dateparser 1.2.1__py3-none-any.whl → 1.2.2__py3-none-any.whl

dateparser/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "1.2.1"
+__version__ = "1.2.2"
 
 from .conf import apply_settings
 from .date import DateDataParser

dateparser/parser.py

@@ -399,8 +399,8 @@ class _parser:
             return datetime(**params)
         except ValueError as e:
             error_text = e.__str__()
-            error_msgs = ["day is out of range", "day must be in"]
-            if error_msgs[0] in error_text or error_msgs[1] in error_text:
+            error_msgs = ["day is out of range", "day must be in", "must be in range"]
+            if any(msg in error_text for msg in error_msgs):
                 if not (self._token_day or hasattr(self, "_token_weekday")):
                     # if day is not available put last day of the month
                     params["day"] = get_last_day_of_month(
@@ -512,6 +512,10 @@ class _parser:
 
             dateobj = dateobj + delta
 
+            # set the token_month here so that it is not subsequently
+            # altered by _correct_for_month
+            self._token_month = dateobj.month
+
         # NOTE: If this assert fires, self.now needs to be made offset-aware in a similar
         # way that dateobj is temporarily made offset-aware.
         assert not (self.now.tzinfo is None and dateobj.tzinfo is not None), (

dateparser/search/__init__.py

@@ -57,6 +57,8 @@ def search_dates(
      ('on May 6th 2004', datetime.datetime(2004, 5, 6, 0, 0))]
 
     """
+    text = _search_with_detection.preprocess_text(text, languages)
+
     result = _search_with_detection.search_dates(
         text=text,
         languages=languages,

dateparser/search/search.py

@@ -295,3 +295,10 @@ class DateSearchWithDetection:
                 language_shortname, text, settings=settings
             ),
         }
+
+    def preprocess_text(self, text, languages):
+        """Preprocess text to handle language-specific quirks."""
+        if languages and "ru" in languages:
+            # Replace "с" (from) before numbers with a placeholder
+            text = re.sub(r"\bс\s+(?=\d)", "[FROM] ", text)
+        return text

dateparser/timezone_parser.py

@@ -1,4 +1,8 @@
+import os
+import pickle
+import zlib
 from datetime import datetime, timedelta, timezone, tzinfo
+from pathlib import Path
 
 import regex as re
 
@@ -84,8 +88,47 @@ def get_local_tz_offset():
     return offset
 
 
-_search_regex_parts = []
-_tz_offsets = list(build_tz_offsets(_search_regex_parts))
-_search_regex = re.compile("|".join(_search_regex_parts))
-_search_regex_ignorecase = re.compile("|".join(_search_regex_parts), re.IGNORECASE)
 local_tz_offset = get_local_tz_offset()
+
+_tz_offsets = None
+_search_regex = None
+_search_regex_ignorecase = None
+
+
+def _load_offsets(cache_path, current_hash):
+    global _tz_offsets, _search_regex, _search_regex_ignorecase
+
+    try:
+        with open(cache_path, mode="rb") as file:
+            (
+                serialized_hash,
+                _tz_offsets,
+                _search_regex,
+                _search_regex_ignorecase,
+            ) = pickle.load(file)
+            if current_hash is None or current_hash == serialized_hash:
+                return
+    except (FileNotFoundError, ValueError, TypeError):
+        pass
+
+    _search_regex_parts = []
+    _tz_offsets = list(build_tz_offsets(_search_regex_parts))
+    _search_regex = re.compile("|".join(_search_regex_parts))
+    _search_regex_ignorecase = re.compile("|".join(_search_regex_parts), re.IGNORECASE)
+
+    with open(cache_path, mode="wb") as file:
+        pickle.dump(
+            (current_hash, _tz_offsets, _search_regex, _search_regex_ignorecase),
+            file,
+            protocol=5,
+        )
+
+
+CACHE_PATH = Path(__file__).parent.joinpath("data", "dateparser_tz_cache.pkl")
+
+if "BUILD_TZ_CACHE" in os.environ:
+    current_hash = zlib.crc32(str(timezone_info_list).encode("utf-8"))
+else:
+    current_hash = None
+
+_load_offsets(CACHE_PATH, current_hash)

{dateparser-1.2.1.dist-info → dateparser-1.2.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
-Metadata-Version: 2.2
+Metadata-Version: 2.4
 Name: dateparser
-Version: 1.2.1
+Version: 1.2.2
 Summary: Date parsing library designed to parse dates from HTML pages
 Home-page: https://github.com/scrapinghub/dateparser
 Author: Scrapinghub
@@ -24,7 +24,7 @@ License-File: LICENSE
 License-File: AUTHORS.rst
 Requires-Dist: python-dateutil>=2.7.0
 Requires-Dist: pytz>=2024.2
-Requires-Dist: regex!=2019.02.19,!=2021.8.27,>=2015.06.24
+Requires-Dist: regex>=2024.9.11
 Requires-Dist: tzlocal>=0.2
 Provides-Extra: calendars
 Requires-Dist: convertdate>=2.2.1; extra == "calendars"
@@ -41,6 +41,7 @@ Dynamic: description
 Dynamic: home-page
 Dynamic: keywords
 Dynamic: license
+Dynamic: license-file
 Dynamic: project-url
 Dynamic: provides-extra
 Dynamic: requires-dist
@@ -318,6 +319,19 @@ To be able to use them you need to install the `calendar` extra by typing:
 History
 =======
 
+1.2.2 (2025-06-26)
+------------------
+
+Fixes:
+
+- Handle the Russian preposition “с” (#1261)
+- Fix weekday search (#1274)
+
+Improvements:
+
+- Add Python 3.14 support (#1273)
+- Cache timezone offsets to improve import time (#1250)
+
 1.2.1 (2025-02-05)
 ------------------
 

{dateparser-1.2.1.dist-info → dateparser-1.2.2.dist-info}/RECORD

@@ -1,10 +1,10 @@
-dateparser/__init__.py,sha256=3ajvpKU-xlhD0gF4Xp1bqU3PGTrMFLfYBlua29tRM1Q,2739
+dateparser/__init__.py,sha256=Wfhc4HQ6pB00SH9DjixNnmfukzzhl-Bsoo-S-sshegk,2739
 dateparser/conf.py,sha256=ynrRajwQ6dtgbSm9T0plfxQEYEarHo6Qsr76ISInX-4,8467
 dateparser/date.py,sha256=AluFz3-uzuRL6Y1cF5uFgjpcxEmTspJAHGMdIR1oBzE,21763
 dateparser/date_parser.py,sha256=LPMD1La3fUTyfa17i0Ng-mii6ucWIBme-zJJY3es-Ug,1782
 dateparser/freshness_date_parser.py,sha256=VgFAxveZqRDXCHTkQzTab7jx197btjv-YFcSOeZwOAY,5081
-dateparser/parser.py,sha256=_PiRb9uoIQNcnymsd0KRO6ZO5M_82Aa418rwp-PHGKU,25838
-dateparser/timezone_parser.py,sha256=98cWyyBl8D71Ry-cAVgB0pa935NPHZQiP451YLvK_zU,2913
+dateparser/parser.py,sha256=KDde8HlyD5wFERiT8ABrIxTakLsEE7VYA4Kif6VWRbQ,26005
+dateparser/timezone_parser.py,sha256=MoMTAH__hkeaoZQ6VaZhJpUPxd-yA9QTczm-mmHztKc,4042
 dateparser/timezones.py,sha256=gtJhZDCy4sw37MpUogWk4LHKmn8DpG5caO3LIaCC264,13981
 dateparser/calendars/__init__.py,sha256=Gf9qLl9xbCFhmuWtuakkaHy_vSuOLOoo1kr33xJGkIY,4323
 dateparser/calendars/hijri.py,sha256=J9IUHs162UslQ4r5OPIr-BgmiyY3_4kj_z1wOIlB57E,168
@@ -16,6 +16,7 @@ dateparser/custom_language_detection/fasttext.py,sha256=uuNN8JlKcUSi5pqUrnywPRrq
 dateparser/custom_language_detection/langdetect.py,sha256=lxSrW6bpB3f4_YaF3nnyi6xmSsvwsnmBmOelHQGKeck,1195
 dateparser/custom_language_detection/language_mapping.py,sha256=HeZLw65ghqXmjLaE8KEOm4BZedkFVrX7dvC_FynXWtk,557
 dateparser/data/__init__.py,sha256=20PfpbkZv8Q9LU6p0T4RA0YNq8oQykhvWThKPhYgp6Y,115
+dateparser/data/dateparser_tz_cache.pkl,sha256=wXxZFkophi11wLJJdVEgUgLpcqf1GS5SAA89e8Mg0-I,134536
 dateparser/data/languages_info.py,sha256=BPHjASz2MUvgc7kF2uXrtYbcj823kjLxrR8X_wLiAEc,13406
 dateparser/data/date_translation_data/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 dateparser/data/date_translation_data/af.py,sha256=LR5yRcwOJ9_QWn5EH_jwfAhwNGtnf4RFhdNoVHqjd7U,4588
@@ -228,12 +229,14 @@ dateparser/languages/dictionary.py,sha256=KiSswl08OsBqW-UiBeST2ff-BtqZyH-sNy1FAs
 dateparser/languages/loader.py,sha256=gTyyVEX8ppeVEQ1SrIjjxwSq8hpTI5HTnS_cyVotegc,7323
 dateparser/languages/locale.py,sha256=3T5VIzS5WHpcs7fScKW-rQiW4refCBrp-BxyXzDYlio,24631
 dateparser/languages/validation.py,sha256=gtLpxvh3XI_nzryBZ8NA6QL8YLcT6HV1j81dV3nDxY4,16750
-dateparser/search/__init__.py,sha256=__laR2Q19meiPpZLILINZMePGRG4DmvOoc2tOncU9ZM,2841
+dateparser/search/__init__.py,sha256=4AixzJu9YID1vzNSKx8uJ0cyKNZGvO8BzMN6Jpoci84,2909
 dateparser/search/detection.py,sha256=YAnTYbM18FGZV2pxn6lNDbfoiQttzM52ZDgOhlV87II,2695
-dateparser/search/search.py,sha256=-c9LIBUnTPRfchEoFrcVeu5JVJBkMsl1jGueB3FhKKY,11846
+dateparser/search/search.py,sha256=v9eH-xwZBu7b4yGsMiwLBBBh8C1l0IS4hmdvd7Tv2V8,12154
 dateparser/search/text_detection.py,sha256=SSQUOr5V3Qm946Tiz7hNWcpYJtcyP9Bxayxec0hC7zc,3238
 dateparser/utils/__init__.py,sha256=X5ssdQuo60F_SnVcKpt-a2ehvPClMaX2pQ2vwGWGSac,7234
 dateparser/utils/strptime.py,sha256=tnPJ_C4wKZAGwW03z5-XHy5ykZ3lP6RLpw6QR-jT-f0,2853
+dateparser-1.2.2.dist-info/licenses/AUTHORS.rst,sha256=xyYeT2AGKVrsHYxoOxgDE8XmSK7VNtJNqFzw25JLE2g,711
+dateparser-1.2.2.dist-info/licenses/LICENSE,sha256=t122Vbt6QqjoVh0NuHMVi7naudRoryRAAPAtfdg-l_Q,1468
 dateparser_cli/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 dateparser_cli/cli.py,sha256=pZSVwZO_N1AfPJfbbTHgvEnV8kmvyKmd6YQLYT3epis,991
 dateparser_cli/exceptions.py,sha256=ifwai5On6opuhUpBHMoz0iFkybtZg7m3gHHnxtOEoEk,58
@@ -247,10 +250,8 @@ dateparser_scripts/order_languages.py,sha256=BkYmAqFvoMvj3I7xcjeutz0Bx5VQQky68Y5
 dateparser_scripts/update_supported_languages_and_locales.py,sha256=BaHzzCiLKvkhKcal44CyedZNIUrBrI4fm0LxHicRJ8g,1460
 dateparser_scripts/utils.py,sha256=Uw4HbgwbKYuUeztUw9OQqtzPtjmBxBBqGytLXjtSia8,2579
 dateparser_scripts/write_complete_data.py,sha256=w1pExoB7Z5oN02pSMvkzN2K-kQ_5UbehAdvTZ_bYUCQ,4540
-dateparser-1.2.1.dist-info/AUTHORS.rst,sha256=xyYeT2AGKVrsHYxoOxgDE8XmSK7VNtJNqFzw25JLE2g,711
-dateparser-1.2.1.dist-info/LICENSE,sha256=t122Vbt6QqjoVh0NuHMVi7naudRoryRAAPAtfdg-l_Q,1468
-dateparser-1.2.1.dist-info/METADATA,sha256=xBE9ekzZoqtRzII8t2YytRzToJ0qJ3suxF0qjBJPyEQ,29406
-dateparser-1.2.1.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
-dateparser-1.2.1.dist-info/entry_points.txt,sha256=QPwc8kOjbaaxJdyQXiix9sGbUjC_G_Ga5LgJ_11KKAs,68
-dateparser-1.2.1.dist-info/top_level.txt,sha256=LujVBIKC69tvws1XkgyOFRDjEEd-E1SjAirYrhEbqn8,61
-dateparser-1.2.1.dist-info/RECORD,,
+dateparser-1.2.2.dist-info/METADATA,sha256=Ji6LM-TkErOhjKfv5MfYuayFv8C-JBsOa0BfjRv0Tks,29635
+dateparser-1.2.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+dateparser-1.2.2.dist-info/entry_points.txt,sha256=QPwc8kOjbaaxJdyQXiix9sGbUjC_G_Ga5LgJ_11KKAs,68
+dateparser-1.2.2.dist-info/top_level.txt,sha256=LujVBIKC69tvws1XkgyOFRDjEEd-E1SjAirYrhEbqn8,61
+dateparser-1.2.2.dist-info/RECORD,,

{dateparser-1.2.1.dist-info → dateparser-1.2.2.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (75.8.0)
+Generator: setuptools (80.9.0)
 Root-Is-Purelib: true
 Tag: py3-none-any