dateparser 1.3.0__tar.gz → 1.4.0__tar.gz

{dateparser-1.3.0 → dateparser-1.4.0}/CONTRIBUTING.rst

@@ -76,7 +76,7 @@ installed, this is how you set up your fork for local development::
 
     $ mkvirtualenv dateparser
     $ cd dateparser/
-    $ python setup.py develop
+    $ pip install -e .
 
 4. Create a branch for local development::
 
@@ -179,16 +179,3 @@ Whenever the content of
 the corresponding documentation table::
 
     dateparser_scripts/update_supported_languages_and_locales.py
-
-
-Updating the Timezone Cache
-----------------------------------------------------
-
-Whenever the content of
-``dateparser/timezones.py`` is modified you need to rebuild the timezone cache.
-
-Run this command:
-``BUILD_TZ_CACHE=1 python -c "import dateparser"``
-
-which should update
-``dateparser/data/dateparser_tz_cache.pkl``

{dateparser-1.3.0 → dateparser-1.4.0}/HISTORY.rst

@@ -3,6 +3,36 @@
 History
 =======
 
+1.4.0 (2026-03-26)
+------------------
+
+Security fixes:
+
+- Remove import-time loading of timezone offset data from pickle to prevent
+  unsafe deserialization from packaged data
+- Replace ``eval()`` use when parsing ``no_word_spacing`` with strict boolean
+  parsing to prevent code execution from locale metadata (#1056)
+
+New features:
+
+- Add support for expressions like "N {interval} from now" in English (#1271)
+- Add support for the ``en-US`` locale (#1222)
+
+Fixes:
+
+- Honor ``REQUIRE_PARTS`` for ambiguous month-number inputs by retrying with a
+  year-biased ``DATE_ORDER`` (#1298)
+- Fix parsing word-number relative phrases such as "two days later" (#1316)
+- Allow md5hash to work in FIPS environments (#1267)
+
+Improvements:
+
+- Add Bosnian Cyrillic (ijekavica) date translations (#1293)
+- Add a new browser-based demo to the project documentation (#1306)
+- Update installation documentation to replace ``setup.py install`` guidance
+  (#1310)
+- Add a project security policy (#1318)
+
 1.3.0 (2026-02-04)
 ------------------
 

{dateparser-1.3.0 → dateparser-1.4.0}/MANIFEST.in

@@ -3,7 +3,6 @@ include CONTRIBUTING.rst
 include HISTORY.rst
 include LICENSE
 include README.rst
-include dateparser/data/dateparser_tz_cache.pkl
 include dateparser_data/settings.py
 include requirements.txt
 

{dateparser-1.3.0 → dateparser-1.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dateparser
-Version: 1.3.0
+Version: 1.4.0
 Summary: Date parsing library designed to parse dates from HTML pages
 Home-page: https://github.com/scrapinghub/dateparser
 Author: Scrapinghub
@@ -330,6 +330,36 @@ To be able to use them you need to install the `calendar` extra by typing:
 History
 =======
 
+1.4.0 (2026-03-26)
+------------------
+
+Security fixes:
+
+- Remove import-time loading of timezone offset data from pickle to prevent
+  unsafe deserialization from packaged data
+- Replace ``eval()`` use when parsing ``no_word_spacing`` with strict boolean
+  parsing to prevent code execution from locale metadata (#1056)
+
+New features:
+
+- Add support for expressions like "N {interval} from now" in English (#1271)
+- Add support for the ``en-US`` locale (#1222)
+
+Fixes:
+
+- Honor ``REQUIRE_PARTS`` for ambiguous month-number inputs by retrying with a
+  year-biased ``DATE_ORDER`` (#1298)
+- Fix parsing word-number relative phrases such as "two days later" (#1316)
+- Allow md5hash to work in FIPS environments (#1267)
+
+Improvements:
+
+- Add Bosnian Cyrillic (ijekavica) date translations (#1293)
+- Add a new browser-based demo to the project documentation (#1306)
+- Update installation documentation to replace ``setup.py install`` guidance
+  (#1310)
+- Add a project security policy (#1318)
+
 1.3.0 (2026-02-04)
 ------------------
 

{dateparser-1.3.0 → dateparser-1.4.0}/README.rst

@@ -5,7 +5,7 @@
     <h1 align="center">
         <br/>
         <a href="https://github.com/scrapinghub/dateparser">
-            <img src="https://raw.githubusercontent.com/scrapinghub/dateparser/master/artwork/dateparser-logo.png" alt="Dateparser" width="500">
+            <img src="artwork/dateparser-logo.png" alt="Dateparser" width="500">
         </a>
         <br/>
     </h1>
@@ -64,7 +64,7 @@ Online demo
 -----------
 
 Do you want to try it out without installing any dependency? Now you can test
-it quickly by visiting `this online demo <https://dateparser-demo.netlify.app/>`__!
+it quickly by visiting `this online demo <https://gnyman.github.io/dateparser-demo-app/>`__!
 
 
 

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "1.3.0"
+__version__ = "1.4.0"
 
 from .conf import apply_settings
 from .date import DateDataParser

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/conf.py

@@ -52,7 +52,9 @@ class Settings:
             return "default"
 
         keys = sorted(["%s-%s" % (key, str(settings[key])) for key in settings])
-        return hashlib.md5("".join(keys).encode("utf-8")).hexdigest()
+        return hashlib.md5(
+            "".join(keys).encode("utf-8"), usedforsecurity=False
+        ).hexdigest()
 
     @classmethod
     def _get_settings_from_pyfile(cls):

dateparser-1.4.0/dateparser/data/date_translation_data/bs-Cyrl.py

@@ -0,0 +1,294 @@
+info = {
+    "name": "bs-Cyrl",
+    "date_order": "DMY",
+    "january": [
+        "јан",
+        "јануар"
+    ],
+    "february": [
+        "феб",
+        "фебруар"
+    ],
+    "march": [
+        "мар",
+        "март"
+    ],
+    "april": [
+        "апр",
+        "април"
+    ],
+    "may": [
+        "мај"
+    ],
+    "june": [
+        "јун",
+        "јуни"
+    ],
+    "july": [
+        "јул",
+        "јули"
+    ],
+    "august": [
+        "авг",
+        "август"
+    ],
+    "september": [
+        "сеп",
+        "септембар"
+    ],
+    "october": [
+        "окт",
+        "октобар"
+    ],
+    "november": [
+        "нов",
+        "новембар"
+    ],
+    "december": [
+        "дец",
+        "децембар"
+    ],
+    "monday": [
+        "пон",
+        "понедељак",
+        "понедјељак"
+    ],
+    "tuesday": [
+        "уто",
+        "уторак"
+    ],
+    "wednesday": [
+        "сри",
+        "сриједа"
+    ],
+    "thursday": [
+        "чет",
+        "четвртак"
+    ],
+    "friday": [
+        "пет",
+        "петак"
+    ],
+    "saturday": [
+        "суб",
+        "субота"
+    ],
+    "sunday": [
+        "нед",
+        "недеља",
+        "недјеља"
+    ],
+    "am": [
+        "пре подне",
+        "пријеподне"
+    ],
+    "pm": [
+        "поподне",
+        "послијеподне"
+    ],
+    "year": [
+        "година",
+        "г",
+        "год",
+        "годинa"
+    ],
+    "month": [
+        "месец",
+        "мј",
+        "мјесец",
+        "мјесеци",
+        "мјесеца"
+    ],
+    "week": [
+        "недеља",
+        "сед",
+        "седмица",
+        "седмице"
+    ],
+    "day": [
+        "дан"
+    ],
+    "hour": [
+        "час",
+        "сат"
+    ],
+    "minute": [
+        "минут",
+        "мин",
+        "минута"
+    ],
+    "second": [
+        "секунд",
+        "с",
+        "сек",
+        "секунда"
+    ],
+    "relative-type": {
+        "0 day ago": [
+            "данас"
+        ],
+        "0 hour ago": [
+            "this hour",
+            "овај сат"
+        ],
+        "0 minute ago": [
+            "this minute",
+            "ова минута"
+        ],
+        "0 month ago": [
+            "овог месеца",
+            "овај мјесец"
+        ],
+        "0 second ago": [
+            "now",
+            "сада"
+        ],
+        "0 week ago": [
+            "ове недеље",
+            "ове седмице"
+        ],
+        "0 year ago": [
+            "ове године"
+        ],
+        "1 day ago": [
+            "јуче"
+        ],
+        "1 month ago": [
+            "прошлог месеца",
+            "прошли мјесец"
+        ],
+        "1 week ago": [
+            "прошле недеље",
+            "прошле седмице"
+        ],
+        "1 year ago": [
+            "прошле године"
+        ],
+        "in 1 day": [
+            "сутра"
+        ],
+        "in 1 month": [
+            "следећег месеца",
+            "сљедећи мјесец"
+        ],
+        "in 1 week": [
+            "следеће недеље",
+            "сљедеће седмице"
+        ],
+        "in 1 year": [
+            "следеће године",
+            "сљедеће године"
+        ]
+    },
+    "relative-type-regex": {
+        "\\1 day ago": [
+            "пре (\\d+[.,]?\\d*) дан",
+            "пре (\\d+[.,]?\\d*) дана",
+            "прије (\\d+[.,]?\\d*) дан",
+            "прије (\\d+[.,]?\\d*) дана"
+        ],
+        "\\1 hour ago": [
+            "пре (\\d+[.,]?\\d*) сат",
+            "пре (\\d+[.,]?\\d*) сати",
+            "прије (\\d+[.,]?\\d*) сат",
+            "прије (\\d+[.,]?\\d*) сати"
+        ],
+        "\\1 minute ago": [
+            "пре (\\d+[.,]?\\d*) минут",
+            "пре (\\d+[.,]?\\d*) минута",
+            "прије (\\d+[.,]?\\d*) мин",
+            "прије (\\d+[.,]?\\d*) минута",
+            "прије (\\d+[.,]?\\d*) минуту"
+        ],
+        "\\1 month ago": [
+            "пре (\\d+[.,]?\\d*) месец",
+            "пре (\\d+[.,]?\\d*) месеци",
+            "прије (\\d+[.,]?\\d*) мј",
+            "прије (\\d+[.,]?\\d*) мјесец",
+            "прије (\\d+[.,]?\\d*) мјесеца",
+            "прије (\\d+[.,]?\\d*) мјесеци"
+        ],
+        "\\1 second ago": [
+            "пре (\\d+[.,]?\\d*) секунд",
+            "пре (\\d+[.,]?\\d*) секунди",
+            "прије (\\d+[.,]?\\d*) сек",
+            "прије (\\d+[.,]?\\d*) секунда",
+            "прије (\\d+[.,]?\\d*) секунду"
+        ],
+        "\\1 week ago": [
+            "пре (\\d+[.,]?\\d*) недеља",
+            "пре (\\d+[.,]?\\d*) недељу",
+            "прије (\\d+[.,]?\\d*) сед",
+            "прије (\\d+[.,]?\\d*) седмица",
+            "прије (\\d+[.,]?\\d*) седмице",
+            "прије (\\d+[.,]?\\d*) седмицу"
+        ],
+        "\\1 year ago": [
+            "пре (\\d+[.,]?\\d*) година",
+            "пре (\\d+[.,]?\\d*) годину",
+            "прије (\\d+[.,]?\\d*) г",
+            "прије (\\d+[.,]?\\d*) год",
+            "прије (\\d+[.,]?\\d*) година",
+            "прије (\\d+[.,]?\\d*) годину"
+        ],
+        "in \\1 day": [
+            "за (\\d+[.,]?\\d*) дан",
+            "за (\\d+[.,]?\\d*) дана"
+        ],
+        "in \\1 hour": [
+            "за (\\d+[.,]?\\d*) сат",
+            "за (\\d+[.,]?\\d*) сати"
+        ],
+        "in \\1 minute": [
+            "за (\\d+[.,]?\\d*) минут",
+            "за (\\d+[.,]?\\d*) минута",
+            "за (\\d+[.,]?\\d*) мин",
+            "за (\\d+[.,]?\\d*) минута",
+            "за (\\d+[.,]?\\d*) минуту"
+        ],
+        "in \\1 month": [
+            "за (\\d+[.,]?\\d*) месец",
+            "за (\\d+[.,]?\\d*) месеци",
+            "за (\\d+[.,]?\\d*) мј",
+            "за (\\d+[.,]?\\d*) мјесец",
+            "за (\\d+[.,]?\\d*) мјесеца",
+            "за (\\d+[.,]?\\d*) мјесеци"
+        ],
+        "in \\1 second": [
+            "за (\\d+[.,]?\\d*) секунд",
+            "за (\\d+[.,]?\\d*) секунди",
+            "за (\\d+[.,]?\\d*) сек",
+            "за (\\d+[.,]?\\d*) секунда",
+            "за (\\d+[.,]?\\d*) секунду"
+        ],
+        "in \\1 week": [
+            "за (\\d+[.,]?\\d*) недеља",
+            "за (\\d+[.,]?\\d*) недељу",
+            "за (\\d+[.,]?\\d*) сед",
+            "за (\\d+[.,]?\\d*) седмица",
+            "за (\\d+[.,]?\\d*) седмицу",
+            "за (\\d+[.,]?\\d*) седмице"
+        ],
+        "in \\1 year": [
+            "за (\\d+[.,]?\\d*) година",
+            "за (\\d+[.,]?\\d*) годину",
+            "за (\\d+[.,]?\\d*) г",
+            "за (\\d+[.,]?\\d*) год"
+        ]
+    },
+    "locale_specific": {},
+    "skip": [
+        " ",
+        "'",
+        ",",
+        "-",
+        ".",
+        "/",
+        ";",
+        "@",
+        "[",
+        "]",
+        "|",
+        "，"
+    ]
+}

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/data/date_translation_data/en.py

@@ -718,6 +718,9 @@ info = {
         "en-UM": {
             "name": "en-UM"
         },
+        "en-US": {
+            "name": "en-US"
+        },
         "en-VC": {
             "name": "en-VC",
             "date_order": "DMY"
@@ -756,6 +759,7 @@ info = {
         "and",
         "at",
         "by",
+        "from",
         "just",
         "m",
         "nd",
@@ -812,13 +816,13 @@ info = {
             "(\\d+[.,]?\\d*)h(\\d+[.,]?\\d*)": "\\1:\\2"
         },
         {
-            "(?<=from\\s+)now": "in"
+            "from\\s+now": "in"
         },
         {
             "less than 1 minute ago": "45 second ago"
         },
         {
-            "(\\d+[.,]?\\d*) (decade|year|month|week|day|hour|minute|second)s? later": "in \\1 \\2"
+            "(\\d+[.,]?\\d*|one|two|three|four|five|six|seven|eight|nine|ten|eleven|twelve) (decade|year|month|week|day|hour|minute|second)s? later": "in \\1 \\2"
         },
         {
             "one": "1"

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/data/languages_info.py

@@ -496,6 +496,7 @@ language_locale_dict = {
         "en-TZ",
         "en-UG",
         "en-UM",
+        "en-US",
         "en-VC",
         "en-VG",
         "en-VI",

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/date.py

@@ -275,26 +275,48 @@ class _DateLocaleParser:
         return self._try_parser(parse_method=_parse_nospaces)
 
     def _try_parser(self, parse_method):
-        _order = self._settings.DATE_ORDER
+        original_order = self._settings.DATE_ORDER
+
+        # Use locale date order unless DATE_ORDER was explicitly set by the caller.
+        if (
+            self._settings.PREFER_LOCALE_DATE_ORDER
+            and "DATE_ORDER" not in self._settings._mod_settings
+        ):
+            first_order = self.locale.info.get("date_order", original_order)
+        else:
+            first_order = original_order
+
+        candidates = [first_order]
+
+        # If the caller requires a year (and not a day) and did not set DATE_ORDER,
+        # retry once or twice with year-biased orders to resolve month-number ambiguity.
+        require_parts = set(getattr(self._settings, "REQUIRE_PARTS", None) or [])
+        if (
+            "DATE_ORDER" not in self._settings._mod_settings
+            and "year" in require_parts
+            and "day" not in require_parts
+        ):
+            for order in ("MYD", "YMD"):
+                if order not in candidates:
+                    candidates.append(order)
+
+        translated = self._get_translated_date()
+
         try:
-            if self._settings.PREFER_LOCALE_DATE_ORDER:
-                if "DATE_ORDER" not in self._settings._mod_settings:
-                    self._settings.DATE_ORDER = self.locale.info.get(
-                        "date_order", _order
+            for order in candidates:
+                self._settings.DATE_ORDER = order
+                try:
+                    date_obj, period = date_parser.parse(
+                        translated,
+                        parse_method=parse_method,
+                        settings=self._settings,
                     )
-            date_obj, period = date_parser.parse(
-                self._get_translated_date(),
-                parse_method=parse_method,
-                settings=self._settings,
-            )
-            self._settings.DATE_ORDER = _order
-            return DateData(
-                date_obj=date_obj,
-                period=period,
-            )
-        except ValueError:
-            self._settings.DATE_ORDER = _order
+                    return DateData(date_obj=date_obj, period=period)
+                except ValueError:
+                    continue
             return None
+        finally:
+            self._settings.DATE_ORDER = original_order
 
     def _try_given_formats(self):
         if not self.date_formats:

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/languages/dictionary.py

@@ -51,6 +51,14 @@ class UnknownTokenError(Exception):
     pass
 
 
+def _parse_bool(value):
+    if isinstance(value, bool):
+        return value
+    if isinstance(value, str):
+        return value.strip().lower() in ("true", "1", "yes")
+    return False
+
+
 class Dictionary:
     """
     Class that modifies and stores translations and handles splitting of date string.
@@ -101,8 +109,8 @@ class Dictionary:
 
         self._dictionary = dictionary
 
-        no_word_spacing = locale_info.get("no_word_spacing", "False")
-        self._no_word_spacing = bool(eval(no_word_spacing))
+        no_word_spacing = locale_info.get("no_word_spacing", False)
+        self._no_word_spacing = _parse_bool(no_word_spacing)
 
         relative_type_regex = locale_info.get("relative-type-regex", {})
         self._relative_strings = list(chain.from_iterable(relative_type_regex.values()))

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/languages/locale.py

@@ -12,6 +12,14 @@ from .dictionary import ALWAYS_KEEP_TOKENS, Dictionary, NormalizedDictionary
 NUMERAL_PATTERN = re.compile(r"(\d+)", re.U)
 
 
+def _parse_bool(value):
+    if isinstance(value, bool):
+        return value
+    if isinstance(value, str):
+        return value.strip().lower() in ("true", "1", "yes")
+    return False
+
+
 class Locale:
     """
     Class that deals with applicability and translation from a locale.
@@ -459,7 +467,7 @@ class Locale:
         return date_string
 
     def _get_simplifications(self, settings=None):
-        no_word_spacing = eval(self.info.get("no_word_spacing", "False"))
+        no_word_spacing = _parse_bool(self.info.get("no_word_spacing", False))
         if settings.NORMALIZE:
             if self._normalized_simplifications is None:
                 self._normalized_simplifications = []

{dateparser-1.3.0 → dateparser-1.4.0}/dateparser/timezone_parser.py

@@ -1,8 +1,4 @@
-import os
-import pickle
-import zlib
 from datetime import datetime, timedelta, timezone, tzinfo
-from pathlib import Path
 
 import regex as re
 
@@ -88,47 +84,8 @@ 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)