Fix tests

Author: serhii73

.github/workflows/cifuzz.yml

@@ -21,7 +21,7 @@ jobs:
         fuzz-seconds: 600
         output-sarif: true
     - name: Upload Crash
-      uses: actions/upload-artifact@v3
+      uses: actions/upload-artifact@v4
       if: failure() && steps.build.outcome == 'success'
       with:
         name: artifacts

dateparser/custom_language_detection/fasttext.py

@@ -1,7 +1,6 @@
 import os
 
-import fasttext
-
+from dateparser.custom_language_detection.fasttext_wrapper import load_model
 from dateparser_cli.exceptions import FastTextModelNotFoundException
 from dateparser_cli.fasttext_manager import fasttext_downloader
 from dateparser_cli.utils import create_data_model_home, dateparser_model_home
@@ -27,7 +26,7 @@ def _load_fasttext_model():
     model_path = os.path.join(dateparser_model_home, downloaded_models[0])
     if not os.path.isfile(model_path):
         raise FastTextModelNotFoundException("Fasttext model file not found")
-    _FastTextCache.model = fasttext.load_model(model_path)
+    _FastTextCache.model = load_model(model_path)
     return _FastTextCache.model
 
 

dateparser/custom_language_detection/fasttext_wrapper.py

@@ -0,0 +1,34 @@
+import fasttext
+import numpy as np
+
+
+class FastTextWrapper:
+    def __init__(self, model_path):
+        self.model = fasttext.load_model(model_path)
+
+    def predict(self, text, k=1, threshold=0.0, on_unicode_error="strict"):
+        def check(entry):
+            if entry.find("\n") != -1:
+                raise ValueError("predict processes one line at a time (remove '\\n')")
+            entry += "\n"
+            return entry
+
+        if isinstance(text, list):
+            text = [check(entry) for entry in text]
+            all_labels, all_probs = self.model.f.multilinePredict(
+                text, k, threshold, on_unicode_error
+            )
+            return all_labels, all_probs
+        else:
+            text = check(text)
+            predictions = self.model.f.predict(text, k, threshold, on_unicode_error)
+            if predictions:
+                probs, labels = zip(*predictions)
+            else:
+                probs, labels = ([], ())
+
+            return labels, np.asarray(probs)
+
+
+def load_model(model_path):
+    return FastTextWrapper(model_path)

fuzzing/requirements.txt

@@ -4,28 +4,17 @@ universal = 1
 [flake8]
 max-line-length = 119
 ignore =
-    # This rule goes against the PEP 8 recommended style and it's incompatible
-    # with W504
     W503
 
-    # Exclude automatically generated files
-    # E501: Line too long
-    dateparser/data/date_translation_data/* E501
-
-    # Exclude files that are meant to provide top-level imports
-    # F401: Module imported but unused
-    dateparser/data/__init__.py F401
-    dateparser/languages/__init__.py F401
-
-    # Issues pending a review:
-    dateparser/freshness_date_parser.py E722
-    dateparser/parser.py E722
-    dateparser/docs/conf.py E402
-
-    # Additional ignored codes
     E203
     E501
     E722
     F401
     E701
     E704
+
+exclude =
+    dateparser/data/date_translation_data/*
+    dateparser/data/__init__.py
+    dateparser/languages/__init__.py
+    docs/conf.py

setup.cfg

@@ -119,7 +119,7 @@ def test_dates_which_match_locales_are_parsed(
                 languages=["en"],
                 region="",
                 date_formats=["%a", "%a", "%a", "%a"],
-                expected_date=datetime(1969, 1, 31, 14, 4),
+                expected_date=datetime(1969, 1, 31, 13, 4),
             )
         ]
     )

tests/test_clean_api.py

@@ -642,7 +642,14 @@ def test_search_and_parse(self, shortname, string, expected, settings=None):
                     (
                         "June 23th 5 pm EST",
                         datetime.datetime(
-                            2023, 6, 23, 17, 0, tzinfo=pytz.timezone("EST")
+                            2023,
+                            6,
+                            23,
+                            17,
+                            0,
+                            tzinfo=datetime.timezone(
+                                datetime.timedelta(hours=-5), name="EST"
+                            ),
                         ),
                     ),
                     ("May 31", datetime.datetime(2023, 5, 31, 0, 0)),

tests/test_search.py

@@ -5,7 +5,7 @@ envlist = flake8, py3
 deps =
     -rdateparser_scripts/requirements.txt
     -rtests/requirements.txt
-    -rfuzzing/requirements.txt
+    atheris; python_version < '3.12'
 
 [testenv]
 deps =
@@ -40,10 +40,9 @@ commands =
 
 [testenv:twinecheck]
 basepython = python3
-extras = []
 deps =
-    twine==4.0.2
-    build==1.0.3
+    build
+    twine
 commands =
-    python -m build --sdist
+    python -m build --sdist --wheel
     twine check dist/*