Python-dwca-reader

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

- fixed temporary directory: previously, it was always created under current dir instead of something chosen by Python such as /tmp.

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

- GBIFResultsReader is now deprecated.
- API new: DwCAReader now provides source_metadata for GBIF-like archives (previously the main perk of GBIFResultsReader)
- API change: dwca.source_metadata is an empty dict (previously: None) when the archive doesn't has source metadata
- API new: dwca.utils._DataFile is now public and renamed to dwca.files.CSVDataFile.

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

- Updated Darwin Core terms for the qualname helper (including Event Core).
- Updated dev. script (https://github.com/BelgianBiodiversityPlatform/python-dwca-reader/issues/45).

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

- API new: DwCAReader.open_included_file(relative_path).
- InvalidArchive exception is now raised when a file descriptor references a non-existing field.

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

- Support for new types of archives:

* DwCA without metafile (see page 2 of http://www.gbif.org/resource/80639), including or not a Metadata document (EML.xml). Fixes #47 and 7.
* DwCA where data fields are enclosed by some character (using fieldsEnclosedBy when the Archive provides a Metafile, autodetection otherwise). Fixes issue 53.
* Archives without a metadata attribute in the Metafile. See 51.
* Tgz archives.

- API change: SectionDescriptor => DataFileDescriptor
- API change: DataFileDescriptor.encoding => DataFileDescriptor.file_encoding
- API change: the reader previously only supported zip archives, and raised BadZipFile when the file format was not understood. It nows also supports .tgz, and throw InvalidArchive when the archive cannot be opened as .zip nor as .tgz.
- API new: DataFileDescriptor.fields_enclosed_by
- API new: DwCAReader.use_extensions

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

- Support for archives contained in a single (sub)directory. See https://github.com/BelgianBiodiversityPlatform/python-dwca-reader/issues/49