Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 13 additions & 6 deletions dateparser/data/date_translation_data/en.py
Original file line number Diff line number Diff line change
Expand Up @@ -51,32 +51,39 @@
],
"monday": [
"mon",
"monday"
"monday",
"mo"
],
"tuesday": [
"tue",
"tuesday",
"tu",
"Tues"
],
"wednesday": [
"wed",
"wednesday"
"wednesday",
"we"
],
"thursday": [
"thu",
"thursday"
"thursday",
"th"
],
"friday": [
"fri",
"friday"
"friday",
"fr"
],
"saturday": [
"sat",
"saturday"
"saturday",
"sa"
],
"sunday": [
"sun",
"sunday"
"sunday",
"su"
],
"am": [
"am"
Expand Down
13 changes: 12 additions & 1 deletion dateparser/languages/dictionary.py
Original file line number Diff line number Diff line change
Expand Up @@ -201,14 +201,25 @@ def _split_by_known_words(self, string: str, keep_formatting: bool):
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)
for token in curr_split:
if (
splitted
and splitted[-1].isdigit()
and token in {"st", "nd", "rd", "th"}
):
continue

splitted.append(token)

return splitted

def _split_by_numerals(self, string, keep_formatting):
Expand Down
1 change: 0 additions & 1 deletion dateparser/languages/locale.py
Original file line number Diff line number Diff line change
Expand Up @@ -144,7 +144,6 @@ def translate(self, date_string, keep_formatting=False, settings=None):
date_string_tokens[i] = dictionary[word] or fallback
if "in" in date_string_tokens:
date_string_tokens = self._clear_future_words(date_string_tokens)

return self._join(
list(filter(bool, date_string_tokens)),
separator="" if keep_formatting else " ",
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -3,9 +3,28 @@ pertain: ["of"]

sentence_splitter_group : 1

monday:
- mo

tuesday:
- tu
- Tues

wednesday:
- we

thursday:
- th

friday:
- fr

saturday:
- sa

sunday:
- su

september:
- sept

Expand Down
69 changes: 63 additions & 6 deletions tests/test_date.py
Original file line number Diff line number Diff line change
@@ -1,5 +1,6 @@
#!/usr/bin/env python

import datetime as real_datetime
import os
import unittest
from collections import OrderedDict
Expand All @@ -9,6 +10,7 @@
from time import tzset
from unittest.mock import Mock, patch

import pytest
from parameterized import param, parameterized

import dateparser
Expand Down Expand Up @@ -832,13 +834,68 @@ def test_get_date_tuple(self, date_string, expected_result):
self.when_get_date_tuple_is_called(date_string)
self.then_returned_tuple_is(expected_result)

@parameterized.expand(
[
param(
"Mo",
datetime(2025, 7, 28, 0, 0),
),
param(
"Tu",
datetime(2025, 7, 29, 0, 0),
),
param(
"We",
datetime(2025, 7, 30, 0, 0),
),
param(
"Th",
datetime(2025, 7, 31, 0, 0),
),
param(
"Fr",
datetime(2025, 8, 1, 0, 0),
),
param(
"Sa",
datetime(2025, 7, 26, 0, 0),
),
param(
"Su",
datetime(2025, 7, 27, 0, 0),
),
]
)
def test_short_weekday_names(self, date_string, expected):
if "Mo" in date_string:
pytest.xfail(
"Known bug: 'Mo' is being interpreted as a month instead of a weekday and needs to be fixed."
)

self.given_parser(["en"])
self.given_now(2025, 8, 1)
self.when_date_string_is_parsed(date_string)
self.then_parsed_datetime_is(expected)

def given_now(self, year, month, day, **time):
now = datetime(year, month, day, **time)
datetime_mock = Mock(wraps=datetime)
datetime_mock.utcnow = Mock(return_value=now)
datetime_mock.now = Mock(return_value=now)
datetime_mock.today = Mock(return_value=now)
self.add_patch(patch("dateparser.date.datetime", new=datetime_mock))
now = real_datetime.datetime(year, month, day, **time)

# Patch the datetime *class* in each target module
class DateParserDateTime(real_datetime.datetime):
@classmethod
def now(cls, tz=None):
return now.replace(tzinfo=tz) if tz else now

@classmethod
def utcnow(cls):
return now

@classmethod
def today(cls):
return now

self.add_patch(patch("dateparser.date.datetime", DateParserDateTime))
self.add_patch(patch("dateparser.parser.datetime", DateParserDateTime))

def given_parser(self, restrict_to_languages=None, **params):
self.parser = date.DateDataParser(languages=restrict_to_languages, **params)
Expand Down
Loading