haystack/test/nodes/test_document_classifier.py

141 lines
5.5 KiB
Python
Raw Normal View History

import pytest
Refactoring of the `haystack` package (#1624) * Files moved, imports all broken * Fix most imports and docstrings into * Fix the paths to the modules in the API docs * Add latest docstring and tutorial changes * Add a few pipelines that were lost in the inports * Fix a bunch of mypy warnings * Add latest docstring and tutorial changes * Create a file_classifier module * Add docs for file_classifier * Fixed most circular imports, now the REST API can start * Add latest docstring and tutorial changes * Tackling more mypy issues * Reintroduce from FARM and fix last mypy issues hopefully * Re-enable old-style imports * Fix some more import from the top-level package in an attempt to sort out circular imports * Fix some imports in tests to new-style to prevent failed class equalities from breaking tests * Change document_store into document_stores * Update imports in tutorials * Add latest docstring and tutorial changes * Probably fixes summarizer tests * Improve the old-style import allowing module imports (should work) * Try to fix the docs * Remove dedicated KnowledgeGraph page from autodocs * Remove dedicated GraphRetriever page from autodocs * Fix generate_docstrings.sh with an updated list of yaml files to look for * Fix some more modules in the docs * Fix the document stores docs too * Fix a small issue on Tutorial14 * Add latest docstring and tutorial changes * Add deprecation warning to old-style imports * Remove stray folder and import Dict into dense.py * Change import path for MLFlowLogger * Add old loggers path to the import path aliases * Fix debug output of convert_ipynb.py * Fix circular import on BaseRetriever * Missed one merge block * re-run tutorial 5 * Fix imports in tutorial 5 * Re-enable squad_to_dpr CLI from the root package and move get_batches_from_generator into document_stores.base * Add latest docstring and tutorial changes * Fix typo in utils __init__ * Fix a few more imports * Fix benchmarks too * New-style imports in test_knowledge_graph * Rollback setup.py * Rollback squad_to_dpr too Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2021-10-25 15:50:23 +02:00
from haystack.schema import Document
from haystack.nodes.document_classifier.base import BaseDocumentClassifier
[CI Refactoring] Workflow refactoring (#2576) * 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>
2022-06-07 09:23:03 +02:00
@pytest.mark.integration
def test_document_classifier(document_classifier):
assert isinstance(document_classifier, BaseDocumentClassifier)
docs = [
Document(
content="""That's good. I like it.""" * 700, # extra long text to check truncation
meta={"name": "0"},
id="1",
),
Document(content="""That's bad. I don't like it.""", meta={"name": "1"}, id="2"),
]
results = document_classifier.predict(documents=docs)
expected_labels = ["joy", "sadness"]
for i, doc in enumerate(results):
assert doc.to_dict()["meta"]["classification"]["label"] == expected_labels[i]
@pytest.mark.integration
def test_document_classifier_details(document_classifier):
docs = [Document(content="""That's good. I like it."""), Document(content="""That's bad. I don't like it.""")]
results = document_classifier.predict(documents=docs)
for doc in results:
assert "details" in doc.meta["classification"]
assert len(doc.meta["classification"]["details"]) == 2 # top_k = 2
[CI Refactoring] Workflow refactoring (#2576) * 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>
2022-06-07 09:23:03 +02:00
@pytest.mark.integration
def test_document_classifier_batch_single_doc_list(document_classifier):
docs = [
Document(content="""That's good. I like it.""", meta={"name": "0"}, id="1"),
Document(content="""That's bad. I don't like it.""", meta={"name": "1"}, id="2"),
]
results = document_classifier.predict_batch(documents=docs)
expected_labels = ["joy", "sadness"]
for i, doc in enumerate(results):
assert doc.to_dict()["meta"]["classification"]["label"] == expected_labels[i]
[CI Refactoring] Workflow refactoring (#2576) * 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>
2022-06-07 09:23:03 +02:00
@pytest.mark.integration
def test_document_classifier_batch_multiple_doc_lists(document_classifier):
docs = [
Document(content="""That's good. I like it.""", meta={"name": "0"}, id="1"),
Document(content="""That's bad. I don't like it.""", meta={"name": "1"}, id="2"),
]
results = document_classifier.predict_batch(documents=[docs, docs])
assert len(results) == 2 # 2 Document lists
expected_labels = ["joy", "sadness"]
for i, doc in enumerate(results[0]):
assert doc.to_dict()["meta"]["classification"]["label"] == expected_labels[i]
[CI Refactoring] Workflow refactoring (#2576) * 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>
2022-06-07 09:23:03 +02:00
@pytest.mark.integration
def test_zero_shot_document_classifier(zero_shot_document_classifier):
assert isinstance(zero_shot_document_classifier, BaseDocumentClassifier)
docs = [
Document(
content="""That's good. I like it.""" * 700, # extra long text to check truncation
meta={"name": "0"},
id="1",
),
Document(content="""That's bad. I don't like it.""", meta={"name": "1"}, id="2"),
]
results = zero_shot_document_classifier.predict(documents=docs)
expected_labels = ["positive", "negative"]
for i, doc in enumerate(results):
assert doc.to_dict()["meta"]["classification"]["label"] == expected_labels[i]
@pytest.mark.integration
def test_zero_shot_document_classifier_details(zero_shot_document_classifier):
docs = [Document(content="""That's good. I like it."""), Document(content="""That's bad. I don't like it.""")]
results = zero_shot_document_classifier.predict(documents=docs)
for doc in results:
assert "details" in doc.meta["classification"]
assert len(doc.meta["classification"]["details"]) == 2 # n_labels = 2
[CI Refactoring] Workflow refactoring (#2576) * 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>
2022-06-07 09:23:03 +02:00
@pytest.mark.integration
def test_document_classifier_batch_size(batched_document_classifier):
assert isinstance(batched_document_classifier, BaseDocumentClassifier)
docs = [
Document(
content="""That's good. I like it.""" * 700, # extra long text to check truncation
meta={"name": "0"},
id="1",
),
Document(content="""That's bad. I don't like it.""", meta={"name": "1"}, id="2"),
]
results = batched_document_classifier.predict(documents=docs)
expected_labels = ["joy", "sadness"]
for i, doc in enumerate(results):
assert doc.to_dict()["meta"]["classification"]["label"] == expected_labels[i]
[CI Refactoring] Workflow refactoring (#2576) * 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>
2022-06-07 09:23:03 +02:00
@pytest.mark.integration
def test_document_classifier_as_index_node(indexing_document_classifier):
assert isinstance(indexing_document_classifier, BaseDocumentClassifier)
docs = [
{
"content": """That's good. I like it.""" * 700, # extra long text to check truncation
"meta": {"name": "0"},
"id": "1",
"class_field": "That's bad.",
},
{"content": """That's bad. I like it.""", "meta": {"name": "1"}, "id": "2", "class_field": "That's good."},
]
output, output_name = indexing_document_classifier.run(documents=docs, root_node="File")
expected_labels = ["sadness", "joy"]
for i, doc in enumerate(output["documents"]):
assert doc["meta"]["classification"]["label"] == expected_labels[i]
[CI Refactoring] Workflow refactoring (#2576) * 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>
2022-06-07 09:23:03 +02:00
@pytest.mark.integration
def test_document_classifier_as_query_node(document_classifier):
assert isinstance(document_classifier, BaseDocumentClassifier)
docs = [
Document(
content="""That's good. I like it.""" * 700, # extra long text to check truncation
meta={"name": "0"},
id="1",
),
Document(content="""That's bad. I don't like it.""", meta={"name": "1"}, id="2"),
]
output, output_name = document_classifier.run(documents=docs, root_node="Query")
expected_labels = ["joy", "sadness"]
for i, doc in enumerate(output["documents"]):
assert doc.to_dict()["meta"]["classification"]["label"] == expected_labels[i]