Python-dwca-reader

-------------------

- New InvalidArchive exception. Currrently, it is only raised when a DwC-A references a metadata file that's not present in the archive.

-------------------

- Performance: an optional 'extension_to_ignores' parameter (List) can be passed to DwCAReader's constructor. In cases where an archive contains large but unneeded extensions, this can greatly improve memory consumption. A typical use-case for that would be the huge 'verbatim.txt' contained in GBIF downloads.

-------------------

- Performance: we now use core_id based indexes for extensions. There's a memory penalty, but extension file parsing is now only done once.

-------------------

- Better performance with extensions.

-------------------

- It can now open not zipped (directory) Darwin Core Archives
- More testing for Descriptor classes.
- Better respect of the standard (http://rs.tdwg.org/dwc/terms/guides/text/):
* We now support default value (\n) for linesTerminatedBy and fieldsTerminatedBy.
- Lower memory use with large archives.

-------------------

- Better performance thanks to a better architecture
- API add: brand new _ArchiveDescriptor and _SectionDescriptor
- API change: DwCAReader.descriptor is an instance of _ArchiveDescriptor (previously BeautifulSoup)
- API remove: DwCAReader.core_rowtype (use DwCAReader.descriptor.core.type instead)
- API remove: DwCAReader.extensions_rowtype (use DwCAReader.descriptor.extensions_type instead)
- API remove: DwCAReader.core_terms (use DwCAReader.descriptor.core.terms instead)