow lazy reading of large files (use generators/iterators)
 * write desktop GUI (prerequisite: round-trip)
 * write web GUI (prerequisite: round-trip)
 * make list with projects/applications/models that use !PyModeldata
 * ''idea'': access to model data via SQLite database or use [http://www.pytables.org PyTables] as intermediate storage
 * more ...