Nlpiper

:pencil: Changelog:

- :hammer_and_wrench: Improvements:
- Support Python 3.7

- :books: Documentation:
- Improve general documentation, add new maintainer

:pencil: Changelog:

- :sparkles: Features:
- Add Spacy and TorchText Embeddings transformer

- :books: Documentation:
- Improve documentation and migrate links to new repository host

:pencil: Changelog:

- :hammer_and_wrench: Improvements:
- Allow `Document` and `Token` to be set extra attributes

- :recycle: Refactoring:
- Move `BaseTransform` out of init for better integration with IDEs

- :books: Documentation:
- Improve documentation `cleaners` for sphinix better compatibility and explanation on transforms

:pencil: Changelog:

- :sparkles: Features:
- Add transformer for embeddings
- Add ner and embeddings token wise
- Add embeddings document wise
- Improve Stanza integration allowing to extract lemma and ner directly to the token object

- 🐛 Bug:
- Fix bug on importing nlpiper

- :books: Documentation:
- Update and improve documentation

:pencil: Changelog:

- :sparkles: Features:
- Create transformation pipeline to a document
- Rollback transformation steps applied to a document
- Add document Cleaners: `CleanURL`, `CleanEmail`, `CleanNumber`, `CleanPunctuation`, `CleanEOF`, `CleanMarkup`, `CleanAccents`
- Add document Tokenizers: `BasicTokenizer`, `MosesTokenizer`, `StanzaTokenizer`
- Add document Normalizers: `CaseTokens`, `RemovePunctuation`, `RemoveStopWords`, `VocabularyFilter`, `SpellCheck`, `Stemmer`
- Add Embeddings: `GensimEmbeddings`