Dimcli

* remove legacy `Dsl.login()` method (should use `dimcli.login` instead)
* fixed BUG in repl that prevented autocompletion of non-filter fields
* remove Warning by updating `pandas.json_normalize` import
* project docs via sphinx
* CLI command `record_notebook` turns latest queries into jupyter notebook

* grammar for new source `Datasets`
* improved `dsl_escape` method desc, when a query with quotes needs to be escaped

* Refactor command line options. Try it with `dimcli -h`
* New command for resolving Dimensions IDs into URLs `dimcli -id [value]`
* Grammar for DSL V 1.22
* Refactor login: invalid credentials raise an error (previously just a message)
* Refactor query_iterative: error while looping provides better infos

* `dimcli.Dataset` is the new name of dimcli.Result and is available at module level
* methods to build `Dataset` from raw data (instead of from a query) so to reuse as_dataframe() methods etc: `Dataset.from_publications_list(data)`
* data can be from list of dictionaries
* or data can be a dataframe (eg obtained from previous query operations)
* experimental feature `Dataset.as_dimensions_url()`
* `dimcli.login()` method now accepts an optional argument `verbose=True/False` to enable/suppress feedback
* `Dsl.query_iterative` takes an extra argument `pause` determining iteration break length (default = 1.5 seconds)
* added more utility functions to `dimcli.shortcuts` module
* new method `Dataset.as_dataframe_concepts()` for publications and grants. Returns a full list of all concepts, including position, score (in the context of a publication/grant) and frequency (across records).

* fixed error with `as_dataframe_investigators`
* fixed error with `limit` keyword wrongly found in iterative_query
* fixed error with verbose flag lost after queries with 403s or automatic retries
* improved autocomplete after magic command %dsl

* fix for double-printed error messages in CLI
* removed deprecated magic commands
* improved printouts on magic methods
* ensuring magic methods with dataframe work only with `search` queries

* Query Warnings and Error messages printed out by default in Jupyter environments
* Added `verbose` option to suppress all messages (default = True)
* `dsl = dimcli.Dls(verbose=False)`: global
* `dsl.query(q, verbose=False)`: local for a specific query
* Improved `query_iterative` to handle max 50k results and timeouts (from too many queries)
* Refactored CLI module and imports