* Enable the `JoinDocuments` node to work with documents with `score=None`
This fixes#2983
As of now, the `JoinDocuments` node will error out if any of the documents has `score=None` - which is possible, as some retriever are not able to provide a score, like the `TfidfRetriever` on Elasticsearch or the `BM25Retriever` on Weaviate.
THe reason for the error is that the `JoinDocuments` always sorts the documents by score and cannot sort when `score=None`.
There was a very similar issue for `JoinAnswers` too, which was addressed by this PR: https://github.com/deepset-ai/haystack/pull/2436
This solution applies the same solution to `JoinDocuments` - so both the `JoinAnswers` and `JoinDocuments` now will have the same additional argument to disable sorting when that is requried.
The solution is to add an argument to `JoinDocuments` called `sort_by_score: bool`, which allows the user to turn off the sorting of documents by score, but keeps the current functionality of sorting being performed as the default.
* Fixing test bug
* Addressing PR review comments
- Extending unit tests
- Simplifying logic
* Making the sorting work even with no scores
By making the no score being sorted as -Inf
* Forgot to commit the change in `join_docs.py`
* [EMPTY] Re-trigger CI
* Added am INFO log if the `JoinDocuments` is sorting while some of the docs have `score=None`
* Adjusting the arguments of `any()`
* [EMPTY] Re-trigger CI
* Add page number to Documents coming from PDFConverters and PreProcessor
* Fix mypy
* Update API Docs
* Update API Docs
* Remove unused imports
* Generate JSON schema
* Generate JSON schema
* Make test variable shorter
* Make regex a separate function
* Move counting of page breaks to a function
* Generate JSON schema
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Update API Documentation
* Don't create instance for testing staticmethod
* Update haystack/nodes/preprocessor/preprocessor.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* fix validation for dynamic outgoing edges
* Update Documentation & Code Style
* use class outgoing_edges as fallback if no instance is provided
* implement classmethod approach
* readd comment
* fix mypy
* fix tests
* set outgoing_edges for all components
* set outgoing_edges for mocks too
* set document store outgoing_edges to 1
* set last missing outgoing_edges
* enforce BaseComponent subclasses to define outgoing_edges
* override _calculate_outgoing_edges for FileTypeClassifier
* remove superfluous test
* set rest_api's custom component's outgoing_edges
* Update docstring
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* remove unnecessary else
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* enable Opensearch unit tests under Win
* move unit tests into a dedicated job
* skip audio tests on missing dependencies
* avoid failing test collection when soundfile is not available
* Update .github/workflows/tests.yml
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* Add support for model folder into BasePreProcessor
* First draft of custom model on PreProcessor
* Update Documentation & Code Style
* Update tests to support custom models
* Update Documentation & Code Style
* Test for wrong models in custom folder
* Default to ISO names on custom model folder
Use long names only when needed
* Update Documentation & Code Style
* Refactoring language names usage
* Update fallback logic
* Check unpickling error
* Updated tests using parametrize
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* Refactored common logic
* Add format control to NLTK load
* Tests improvements
Add a sample for specialized model
* Update Documentation & Code Style
* Minor log text update
* Log model format exception details
* Change pickle protocol version to 4 for 3.7 compat
* Removed unnecessary model folder parameter
Changed logic comparisons
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* Update Documentation & Code Style
* Removed unused import
* Change errors with warnings
* Change to absolute path
* Rename sentence tokenizer method
Co-authored-by: tstadel
* Check document content is a string before process
* Change to log errors and not warnings
* Update Documentation & Code Style
* Improve split sentences method
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* Update Documentation & Code Style
* Empty commit - trigger workflow
* Remove superfluous parameters
Co-authored-by: tstadel
* Explicit None checking
Co-authored-by: tstadel
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* Passing the all the meta-data in the summerizer
* Disable metadata forwarding if `generate_single_summary` is `True`
* Update Documentation & Code Style
* simplify tests
* Update Documentation & Code Style
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Change split logic to list
* Fix wrong parameter for run
* Fix mypy error
* Fix layout/raw parameter
* Add test for filename with whitespaces on PDFToText
* Update Documentation & Code Style
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Changing the name that crawled page is saved to avoid long file names error on some file systems
* Custom naming function for saving crawled files
* Update Documentation & Code Style
* Remove bad characters on file name and preffix
* Add test for naming function
* Update Documentation & Code Style
* Fix expensive regex recalculation and linter warns
* Check for exceptions on file dump
* Remove param_naming variable
* Fix file paths on Windows, Linux and Mac
* Update Documentation & Code Style
* Test using one of the docstrings examples
* Change default naming function
Update docstrings
* Applying formatting rules
* Update Documentation & Code Style
* Fix mypy incompatible assignment error
* Remove unused type declaration
* Fix typo
* Update tests for naming function
* Update Documentation & Code Style
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Use AutoTokenizer by default, to easily adapt to new models and tokenizers
* Add missing AutoTokenizer import
* Apply Black
* Missing import
* Fix DPR tests
* Remove tests on max length
* Update Documentation & Code Style
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Add new audio answer primitives
* Add AnswerToSpeech
* Add dependency group
* Update Documentation & Code Style
* Extract TextToSpeech in a helper class, create DocumentToSpeech and primitives
* Add tests
* Update Documentation & Code Style
* Add ability to compress audio and more tests
* Add audio group to test, all and all-gpu
* fix pylint
* Update Documentation & Code Style
* Accidental git tag
* Try pleasing mypy
* Update Documentation & Code Style
* fix pylint
* Add warning for missing OS library and support in CI
* Try fixing mypy
* Update Documentation & Code Style
* Add docs, simplify args for audio nodes and add tutorials
* Fix mypy
* Fix run_batch
* Feedback on tutorials
* fix mypy and pylint
* Fix mypy again
* Fix mypy yet again
* Fix the ci
* Fix dicts merge and install ffmpeg on CI
* Make the audio nodes import safe
* Trying to increase tolerance in audio test
* Fix import paths
* fix linter
* Update Documentation & Code Style
* Add audio libs in unit tests
* Update _text_to_speech.py
* Update answer_to_speech.py
* Use dedicated dataset & update telemetry
* Remove and use distilled roberta
* Revert special primitives so that the nodes run in indexing
* Improve tutorials and fix smaller bugs
* Update Documentation & Code Style
* Fix serialization issue
* Update Documentation & Code Style
* Improve tutorial
* Update Documentation & Code Style
* Update _text_to_speech.py
* Minor lg updates
* Minor lg updates to tutorial
* Making indexing work in tutorials
* Update Documentation & Code Style
* Improve docstrings
* Try to use GPU when available
* Update Documentation & Code Style
* Fixi mypy and pylint
* Try to pass the device correctly
* Update Documentation & Code Style
* Use type of device
* use .cpu()
* Improve .ipynb
* update apt index to be able to download libsndfile1
* Fix SpeechDocument.from_dict()
* Change pip URL
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* make crawler extract also hidden text
* Update Documentation & Code Style
* try to adapt test for extract_hidden_text
* Update Documentation & Code Style
* fix test bug
* fix bug in test
* added test for hidden text"
* Update Documentation & Code Style
* fix bug in test
* Update Documentation & Code Style
* fix test
* Update Documentation & Code Style
* fix other test bug
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Unify CI tests (from #2466)
* Update Documentation & Code Style
* Change folder names
* Fix markers list
* Remove marker 'slow', replaced with 'integration'
* Soften children check
* Start ES first so it has time to boot while Python is setup
* Run the full workflow
* Try to make pip upgrade on Windows
* Set KG tests as integration
* Update Documentation & Code Style
* typo
* faster pylint
* Make Pylint use the cache
* filter diff files for pylint
* debug pylint statement
* revert pylint changes
* Remove path from asserted log (fails on Windows)
* Skip preprocessor test on Windows
* Tackling Windows specific failures
* Fix pytest command for windows suites
* Remove \ from command
* Move poppler test into integration
* Skip opensearch test on windows
* Add tolerance in reader sas score for Windows
* Another pytorch approx
* Raise time limit for unit tests :(
* Skip poppler test on Windows CI
* Specify to pull with FF only in docs check
* temporarily run the docs check immediately
* Allow merge commit for now
* Try without fetch depth
* Accelerating test
* Accelerating test
* Add repository and ref alongside fetch-depth
* Separate out code&docs check from tests
* Use setup-python cache
* Delete custom action
* Remove the pull step in the docs check, will find a way to run on bot commits
* Add requirements.txt in .github for caching
* Actually install dependencies
* Change deps group for pylint
* Unclear why the requirements.txt is still required :/
* Fix the code check python setup
* Install all deps for pylint
* Make the autoformat check depend on tests and doc updates workflows
* Try installing dependencies in another order
* Try again to install the deps
* quoting the paths
* Ad back the requirements
* Try again to install rest_api and ui
* Change deps group
* Duplicate haystack install line
* See if the cache is the problem
* Disable also in mypy, who knows
* split the install step
* Split install step everywhere
* Revert "Separate out code&docs check from tests"
This reverts commit 1cd59b15ffc5b984e1d642dcbf4c8ccc2bb6c9bd.
* Add back the action
* Proactive support for audio (see text2speech branch)
* Fix label generator tests
* Remove install of libsndfile1 on win temporarily
* exclude audio tests on win
* install ffmpeg for integration tests
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Rewrite crawler tests (very slow) and fix small crawler bug
* Update Documentation & Code Style
* compile the regex only once
* Factor out the html files & add content check to most tests
* Clarify that even starting URLs can be excluded
* Update Documentation & Code Style
* Change signature
* Fix failing test
* Update Documentation & Code Style
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
