dateparser 1.2.0__tar.gz → 1.2.2__tar.gz

{dateparser-1.2.0 → dateparser-1.2.2}/CONTRIBUTING.rst

@@ -179,3 +179,16 @@ 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-1.2.0 → dateparser-1.2.2}/HISTORY.rst

@@ -3,6 +3,43 @@
 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)
+------------------
+
+Fixes:
+
+- Fix PytzUsageWarning (#1109)
+- Fix date_parser with prefer_month_of_year wrong results (#1224)
+- Fix skipped day when UTC and tz are different days (#1183)
+
+Improvements:
+
+- Avoid repeated loop over timezones (#1238)
+- Proofread README.rst (#1234)
+- Check for derived types for configuration (#1223)
+- Parse some abbreviated strings as relative dates (#1219)
+- Migrate from hijri-converter to hijridate (#1211)
+- Fixed ClusterFuzz build error by adding dateparser.data as a binary (#1208)
+- Fix an issue detected by OSSFuzz (#1203)
+- Support two-digit years in non-Gregorian calendars (#1187)
+- Refactored CI to run extras separately and test minimum versions of dependencies, replaced flake8 with ruff, fixed tests (#1248)
+- Set minimum versions for dependencies (#1248)
+- Limited ``numpy`` to 1.x when installing ``dateparser[fasttext]`` (#1248)
+
+
 1.2.0 (2023-11-17)
 ------------------
 

{dateparser-1.2.0 → dateparser-1.2.2}/MANIFEST.in

@@ -3,6 +3,7 @@ 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-1.2.0 → dateparser-1.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: dateparser
-Version: 1.2.0
+Version: 1.2.2
 Summary: Date parsing library designed to parse dates from HTML pages
 Home-page: https://github.com/scrapinghub/dateparser
 Author: Scrapinghub
@@ -13,26 +13,40 @@ Classifier: Intended Audience :: Developers
 Classifier: License :: OSI Approved :: BSD License
 Classifier: Natural Language :: English
 Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.7
-Classifier: Programming Language :: Python :: 3.8
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Classifier: Programming Language :: Python :: Implementation :: CPython
-Requires-Python: >=3.7
+Requires-Python: >=3.8
 License-File: LICENSE
 License-File: AUTHORS.rst
-Requires-Dist: python-dateutil
-Requires-Dist: pytz
-Requires-Dist: regex!=2019.02.19,!=2021.8.27
-Requires-Dist: tzlocal
+Requires-Dist: python-dateutil>=2.7.0
+Requires-Dist: pytz>=2024.2
+Requires-Dist: regex>=2024.9.11
+Requires-Dist: tzlocal>=0.2
 Provides-Extra: calendars
-Requires-Dist: hijri-converter; extra == "calendars"
-Requires-Dist: convertdate; extra == "calendars"
+Requires-Dist: convertdate>=2.2.1; extra == "calendars"
+Requires-Dist: hijridate; extra == "calendars"
 Provides-Extra: fasttext
-Requires-Dist: fasttext; extra == "fasttext"
+Requires-Dist: fasttext>=0.9.1; extra == "fasttext"
+Requires-Dist: numpy<2,>=1.19.3; extra == "fasttext"
 Provides-Extra: langdetect
-Requires-Dist: langdetect; extra == "langdetect"
+Requires-Dist: langdetect>=1.0.0; extra == "langdetect"
+Dynamic: author
+Dynamic: author-email
+Dynamic: classifier
+Dynamic: description
+Dynamic: home-page
+Dynamic: keywords
+Dynamic: license
+Dynamic: license-file
+Dynamic: project-url
+Dynamic: provides-extra
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary
 
 ==========================
 Introduction to dateparser
@@ -259,13 +273,13 @@ Dependencies
 
   * dateutil_'s module ``relativedelta`` for its freshness parser.
   * convertdate_ to convert *Jalali* dates to *Gregorian*.
-  * hijri-converter_ to convert *Hijri* dates to *Gregorian*.
+  * hijridate_ to convert *Hijri* dates to *Gregorian*.
   * tzlocal_ to reliably get local timezone.
   * ruamel.yaml_ (optional) for operations on language files.
 
 .. _dateutil: https://pypi.python.org/pypi/python-dateutil
 .. _convertdate: https://pypi.python.org/pypi/convertdate
-.. _hijri-converter: https://pypi.python.org/pypi/hijri-converter
+.. _hijridate: https://pypi.python.org/pypi/hijridate
 .. _tzlocal: https://pypi.python.org/pypi/tzlocal
 .. _ruamel.yaml: https://pypi.python.org/pypi/ruamel.yaml
 
@@ -297,7 +311,7 @@ To be able to use them you need to install the `calendar` extra by typing:
     >>> HijriCalendar('17-01-1437 هـ 08:30 مساءً').get_date()
     DateData(date_obj=datetime.datetime(2015, 10, 30, 20, 30), period='day', locale=None)
 
-.. note:: `HijriCalendar` only works with Python ≥ 3.6.
+.. note:: `HijriCalendar` only works with Python ≥ 3.7.
 
 
 .. :changelog:
@@ -305,6 +319,43 @@ 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)
+------------------
+
+Fixes:
+
+- Fix PytzUsageWarning (#1109)
+- Fix date_parser with prefer_month_of_year wrong results (#1224)
+- Fix skipped day when UTC and tz are different days (#1183)
+
+Improvements:
+
+- Avoid repeated loop over timezones (#1238)
+- Proofread README.rst (#1234)
+- Check for derived types for configuration (#1223)
+- Parse some abbreviated strings as relative dates (#1219)
+- Migrate from hijri-converter to hijridate (#1211)
+- Fixed ClusterFuzz build error by adding dateparser.data as a binary (#1208)
+- Fix an issue detected by OSSFuzz (#1203)
+- Support two-digit years in non-Gregorian calendars (#1187)
+- Refactored CI to run extras separately and test minimum versions of dependencies, replaced flake8 with ruff, fixed tests (#1248)
+- Set minimum versions for dependencies (#1248)
+- Limited ``numpy`` to 1.x when installing ``dateparser[fasttext]`` (#1248)
+
+
 1.2.0 (2023-11-17)
 ------------------
 

{dateparser-1.2.0 → dateparser-1.2.2}/README.rst

@@ -50,7 +50,7 @@ Key Features
    locales <https://dateparser.readthedocs.io/en/latest/supported_locales.html>`__.
 -  Language autodetection
 -  Customizable behavior through
-   `settings <https://dateparser.readthedocs.io/en/latest/usage.html#settings>`__.
+   `settings <https://dateparser.readthedocs.io/en/latest/settings.html>`__.
 -  Support for `non-Gregorian calendar
    systems <https://dateparser.readthedocs.io/en/latest/introduction.html#supported-calendars>`__.
 -  Support for dates with timezones abbreviations or UTC offsets
@@ -144,8 +144,8 @@ For that reason it is important that the input is a valid date, otherwise it cou
 
 To reduce the possibility of receiving false positives, make sure that:
 
-- The input string it's a valid date and it doesn't contain any other words or numbers.
-- If you know the language or languages beforehand you add them through the ``languages`` or ``locales`` properties.
+- The input string is a valid date and doesn't contain any other words or numbers.
+- If you know the language or languages beforehand, you add them through the ``languages`` or ``locales`` properties.
 
 
 On the other hand, if you want to exclude any of the default parsers
@@ -156,7 +156,7 @@ are executed, you can do so through the
 Installation
 ------------
 
-Dateparser supports Python >= 3.7. You can install it by doing:
+Dateparser supports Python >= 3.9. You can install it by doing:
 
 ::
 
@@ -172,7 +172,7 @@ If you want to use the jalali or hijri calendar, you need to install the
 Common use cases
 ----------------
 
-**dateparser** can be used with a really different number of purposes,
+**dateparser** can be used for a wide variety of purposes,
 but it stands out when it comes to:
 
 Consuming data from different sources:
@@ -191,7 +191,7 @@ Offering natural interaction with users:
 
 -  **Tooling and CLI**: allow users to write “3 days ago” to retrieve
    information.
--  **Search engine**: allow people to search by date in an easiest /
+-  **Search engine**: allow people to search by date in an easy /
    natural format.
 -  **Bots**: allow users to interact with a bot easily
 

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "1.2.0"
+__version__ = "1.2.2"
 
 from .conf import apply_settings
 from .date import DateDataParser

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/calendars/__init__.py

@@ -76,6 +76,9 @@ class non_gregorian_parser(_parser):
 
         return result
 
+    def handle_two_digit_year(self, year):
+        raise ValueError
+
     def _get_datetime_obj(self, **params):
         day = params["day"]
         year = params["year"]
@@ -129,6 +132,8 @@ class non_gregorian_parser(_parser):
             day = int(token)
         elif directive == "%Y" and token_len == 4 and is_digit:
             year = int(token)
+        elif directive == "%Y" and token_len == 2 and is_digit:
+            year = self.handle_two_digit_year(int(token))
         else:
             raise ValueError
         return self.non_gregorian_date_cls(year, month, day)

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/calendars/hijri_parser.py

@@ -1,4 +1,4 @@
-from hijri_converter import convert
+from hijridate import Gregorian, Hijri
 
 from dateparser.calendars import non_gregorian_parser
 
@@ -6,19 +6,17 @@ from dateparser.calendars import non_gregorian_parser
 class hijri:
     @classmethod
     def to_gregorian(cls, year=None, month=None, day=None):
-        g = convert.Hijri(
-            year=year, month=month, day=day, validate=False
-        ).to_gregorian()
+        g = Hijri(year=year, month=month, day=day, validate=False).to_gregorian()
         return g.datetuple()
 
     @classmethod
     def from_gregorian(cls, year=None, month=None, day=None):
-        h = convert.Gregorian(year, month, day).to_hijri()
+        h = Gregorian(year, month, day).to_hijri()
         return h.datetuple()
 
     @classmethod
     def month_length(cls, year, month):
-        h = convert.Hijri(year=year, month=month, day=1)
+        h = Hijri(year=year, month=month, day=1)
         return h.month_length()
 
 
@@ -54,3 +52,9 @@ class hijri_parser(non_gregorian_parser):
             for arabic in arabics:
                 result = result.replace(arabic, latin)
         return result
+
+    def handle_two_digit_year(self, year):
+        if year >= 90:
+            return year + 1300
+        else:
+            return year + 1400

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/calendars/jalali_parser.py

@@ -176,3 +176,9 @@ class jalali_parser(non_gregorian_parser):
         ):
             result = result.replace(persian_number, str(number))
         return result
+
+    def handle_two_digit_year(self, year):
+        if year > 60:
+            return year + 1300
+        else:
+            return year + 1400

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/conf.py

@@ -243,7 +243,7 @@ def check_settings(settings):
         setting_props = settings_values[setting_name]
 
         # check type:
-        if not setting_type == setting_props["type"]:
+        if not isinstance(setting_value, setting_props["type"]):
             raise SettingValidationError(
                 '"{}" must be "{}", not "{}".'.format(
                     setting_name, setting_props["type"].__name__, setting_type.__name__

dateparser-1.2.2/dateparser/data/__init__.py

@@ -0,0 +1,2 @@
+from dateparser.data import date_translation_data
+from .languages_info import language_order, language_locale_dict

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/data/date_translation_data/en.py

@@ -800,7 +800,7 @@ info = {
             "(?:12\\s+)?midnight": "00:00"
         },
         {
-            "(\\d+[.,]?\\d*)h(\\d+[.,]?\\d*)m?": "\\1:\\2"
+            "(\\d+[.,]?\\d*)h(\\d+[.,]?\\d*)": "\\1:\\2"
         },
         {
             "(?<=from\\s+)now": "in"

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/date.py

@@ -1,5 +1,4 @@
 import collections
-import sys
 from collections.abc import Set
 from datetime import datetime, timedelta
 

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/freshness_date_parser.py

@@ -31,7 +31,7 @@ class FreshnessDateDataParser:
         date_string = re.sub(r"\b(?:ago|in)\b", "", date_string)
         try:
             return time_parser(date_string)
-        except:
+        except Exception:
             pass
 
     def get_local_tz(self):
@@ -71,10 +71,7 @@ class FreshnessDateDataParser:
                         now = now.replace(tzinfo=ptz)
 
             if not now.tzinfo:
-                if hasattr(self.get_local_tz(), "localize"):
-                    now = self.get_local_tz().localize(now)
-                else:
-                    now = now.replace(tzinfo=self.get_local_tz())
+                now = now.replace(tzinfo=self.get_local_tz())
 
         elif ptz:
             localized_now = datetime.now(ptz)

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/languages/dictionary.py

@@ -181,26 +181,34 @@ class Dictionary:
         ):
             cache.pop(list(cache.keys())[0])
 
-    def _split_by_known_words(self, string, keep_formatting):
-        if not string:
-            return string
-
+    def _split_by_known_words(self, string: str, keep_formatting: bool):
         regex = self._get_split_regex_cache()
-        match = regex.match(string)
-        if not match:
-            return (
-                self._split_by_numerals(string, keep_formatting)
-                if self._should_capture(string, keep_formatting)
-                else []
-            )
-
-        unparsed, known, unknown = match.groups()
-        splitted = [known] if self._should_capture(known, keep_formatting) else []
-        if unparsed and self._should_capture(unparsed, keep_formatting):
-            splitted = self._split_by_numerals(unparsed, keep_formatting) + splitted
-        if unknown:
-            splitted.extend(self._split_by_known_words(unknown, keep_formatting))
-
+        splitted = []
+        unknown = string
+
+        while unknown:
+            match = regex.match(string)
+
+            if not match:
+                curr_split = (
+                    self._split_by_numerals(string, keep_formatting)
+                    if self._should_capture(string, keep_formatting)
+                    else []
+                )
+                unknown = ""
+            else:
+                unparsed, known, unknown = match.groups()
+                curr_split = (
+                    [known] if self._should_capture(known, keep_formatting) else []
+                )
+                if unparsed and self._should_capture(unparsed, keep_formatting):
+                    curr_split = (
+                        self._split_by_numerals(unparsed, keep_formatting) + curr_split
+                    )
+                if unknown:
+                    string = unknown if string != unknown else ""
+
+            splitted.extend(curr_split)
         return splitted
 
     def _split_by_numerals(self, string, keep_formatting):

{dateparser-1.2.0 → dateparser-1.2.2}/dateparser/parser.py

@@ -182,7 +182,7 @@ class _no_spaces_parser:
                 dt = strptime(datestring, fmt), cls._get_period(fmt)
                 if len(str(dt[0].year)) == 4:
                     return dt
-            except:
+            except Exception:
                 pass
         return None
 
@@ -216,7 +216,7 @@ class _no_spaces_parser:
                     missing = _get_missing_parts(fmt)
                     _check_strict_parsing(missing, settings)
                     return dt
-                except:
+                except Exception:
                     pass
         else:
             if ambiguous_date:
@@ -332,7 +332,7 @@ class _parser:
                     token.index(":")
                     # Is after period? raise ValueError if '.' can't be found:
                     self.tokens[self.tokens.index((token, 0)) + 1][0].index(".")
-                except:
+                except Exception:
                     microsecond = None
 
                 if microsecond:
@@ -342,7 +342,7 @@ class _parser:
                     meridian = MERIDIAN.search(
                         self.filtered_tokens[meridian_index][0]
                     ).group()
-                except:
+                except Exception:
                     meridian = None
 
                 if any([":" in token, meridian, microsecond]):
@@ -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(
@@ -512,11 +512,15 @@ 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
-        ), "`self.now` doesn't have `tzinfo`. Review comment in code for details."
+        assert not (self.now.tzinfo is None and dateobj.tzinfo is not None), (
+            "`self.now` doesn't have `tzinfo`. Review comment in code for details."
+        )
 
         # Store the original dateobj values so that upon subsequent parsing everything is not
         # treated as offset-aware if offset awareness is changed.
@@ -563,18 +567,15 @@ class _parser:
             # Convert dateobj to utc time to compare with self.now
             try:
                 tz = tz or get_timezone_from_tz_string(self.settings.TIMEZONE)
-            except pytz.UnknownTimeZoneError:
-                tz = None
+                tz_offset = tz.utcoffset(dateobj)
+            except (pytz.UnknownTimeZoneError, pytz.NonExistentTimeError):
+                tz_offset = timedelta(hours=0)
 
-            if tz:
-                dateobj_time = (dateobj - tz.utcoffset(dateobj)).time()
-            else:
-                dateobj_time = dateobj.time()
             if "past" in self.settings.PREFER_DATES_FROM:
-                if self.now.time() < dateobj_time:
+                if self.now < dateobj - tz_offset:
                     dateobj = dateobj + timedelta(days=-1)
             if "future" in self.settings.PREFER_DATES_FROM:
-                if self.now.time() > dateobj_time:
+                if self.now > dateobj - tz_offset:
                     dateobj = dateobj + timedelta(days=1)
 
         # Reset dateobj to the original value, thus removing any offset awareness that may
@@ -601,10 +602,13 @@ class _parser:
         relative_base_month = (
             relative_base.month if hasattr(relative_base, "month") else relative_base
         )
-        if getattr(self, "_token_month", None) or relative_base_month:
+
+        if getattr(self, "_token_month", None):
             return dateobj
 
-        dateobj = set_correct_month_from_settings(dateobj, self.settings)
+        dateobj = set_correct_month_from_settings(
+            dateobj, self.settings, relative_base_month
+        )
         return dateobj
 
     @classmethod
@@ -616,11 +620,13 @@ class _parser:
         # correction for past, future if applicable
         dateobj = po._correct_for_time_frame(dateobj, tz)
 
+        # correction for preference of month: beginning, current, end
+        # must happen before day so that day is derived from the correct month
+        dateobj = po._correct_for_month(dateobj)
+
         # correction for preference of day: beginning, current, end
         dateobj = po._correct_for_day(dateobj)
 
-        # correction for preference of month: beginning, current, end
-        dateobj = po._correct_for_month(dateobj)
         period = po._get_period()
 
         return dateobj, period
@@ -684,7 +690,7 @@ class _parser:
                                 (component, getattr(do, component)),
                                 ("day", prev_value),
                             ]
-                    except:
+                    except Exception:
                         pass
             else:
                 raise ValueError("Unable to parse: %s" % token)

{dateparser-1.2.0 → dateparser-1.2.2}/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-1.2.0 → dateparser-1.2.2}/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-1.2.0 → dateparser-1.2.2}/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
 
@@ -72,9 +76,8 @@ def build_tz_offsets(search_regex_parts):
                 search_regex_parts.append(tz_obj[0])
                 yield get_offset(tz_obj, regex)
 
-            # alternate patterns
-            for replace, replacewith in tz_info.get("replace", []):
-                for tz_obj in tz_info["timezones"]:
+                # alternate patterns
+                for replace, replacewith in tz_info.get("replace", []):
                     search_regex_parts.append(re.sub(replace, replacewith, tz_obj[0]))
                     yield get_offset(tz_obj, regex, repl=replace, replw=replacewith)
 
@@ -85,8 +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)