Add mxnet.text APIs (apache#8763)

* Add text utils

* Leftovers

* revise

* before load embeddings

* glossary done

* Add/revise text utils, revise test cases

* Add docstrings

* clean package init

* remove play

* Resolve issues and complete docstrings

* disable pylint

* Remove tqdm dependency

* Add encoding

utf8

utf

utf

utf

* remove non-ascii

* fix textcase

* remove decode in glossary

* py2 unicode

* Fix py2 error

* add tests

* Test all embds

* test some embeds

* Add getter for glossary

* remove util from path, revise interfaces of glossary

* skip some test, before major revise

* Add TextIndexer, only TextEmbed needs revised before major revise

* before major revise

* minor update

* Revise TextIndexer with test

* lint

* lint

* Revise TextEmbed, FastText, Glove, CustmonEmbed with test

* Revision done except for docstr

* Add unit tests for utils

* almost no pylint disable, yeah

* doc minor updates

* re-run

* re-run

* except for register

* except for register

* Revise register/create, add get_registry

* revise

* More readability

* py2 compatibility

* Update doc

* Revise based on feedbacks from NLP team

* add init

* Support indexing for any hashable and comparable token

* Add test cases

* remove type cmp

* Fix doc error and add API descriptions

* Fix api doc error

* add members explicitly

* re-order modules in text.md

* url in one line

* add property desc for all inherited classes for rst parsing

* escape \n

* update glossary example

* escape \n

* add use case

* Make doc more user-friendly

* proper imports, gluon.nn.Embedding use case

* fix links

* re-org link level

* tokens_to_indices

* to_indices, to_tokens

docs/api/python/index.md

@@ -98,6 +98,15 @@ imported by running:
    io/io.md
 ```
 
+## Text API
+
+```eval_rst
+.. toctree::
+   :maxdepth: 1
+
+   text/text.md
+```
+
 ## Image API
 
 ```eval_rst