Merge branch 'scrapinghub:master' into master

.flake8

@@ -1,7 +1,7 @@
 [flake8]
 
 max-line-length = 119
-ignore = W503, E203, E501, E722, F401
+ignore = W503, E203, E501, E722, F401, E701, E704
 
 exclude =
     docs/conf.py

.github/workflows/main.yml

@@ -14,15 +14,9 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          - python-version: 3.8
+          - python-version: 3.9
             env:
               TOXENV: flake8
-          - python-version: 3.7
-            env:
-              TOXENV: py
-          - python-version: 3.8
-            env:
-              TOXENV: py
           - python-version: 3.9
             env:
               TOXENV: py
@@ -38,30 +32,44 @@ jobs:
           - python-version: "3.11"
             env:
               TOXENV: latest
-          - python-version: "3.11"
+          - python-version: "3.12"
+            env:
+              TOXENV: py
+          - python-version: "3.12"
+            env:
+              TOXENV: latest
+          - python-version: "3.13"
+            env:
+              TOXENV: py
+          - python-version: "3.13"
+            env:
+              TOXENV: latest
+          - python-version: "3.13"
             env:
               TOXENV: twinecheck
-          - python-version: "3.11"  # Keep in sync with tox.ini
+          - python-version: "3.12"  # Keep in sync with tox.ini
             env:
               TOXENV: docs
     steps:
       - uses: actions/checkout@v3
       - name: 'Set up Python ${{ matrix.python-version }}'
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         with:
-          python-version: '${{ matrix.python-version }}'
+          python-version: ${{ matrix.python-version }}
       - name: Install language-pack-fr
-        run: sudo apt-get update && sudo apt-get install language-pack-fr
-      - name: Install python dependencies
-        run: pip install tox
-      - name: tox
-        run: tox -e py
+        run: sudo apt-get update && sudo apt-get install -y language-pack-fr build-essential
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install tox
+      - name: Run tests
+        run: tox -e ${{ matrix.env.TOXENV }}
       - name: Upload coverage.xml to codecov
-        if: ${{ matrix.python-version == '3.9' && matrix.env.TOXENV == 'latest'}}
+        if: ${{ matrix.env.python-version == '3.9' && matrix.env.TOXENV == 'latest' }}
         uses: codecov/codecov-action@v3
 
   pre-commit:
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v3
-    - uses: pre-commit/[email protected]
+      - uses: actions/checkout@v3
+      - uses: pre-commit/[email protected]

.github/workflows/publish.yml

@@ -11,7 +11,7 @@ jobs:
     steps:
       - uses: actions/checkout@v3
       - name: Set up Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         with:
           python-version: "3.x"
       - name: Install dependencies

README.rst

@@ -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/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/parser.py

@@ -598,10 +598,13 @@ def _correct_for_month(self, dateobj):
         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
@@ -613,11 +616,13 @@ def parse(cls, datestring, settings, tz=None):
         # 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

dateparser/timezone_parser.py

@@ -72,9 +72,8 @@ def get_offset(tz_obj, regex, repl="", replw=""):
                 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)
 

setup.cfg

@@ -21,3 +21,11 @@ ignore =
     dateparser/freshness_date_parser.py E722
     dateparser/parser.py E722
     dateparser/docs/conf.py E402
+
+    # Additional ignored codes
+    E203
+    E501
+    E722
+    F401
+    E701
+    E704

setup.py

@@ -48,18 +48,18 @@
     license="BSD",
     zip_safe=False,
     keywords="dateparser",
-    python_requires=">=3.7",
+    python_requires=">=3.8",  # Python 3.8 is required for fuzzing
     classifiers=[
         "Development Status :: 5 - Production/Stable",
         "Intended Audience :: Developers",
         "License :: OSI Approved :: BSD License",
         "Natural Language :: English",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.7",
-        "Programming Language :: Python :: 3.8",
         "Programming Language :: Python :: 3.9",
         "Programming Language :: Python :: 3.10",
         "Programming Language :: Python :: 3.11",
+        "Programming Language :: Python :: 3.12",
+        "Programming Language :: Python :: 3.13",
         "Programming Language :: Python :: Implementation :: CPython",
     ],
 )

tests/test_clean_api.py

@@ -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, 12, 31, 14, 4),
+                expected_date=datetime(1969, 1, 31, 14, 4),
             )
         ]
     )

tests/test_date_parser.py

@@ -1265,6 +1265,66 @@ def test_prefer_dates_from_with_timezone(
         self.then_date_was_parsed_by_date_parser()
         self.then_date_obj_exactly_is(expected)
 
+    @parameterized.expand(
+        [
+            param(
+                "2015",
+                prefer_day="current",
+                prefer_month="current",
+                today=datetime(2010, 2, 10),
+                expected=datetime(2015, 2, 10),
+            ),
+            param(
+                "2015",
+                prefer_day="last",
+                prefer_month="current",
+                today=datetime(2010, 2, 10),
+                expected=datetime(2015, 2, 28),
+            ),
+            param(
+                "2015",
+                prefer_day="first",
+                prefer_month="current",
+                today=datetime(2010, 2, 10),
+                expected=datetime(2015, 2, 1),
+            ),
+            param(
+                "2015",
+                prefer_day="current",
+                prefer_month="last",
+                today=datetime(2010, 2, 10),
+                expected=datetime(2015, 12, 10),
+            ),
+            param(
+                "2015",
+                prefer_day="last",
+                prefer_month="last",
+                today=datetime(2010, 2, 10),
+                expected=datetime(2015, 12, 31),
+            ),
+            param(
+                "2020",  # Leap year last day test
+                prefer_day="last",
+                prefer_month="current",
+                today=datetime(2010, 2, 10),
+                expected=datetime(2020, 2, 29),
+            ),
+        ]
+    )
+    def test_dates_with_no_day_or_month(
+        self, date_string, prefer_day, prefer_month, today=None, expected=None
+    ):
+        self.given_parser(
+            settings={
+                "PREFER_DAY_OF_MONTH": prefer_day,
+                "PREFER_MONTH_OF_YEAR": prefer_month,
+                "RELATIVE_BASE": today,
+            }
+        )
+        self.when_date_is_parsed(date_string)
+        self.then_date_was_parsed_by_date_parser()
+        self.then_date_obj_exactly_is(expected)
+
     def given_local_tz_offset(self, offset):
         self.add_patch(
             patch.object(

tests/test_search.py

@@ -410,7 +410,7 @@ def test_search_date_string(self, shortname, datetime_string):
                 "Die UdSSR blieb gemäß dem Neutralitätspakt "
                 "vom 13. April 1941 gegenüber Japan vorerst neutral.",
                 [
-                    ("Die", datetime.datetime(1999, 12, 28, 0, 0)),
+                    ("Die", datetime.datetime(1999, 1, 28, 0, 0)),
                     ("13. April 1941", datetime.datetime(1941, 4, 13, 0, 0)),
                 ],
                 settings={"RELATIVE_BASE": datetime.datetime(2000, 1, 1)},
@@ -825,7 +825,10 @@ def test_splitting_of_not_parsed(self, shortname, string, expected, settings=Non
                 "бомбардировки срещу Япония, използувайки новозавладените острови като бази.",
             ),
             # Chinese
-            param("zh", "不過大多數人仍多把第二次世界大戰的爆發定為1939年9月1日德國入侵波蘭開始，2015年04月08日10点05。"),
+            param(
+                "zh",
+                "不過大多數人仍多把第二次世界大戰的爆發定為1939年9月1日德國入侵波蘭開始，2015年04月08日10点05。",
+            ),
             # Czech
             param(
                 "cs",
@@ -897,7 +900,10 @@ def test_splitting_of_not_parsed(self, shortname, string, expected, settings=Non
                 "d'Etiopia. Il 9 maggio 1936 venne proclamato l'Impero. ",
             ),
             # Japanese
-            param("ja", "1933年（昭和8年）12月23日午前6時39分、宮城（現：皇居）内の産殿にて誕生。"),
+            param(
+                "ja",
+                "1933年（昭和8年）12月23日午前6時39分、宮城（現：皇居）内の産殿にて誕生。",
+            ),
             # Persian
             param("fa", "نگ جهانی دوم جنگ جدی بین سپتامبر 1939 و 2 سپتامبر 1945 بود."),
             # Polish