Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/dateparser/parser.py: 91%

1import calendar

2from collections import OrderedDict

3from datetime import datetime, timedelta, timezone

4from io import StringIO

6import pytz

7import regex as re

9from dateparser.utils import (

10 _get_missing_parts,

11 get_last_day_of_month,

12 get_next_leap_year,

13 get_previous_leap_year,

14 get_timezone_from_tz_string,

15 set_correct_day_from_settings,

16 set_correct_month_from_settings,

17)

18from dateparser.utils.strptime import strptime

20NSP_COMPATIBLE = re.compile(r"\D+")

21MERIDIAN = re.compile(r"am|pm")

22MICROSECOND = re.compile(r"\d{1,6}")

23EIGHT_DIGIT = re.compile(r"^\d{8}$")

24HOUR_MINUTE_REGEX = re.compile(r"^([0-9]|0[0-9]|1[0-9]|2[0-3]):[0-5][0-9]$")

27def no_space_parser_eligibile(datestring):

28 src = NSP_COMPATIBLE.search(datestring)

29 if not src or ":" == src.group():

30 return True

31 return False

34def get_unresolved_attrs(parser_object):

35 attrs = ["year", "month", "day"]

36 seen = []

37 unseen = []

38 for attr in attrs:

39 if getattr(parser_object, attr, None) is not None:

40 seen.append(attr)

41 else:

42 unseen.append(attr)

43 return seen, unseen

46date_order_chart = {

47 "DMY": "%d%m%y",

48 "DYM": "%d%y%m",

49 "MDY": "%m%d%y",

50 "MYD": "%m%y%d",

51 "YDM": "%y%d%m",

52 "YMD": "%y%m%d",

53}

56def resolve_date_order(order, lst=None):

57 chart_list = {

58 "DMY": ["day", "month", "year"],

59 "DYM": ["day", "year", "month"],

60 "MDY": ["month", "day", "year"],

61 "MYD": ["month", "year", "day"],

62 "YDM": ["year", "day", "month"],

63 "YMD": ["year", "month", "day"],

64 }

66 return chart_list[order] if lst else date_order_chart[order]

69def _parse_absolute(datestring, settings, tz=None):

70 return _parser.parse(datestring, settings, tz)

73def _parse_nospaces(datestring, settings, tz=None):

74 return _no_spaces_parser.parse(datestring, settings)

77class _time_parser:

78 time_directives = [

79 "%H:%M:%S",

80 "%I:%M:%S %p",

81 "%H:%M",

82 "%I:%M %p",

83 "%I %p",

84 "%H:%M:%S.%f",

85 "%I:%M:%S.%f %p",

86 "%H:%M %p",

87 ]

89 def __call__(self, timestring):

90 _timestring = timestring

91 for directive in self.time_directives:

92 try:

93 return strptime(timestring.strip(), directive).time()

94 except ValueError:

95 pass

96 else:

97 raise ValueError("%s does not seem to be a valid time string" % _timestring)

100time_parser = _time_parser()

101

102

103class _no_spaces_parser:

104 _dateformats = [

105 "%Y%m%d",

106 "%Y%d%m",

107 "%m%Y%d",

108 "%m%d%Y",

109 "%d%Y%m",

110 "%d%m%Y",

111 "%y%m%d",

112 "%y%d%m",

113 "%m%y%d",

114 "%m%d%y",

115 "%d%y%m",

116 "%d%m%y",

117 ]

118

119 _preferred_formats = ["%Y%m%d%H%M", "%Y%m%d%H%M%S", "%Y%m%d%H%M%S.%f"]

120

121 _preferred_formats_ordered_8_digit = [

122 "%m%d%Y",

123 "%d%m%Y",

124 "%Y%m%d",

125 "%Y%d%m",

126 "%m%Y%d",

127 "%d%Y%m",

128 ]

129

130 _timeformats = ["%H%M%S.%f", "%H%M%S", "%H%M", "%H"]

131

132 period = {"day": ["%d", "%H", "%M", "%S"], "month": ["%m"]}

133

134 _default_order = resolve_date_order("MDY")

135

136 def __init__(self, *args, **kwargs):

137 self._all = (

138 self._dateformats

139 + [x + y for x in self._dateformats for y in self._timeformats]

140 + self._timeformats

141 )

142

143 self.date_formats = {

144 "%m%d%y": (

145 self._preferred_formats

146 + sorted(

147 self._all,

148 key=lambda x: x.lower().startswith("%m%d%y"),

149 reverse=True,

150 )

151 ),

152 "%m%y%d": sorted(

153 self._all, key=lambda x: x.lower().startswith("%m%y%d"), reverse=True

154 ),

155 "%y%m%d": sorted(

156 self._all, key=lambda x: x.lower().startswith("%y%m%d"), reverse=True

157 ),

158 "%y%d%m": sorted(

159 self._all, key=lambda x: x.lower().startswith("%y%d%m"), reverse=True

160 ),

161 "%d%m%y": sorted(

162 self._all, key=lambda x: x.lower().startswith("%d%m%y"), reverse=True

163 ),

164 "%d%y%m": sorted(

165 self._all, key=lambda x: x.lower().startswith("%d%y%m"), reverse=True

166 ),

167 }

168

169 @classmethod

170 def _get_period(cls, format_string):

171 for pname, pdrv in sorted(cls.period.items(), key=lambda x: x[0]):

172 for drv in pdrv:

173 if drv in format_string:

174 return pname

175 else:

176 return "year"

177

178 @classmethod

179 def _find_best_matching_date(cls, datestring):

180 for fmt in cls._preferred_formats_ordered_8_digit:

181 try:

182 dt = strptime(datestring, fmt), cls._get_period(fmt)

183 if len(str(dt[0].year)) == 4:

184 return dt

185 except:

186 pass

187 return None

188

189 @classmethod

190 def parse(cls, datestring, settings):

191 if not no_space_parser_eligibile(datestring):

192 raise ValueError("Unable to parse date from: %s" % datestring)

193

194 datestring = datestring.replace(":", "")

195 if not datestring:

196 raise ValueError("Empty string")

197 tokens = tokenizer(datestring)

198 if settings.DATE_ORDER:

199 order = resolve_date_order(settings.DATE_ORDER)

200 else:

201 order = cls._default_order

202 if EIGHT_DIGIT.match(datestring):

203 dt = cls._find_best_matching_date(datestring)

204 if dt is not None:

205 return dt

206 nsp = cls()

207 ambiguous_date = None

208 for token, _ in tokens.tokenize():

209 for fmt in nsp.date_formats[order]:

210 try:

211 dt = strptime(token, fmt), cls._get_period(fmt)

212 if len(str(dt[0].year)) < 4:

213 ambiguous_date = dt

214 continue

215

216 missing = _get_missing_parts(fmt)

217 _check_strict_parsing(missing, settings)

218 return dt

219 except:

220 pass

221 else:

222 if ambiguous_date:

223 return ambiguous_date

224 else:

225 raise ValueError("Unable to parse date from: %s" % datestring)

226

227

228def _get_missing_error(missing):

229 return "Fields missing from the date string: {}".format(", ".join(missing))

230

231

232def _check_strict_parsing(missing, settings):

233 if settings.STRICT_PARSING and missing:

234 raise ValueError(_get_missing_error(missing))

235 elif settings.REQUIRE_PARTS and missing:

236 errors = [part for part in settings.REQUIRE_PARTS if part in missing]

237 if errors:

238 raise ValueError(_get_missing_error(errors))

239

240

241class _parser:

242 alpha_directives = OrderedDict(

243 [

244 ("weekday", ["%A", "%a"]),

245 ("month", ["%B", "%b"]),

246 ]

247 )

248

249 num_directives = {

250 "month": ["%m"],

251 "day": ["%d"],

252 "year": ["%y", "%Y"],

253 }

254

255 def __init__(self, tokens, settings):

256 self.settings = settings

257 self.tokens = [(t[0].strip(), t[1]) for t in list(tokens)]

258 self.filtered_tokens = [

259 (t[0], t[1], i) for i, t in enumerate(self.tokens) if t[1] <= 1

260 ]

261

262 self.unset_tokens = []

263

264 self.day = None

265 self.month = None

266 self.year = None

267 self.time = None

268

269 self.auto_order = []

270

271 self._token_day = None

272 self._token_month = None

273 self._token_year = None

274 self._token_time = None

275

276 self.ordered_num_directives = OrderedDict(

277 (k, self.num_directives[k])

278 for k in (resolve_date_order(settings.DATE_ORDER, lst=True))

279 )

280

281 skip_index = []

282 skip_component = None

283 skip_tokens = ["t", "year", "hour", "minute"]

284

285 for index, token_type_original_index in enumerate(self.filtered_tokens):

286 if index in skip_index:

287 continue

288

289 token, type, original_index = token_type_original_index

290

291 if token in skip_tokens:

292 continue

293

294 if self.time is None:

295 meridian_index = index + 1

296

297 try:

298 # try case where hours and minutes are separated by a period. Example: 13.20.

299 _is_before_period = self.tokens[original_index + 1][0] == "."

300 _is_after_period = (

301 original_index != 0

302 and self.tokens[original_index - 1][0] == "."

303 )

304

305 if _is_before_period and not _is_after_period:

306 index_next_token = index + 1

307 next_token = self.filtered_tokens[index_next_token][0]

308 index_in_tokens_for_next_token = self.filtered_tokens[

309 index_next_token

310 ][2]

311

312 next_token_is_last = (

313 index_next_token == len(self.filtered_tokens) - 1

314 )

315 if (

316 next_token_is_last

317 or self.tokens[index_in_tokens_for_next_token + 1][0] != "."

318 ):

319 new_token = token + ":" + next_token

320 if re.match(HOUR_MINUTE_REGEX, new_token):

321 token = new_token

322 skip_index.append(index + 1)

323 meridian_index += 1

324 except Exception:

325 pass

326

327 try:

328 microsecond = MICROSECOND.search(

329 self.filtered_tokens[index + 1][0]

330 ).group()

331 # Is after time token? raise ValueError if ':' can't be found:

332 token.index(":")

333 # Is after period? raise ValueError if '.' can't be found:

334 self.tokens[self.tokens.index((token, 0)) + 1][0].index(".")

335 except:

336 microsecond = None

337

338 if microsecond:

339 meridian_index += 1

340

341 try:

342 meridian = MERIDIAN.search(

343 self.filtered_tokens[meridian_index][0]

344 ).group()

345 except:

346 meridian = None

347

348 if any([":" in token, meridian, microsecond]):

349 if meridian and not microsecond:

350 self._token_time = "%s %s" % (token, meridian)

351 skip_index.append(meridian_index)

352 elif microsecond and not meridian:

353 self._token_time = "%s.%s" % (token, microsecond)

354 skip_index.append(index + 1)

355 elif meridian and microsecond:

356 self._token_time = "%s.%s %s" % (token, microsecond, meridian)

357 skip_index.append(index + 1)

358 skip_index.append(meridian_index)

359 else:

360 self._token_time = token

361 self.time = lambda: time_parser(self._token_time)

362 continue

363

364 results = self._parse(type, token, skip_component=skip_component)

365 for res in results:

366 if len(token) == 4 and res[0] == "year":

367 skip_component = "year"

368 setattr(self, *res)

369

370 known, unknown = get_unresolved_attrs(self)

371 params = {}

372 for attr in known:

373 params.update({attr: getattr(self, attr)})

374 for attr in unknown:

375 for token, type, _ in self.unset_tokens:

376 if type == 0:

377 params.update({attr: int(token)})

378 setattr(self, "_token_%s" % attr, token)

379 setattr(self, attr, int(token))

380

381 def _get_period(self):

382 if self.settings.RETURN_TIME_AS_PERIOD:

383 if getattr(self, "time", None):

384 return "time"

385

386 for period in ["time", "day"]:

387 if getattr(self, period, None):

388 return "day"

389

390 for period in ["month", "year"]:

391 if getattr(self, period, None):

392 return period

393

394 if self._results():

395 return "day"

396

397 def _get_datetime_obj(self, **params):

398 try:

399 return datetime(**params)

400 except ValueError as e:

401 error_text = e.__str__()

402 error_msgs = ["day is out of range", "day must be in"]

403 if error_msgs[0] in error_text or error_msgs[1] in error_text:

404 if not (self._token_day or hasattr(self, "_token_weekday")):

405 # if day is not available put last day of the month

406 params["day"] = get_last_day_of_month(

407 params["year"], params["month"]

408 )

409 return datetime(**params)

410 elif (

411 not self._token_year

412 and params["day"] == 29

413 and params["month"] == 2

414 and not calendar.isleap(params["year"])

415 ):

416 # fix the year when year is not present and it is 29 of February

417 params["year"] = self._get_correct_leap_year(

418 self.settings.PREFER_DATES_FROM, params["year"]

419 )

420 return datetime(**params)

421 raise e

422

423 def _get_correct_leap_year(self, prefer_dates_from, current_year):

424 if prefer_dates_from == "future":

425 return get_next_leap_year(current_year)

426 if prefer_dates_from == "past":

427 return get_previous_leap_year(current_year)

428

429 # Default case ('current_period'): return closer leap year

430 next_leap_year = get_next_leap_year(current_year)

431 previous_leap_year = get_previous_leap_year(current_year)

432 next_leap_year_is_closer = (

433 next_leap_year - current_year < current_year - previous_leap_year

434 )

435 return next_leap_year if next_leap_year_is_closer else previous_leap_year

436

437 def _set_relative_base(self):

438 self.now = self.settings.RELATIVE_BASE

439 if not self.now:

440 self.now = datetime.now(tz=timezone.utc).replace(tzinfo=None)

441

442 def _get_datetime_obj_params(self):

443 if not self.now:

444 self._set_relative_base()

445

446 params = {

447 "day": self.day or self.now.day,

448 "month": self.month or self.now.month,

449 "year": self.year or self.now.year,

450 "hour": 0,

451 "minute": 0,

452 "second": 0,

453 "microsecond": 0,

454 }

455 return params

456

457 def _get_date_obj(self, token, directive):

458 return strptime(token, directive)

459

460 def _results(self):

461 missing = [

462 field for field in ("day", "month", "year") if not getattr(self, field)

463 ]

464 _check_strict_parsing(missing, self.settings)

465 self._set_relative_base()

466

467 time = self.time() if self.time is not None else None

468 params = self._get_datetime_obj_params()

469

470 if time:

471 params.update(

472 dict(

473 hour=time.hour,

474 minute=time.minute,

475 second=time.second,

476 microsecond=time.microsecond,

477 )

478 )

479

480 return self._get_datetime_obj(**params)

481

482 def _correct_for_time_frame(self, dateobj, tz):

483 days = ["mon", "tue", "wed", "thu", "fri", "sat", "sun"]

484

485 token_weekday, _ = getattr(self, "_token_weekday", (None, None))

486

487 if token_weekday and not (

488 self._token_year or self._token_month or self._token_day

489 ):

490 day_index = calendar.weekday(dateobj.year, dateobj.month, dateobj.day)

491 day = token_weekday[:3].lower()

492 steps = 0

493 if "future" in self.settings.PREFER_DATES_FROM:

494 if days[day_index] == day:

495 steps = 7

496 else:

497 while days[day_index] != day:

498 day_index = (day_index + 1) % 7

499 steps += 1

500 delta = timedelta(days=steps)

501 else:

502 if days[day_index] == day:

503 if self.settings.PREFER_DATES_FROM == "past":

504 steps = 7

505 else:

506 steps = 0

507 else:

508 while days[day_index] != day:

509 day_index -= 1

510 steps += 1

511 delta = timedelta(days=-steps)

512

513 dateobj = dateobj + delta

514

515 # NOTE: If this assert fires, self.now needs to be made offset-aware in a similar

516 # way that dateobj is temporarily made offset-aware.

517 assert not (

518 self.now.tzinfo is None and dateobj.tzinfo is not None

519 ), "`self.now` doesn't have `tzinfo`. Review comment in code for details."

520

521 # Store the original dateobj values so that upon subsequent parsing everything is not

522 # treated as offset-aware if offset awareness is changed.

523 original_dateobj = dateobj

524

525 # Since date comparisons must be either offset-naive or offset-aware, normalize dateobj

526 # to be offset-aware if one or the other is already offset-aware.

527 if self.now.tzinfo is not None and dateobj.tzinfo is None:

528 dateobj = pytz.utc.localize(dateobj)

529

530 if self.month and not self.year:

531 try:

532 if self.now < dateobj:

533 if self.settings.PREFER_DATES_FROM == "past":

534 dateobj = dateobj.replace(year=dateobj.year - 1)

535 else:

536 if self.settings.PREFER_DATES_FROM == "future":

537 dateobj = dateobj.replace(year=dateobj.year + 1)

538 except ValueError as e:

539 if dateobj.day == 29 and dateobj.month == 2:

540 valid_year = self._get_correct_leap_year(

541 self.settings.PREFER_DATES_FROM, dateobj.year

542 )

543 dateobj = dateobj.replace(year=valid_year)

544 else:

545 raise e

546

547 if self._token_year and len(self._token_year[0]) == 2:

548 if self.now < dateobj:

549 if "past" in self.settings.PREFER_DATES_FROM:

550 dateobj = dateobj.replace(year=dateobj.year - 100)

551 else:

552 if "future" in self.settings.PREFER_DATES_FROM:

553 dateobj = dateobj.replace(year=dateobj.year + 100)

554

555 if self._token_time and not any(

556 [

557 self._token_year,

558 self._token_month,

559 self._token_day,

560 hasattr(self, "_token_weekday"),

561 ]

562 ):

563 # Convert dateobj to utc time to compare with self.now

564 try:

565 tz = tz or get_timezone_from_tz_string(self.settings.TIMEZONE)

566 tz_offset = tz.utcoffset(dateobj)

567 except (pytz.UnknownTimeZoneError, pytz.NonExistentTimeError):

568 tz_offset = timedelta(hours=0)

569

570 if "past" in self.settings.PREFER_DATES_FROM:

571 if self.now < dateobj - tz_offset:

572 dateobj = dateobj + timedelta(days=-1)

573 if "future" in self.settings.PREFER_DATES_FROM:

574 if self.now > dateobj - tz_offset:

575 dateobj = dateobj + timedelta(days=1)

576

577 # Reset dateobj to the original value, thus removing any offset awareness that may

578 # have been set earlier.

579 dateobj = dateobj.replace(tzinfo=original_dateobj.tzinfo)

580

581 return dateobj

582

583 def _correct_for_day(self, dateobj):

584 if (

585 getattr(self, "_token_day", None)

586 or getattr(self, "_token_weekday", None)

587 or getattr(self, "_token_time", None)

588 ):

589 return dateobj

590

591 dateobj = set_correct_day_from_settings(

592 dateobj, self.settings, current_day=self.now.day

593 )

594 return dateobj

595

596 def _correct_for_month(self, dateobj):

597 relative_base = getattr(self.settings, "RELATIVE_BASE", None)

598 relative_base_month = (

599 relative_base.month if hasattr(relative_base, "month") else relative_base

600 )

601

602 if getattr(self, "_token_month", None):

603 return dateobj

604

605 dateobj = set_correct_month_from_settings(

606 dateobj, self.settings, relative_base_month

607 )

608 return dateobj

609

610 @classmethod

611 def parse(cls, datestring, settings, tz=None):

612 tokens = tokenizer(datestring)

613 po = cls(tokens.tokenize(), settings)

614 dateobj = po._results()

615

616 # correction for past, future if applicable

617 dateobj = po._correct_for_time_frame(dateobj, tz)

618

619 # correction for preference of month: beginning, current, end

620 # must happen before day so that day is derived from the correct month

621 dateobj = po._correct_for_month(dateobj)

622

623 # correction for preference of day: beginning, current, end

624 dateobj = po._correct_for_day(dateobj)

625

626 period = po._get_period()

627

628 return dateobj, period

629

630 def _parse(self, type, token, skip_component=None):

631 def set_and_return(token, type, component, dateobj, skip_date_order=False):

632 if not skip_date_order:

633 self.auto_order.append(component)

634 setattr(self, "_token_%s" % component, (token, type))

635 return [(component, getattr(dateobj, component))]

636

637 def parse_number(token, skip_component=None):

638 type = 0

639

640 for component, directives in self.ordered_num_directives.items():

641 if skip_component == component:

642 continue

643 for directive in directives:

644 try:

645 do = self._get_date_obj(token, directive)

646 prev_value = getattr(self, component, None)

647 if not prev_value:

648 return set_and_return(token, type, component, do)

649 else:

650 try:

651 prev_token, prev_type = getattr(

652 self, "_token_%s" % component

653 )

654 if prev_type == type:

655 do = self._get_date_obj(prev_token, directive)

656 except ValueError:

657 self.unset_tokens.append(

658 (prev_token, prev_type, component)

659 )

660 return set_and_return(token, type, component, do)

661 except ValueError:

662 pass

663 else:

664 raise ValueError("Unable to parse: %s" % token)

665

666 def parse_alpha(token, skip_component=None):

667 type = 1

668

669 for component, directives in self.alpha_directives.items():

670 if skip_component == component:

671 continue

672 for directive in directives:

673 try:

674 do = self._get_date_obj(token, directive)

675 prev_value = getattr(self, component, None)

676 if not prev_value:

677 return set_and_return(

678 token, type, component, do, skip_date_order=True

679 )

680 elif component == "month":

681 index = self.auto_order.index("month")

682 self.auto_order[index] = "day"

683 setattr(self, "_token_day", self._token_month)

684 setattr(self, "_token_month", (token, type))

685 return [

686 (component, getattr(do, component)),

687 ("day", prev_value),

688 ]

689 except:

690 pass

691 else:

692 raise ValueError("Unable to parse: %s" % token)

693

694 handlers = {0: parse_number, 1: parse_alpha}

695 return handlers[type](token, skip_component)

696

697

698class tokenizer:

699 digits = "0123456789:"

700 letters = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"

701

702 def _isletter(self, tkn):

703 return tkn in self.letters

704

705 def _isdigit(self, tkn):

706 return tkn in self.digits

707

708 def __init__(self, ds):

709 self.instream = StringIO(ds)

710

711 def _switch(self, chara, charb):

712 if self._isdigit(chara):

713 return 0, not self._isdigit(charb)

714

715 if self._isletter(chara):

716 return 1, not self._isletter(charb)

717

718 return 2, self._isdigit(charb) or self._isletter(charb)

719

720 def tokenize(self):

721 token = ""

722 EOF = False

723

724 while not EOF:

725 nextchar = self.instream.read(1)

726

727 if not nextchar:

728 EOF = True

729 type, _ = self._switch(token[-1], nextchar)

730 yield token, type

731 return

732

733 if token:

734 type, switch = self._switch(token[-1], nextchar)

735

736 if not switch:

737 token += nextchar

738 else:

739 yield token, type

740 token = nextchar

741 else:

742 token += nextchar