* feat: add SQL connectors

* feat: add .env configuration for mysql and postgres

* feat: a smart df can now accept a connector as a param

* fix: correct error prompt now handles multiple dataframes

* feat: cache the dataframes from third parties

* fix: truncate long text from the sample

* feat: extracting filters (#451) (#483)

* fix: setters were not overriding some properties in some cases

* fix: remove multiple df overwrites (#436)

* improved on _is_df_overwrite method to check all targets

* test: add test for multiple df overwrite

---------

Co-authored-by: Gabriele Venturi <[email protected]>

* fix: prevent to_* pandas methods from running

* Release v0.8.3

* fix: environment for executing code (#419)

* fix: environment for executing code

* (fix): set `matplot.pyplot` package under "plt" alias in
  `_get_environment()`

* feat: Try to import a package when NameError (#430)

* (refactor): split calling of `exec()` for code and handling of errors
  into separate methods;
* (feat): add trying to import the `exc.name` and add one to context
  when `exc` is an instance of `NameError`;
* (tests): add tests for new methods;

* fix: add logging for `NameError` handling (#430)

* (fix): add log a general exception raised when trying to fix
  `NameError`

* fix: compatibility with python 3.9 (#430)

* (fix): update `handle_error()` method, add parsing `exc.args[0]` for
  cases with NameError if there is no attribute "name" of object of the
  exception.

* fix: update for `handle_error()`(#430)

* (fix): add checking if the library to be imported is preset in the
  whitelist of third-packages;
* (fix): remove "plt" when forming environment for code to be run;
* (fix): add raising an exception that was caught during fixing code
  with the error correction framework;
* (tests): add test-case for retrying to fix code with the error
  correction framework unsuccessfully;
* (tests): fix issue with inappropriate instantiation of numpy's `array`
  object within several test methods;

* Release v0.8.4

* Release v1.0b1

* Release v1.0

* fix: convert polars to pandas df and back to polars after exec the code

* docs: update docs for v1

* Release v1.0.1

* fix: remove shortcuts from backwords compatibile version

* Release v1.0.2

* refactor: minor updates (#446)

* (refactor): inappropriate type hint for `_has_run` in base class for
  middlewares;
* (style): remove built-ins name shadowing in `__init__()` method of
  `Prompt` class;

* docs: update LLMs documentation code to match the last version (#445)

I have updated OpenAI, Starcoder, Falcon, and AzureOpenAI.

But I haven't edited GooglePalm because it still uses api_key in the last version not api_token

Co-authored-by: Gabriele Venturi <[email protected]>

* fix: do not override setters and getters with __getattr__

* Release v1.0.3

* fix: huggingface models (#454)

* Release v1.0.4

* fix: update code manager to convert prompt id to string (#462)

* test: increase test coverage (#460)

* Update LLMs documentation code to match the last version

I have updated OpenAI, Starcoder, Falcon, and AzureOpenAI.

But I haven't edited GooglePalm because it still uses api_key in the last version not api_token

* Create test_google_vertexai.py

Create test cases for Class Google Vertexai. I have tested initailization function with default model provided and custom one as well.
Then I have built test cases for validation of class when model provided and without model.

* Update test_smartdataframe.py

Created test cases for _load_df() and _import_from_file() functions

* test: remove verbose test

* test: unskip skipped test

* test: test multiple invalid file formats

* test: fix tests

---------

Co-authored-by: Gabriele Venturi <[email protected]>

* fix: recursive error on df last_result property (#468) (#469)

* fix: recursive error on df last_result property (#468)

* fix: change last_result type to str

* test: add tests

---------

Co-authored-by: Gabriele Venturi <[email protected]>

* Release v1.0.5

* fix(llms): AzureOpenAI call method signature (#476)

* fix: added last_error property reference in smart_datalake (#475)

* fix: temporarely disable multi-turn capacity, causing hallucinations

* Release v1.0.6

* fix: create required folders as the datalake is initialized

* Release v1.0.7

* fix: create required folders in the project folder

* Release v1.0.8

* chore: fix typing in the logger logs (#464)

* docs: add video explanation for SmartDataframe

* feat: extracting filters (#451)

* (feat): add `_extract_filters()`, `_extract_comparisons()`,
  `_tokenize_operand()` methods to `CodeManager`;
* (tests): add basic tests;

* test: extracting filters (#451)

* (tests): rename `test_extract_filters()` to
  `test_extract_filters_col_index()`, update asserting according to
  task requirements;
* (tests): add `test_extract_filters_col_index_multiple_df()` for
  multiple dfs;

* fix: update according to requirements (#451)

* (feat): add module `node_visitor.py`, add `AssignmentVisitor` in
  `node_visitors.py`;
* (fix): `_extract_filters()` now returns dictionary;
* (feat): add grouping by df number when forming;
* (tests): update naming in test cases;
* (tests): add test method with non default (not `df`) variable name;

* feat: extracting filters (#451)

* (feat): add `CallVisitor` to `node_visitors.py` module;
* (feat): add extracting comparisions in `filter()` method being called
  as an attribute of object named `pl` or `polars`;
* (tests): test methods for extracting polars-like filters;

* refactor: extracting filters (#451)

* (refactor): add handling exceptions in `_extract_filters()` method;

* docs: extracting filters (#451)

* (docs): add documentations for implemented functions;
* (style): type hinting;

* test: merge similar tests

---------

Co-authored-by: Gabriele Venturi <[email protected]>
Co-authored-by: Aymane Hachcham <[email protected]>
Co-authored-by: Omar Elsherif <[email protected]>
Co-authored-by: Hanchung Lee <[email protected]>
Co-authored-by: Sanchit Bhavsar <[email protected]>
Co-authored-by: Massimiliano Pronesti <[email protected]>

* docs: add example of usage with connectors

* feat: allow to save and load config for connectors

* feat: add dynamic filters to connectors

* feat: add yahoo finance connector

* fix: prevent sql injections

* refactor: improve sql connector

* test: test yahoo finance connector

* refactor: update `DataframeAbstract` (#547)

* refactor: `to_dict()` parameters

* (refactor): update behaviour of `to_dict()` method, make it possible
  to accept `into` and `as_series` parameters (the last one is for
  polars dataframes).
* (tests): add tests for casting the dataframe to a dictionary, add
  tests for passing parameters in the proxy-call to dataframe's
  `to_dict()` method.

* docs: docstrings for `DataframeAbstract`

* (docs): add docstrings for proxy-methods in `DataframeAbstract` class

* chore: type hint issue

* (chore): update return type for `df_type` function

* feat: improve prompt in the way it handles conversation and dataframes

---------

Co-authored-by: Ihor <[email protected]>
Co-authored-by: Aymane Hachcham <[email protected]>
Co-authored-by: Omar Elsherif <[email protected]>
Co-authored-by: Hanchung Lee <[email protected]>
Co-authored-by: Sanchit Bhavsar <[email protected]>
Co-authored-by: Massimiliano Pronesti <[email protected]>