imk/node_modules/date-fns/parse.js

514 lines
28 KiB
JavaScript
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

import { defaultLocale } from "./_lib/defaultLocale.js";
import { longFormatters } from "./_lib/format/longFormatters.js";
import {
isProtectedDayOfYearToken,
isProtectedWeekYearToken,
warnOrThrowProtectedError,
} from "./_lib/protectedTokens.js";
import { constructFrom } from "./constructFrom.js";
import { getDefaultOptions } from "./getDefaultOptions.js";
import { toDate } from "./toDate.js";
import { DateTimezoneSetter } from "./parse/_lib/Setter.js";
import { parsers } from "./parse/_lib/parsers.js";
// Rexports of internal for libraries to use.
// See: https://github.com/date-fns/date-fns/issues/3638#issuecomment-1877082874
export { longFormatters, parsers };
/**
* The {@link parse} function options.
*/
// This RegExp consists of three parts separated by `|`:
// - [yYQqMLwIdDecihHKkms]o matches any available ordinal number token
// (one of the certain letters followed by `o`)
// - (\w)\1* matches any sequences of the same letter
// - '' matches two quote characters in a row
// - '(''|[^'])+('|$) matches anything surrounded by two quote characters ('),
// except a single quote symbol, which ends the sequence.
// Two quote characters do not end the sequence.
// If there is no matching single quote
// then the sequence will continue until the end of the string.
// - . matches any single character unmatched by previous parts of the RegExps
const formattingTokensRegExp =
/[yYQqMLwIdDecihHKkms]o|(\w)\1*|''|'(''|[^'])+('|$)|./g;
// This RegExp catches symbols escaped by quotes, and also
// sequences of symbols P, p, and the combinations like `PPPPPPPppppp`
const longFormattingTokensRegExp = /P+p+|P+|p+|''|'(''|[^'])+('|$)|./g;
const escapedStringRegExp = /^'([^]*?)'?$/;
const doubleQuoteRegExp = /''/g;
const notWhitespaceRegExp = /\S/;
const unescapedLatinCharacterRegExp = /[a-zA-Z]/;
/**
* @name parse
* @category Common Helpers
* @summary Parse the date.
*
* @description
* Return the date parsed from string using the given format string.
*
* > Please note that the `format` tokens differ from Moment.js and other libraries.
* > See: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
*
* The characters in the format string wrapped between two single quotes characters (') are escaped.
* Two single quotes in a row, whether inside or outside a quoted sequence, represent a 'real' single quote.
*
* Format of the format string is based on Unicode Technical Standard #35:
* https://www.unicode.org/reports/tr35/tr35-dates.html#Date_Field_Symbol_Table
* with a few additions (see note 5 below the table).
*
* Not all tokens are compatible. Combinations that don't make sense or could lead to bugs are prohibited
* and will throw `RangeError`. For example usage of 24-hour format token with AM/PM token will throw an exception:
*
* ```javascript
* parse('23 AM', 'HH a', new Date())
* //=> RangeError: The format string mustn't contain `HH` and `a` at the same time
* ```
*
* See the compatibility table: https://docs.google.com/spreadsheets/d/e/2PACX-1vQOPU3xUhplll6dyoMmVUXHKl_8CRDs6_ueLmex3SoqwhuolkuN3O05l4rqx5h1dKX8eb46Ul-CCSrq/pubhtml?gid=0&single=true
*
* Accepted format string patterns:
* | Unit |Prior| Pattern | Result examples | Notes |
* |---------------------------------|-----|---------|-----------------------------------|-------|
* | Era | 140 | G..GGG | AD, BC | |
* | | | GGGG | Anno Domini, Before Christ | 2 |
* | | | GGGGG | A, B | |
* | Calendar year | 130 | y | 44, 1, 1900, 2017, 9999 | 4 |
* | | | yo | 44th, 1st, 1900th, 9999999th | 4,5 |
* | | | yy | 44, 01, 00, 17 | 4 |
* | | | yyy | 044, 001, 123, 999 | 4 |
* | | | yyyy | 0044, 0001, 1900, 2017 | 4 |
* | | | yyyyy | ... | 2,4 |
* | Local week-numbering year | 130 | Y | 44, 1, 1900, 2017, 9000 | 4 |
* | | | Yo | 44th, 1st, 1900th, 9999999th | 4,5 |
* | | | YY | 44, 01, 00, 17 | 4,6 |
* | | | YYY | 044, 001, 123, 999 | 4 |
* | | | YYYY | 0044, 0001, 1900, 2017 | 4,6 |
* | | | YYYYY | ... | 2,4 |
* | ISO week-numbering year | 130 | R | -43, 1, 1900, 2017, 9999, -9999 | 4,5 |
* | | | RR | -43, 01, 00, 17 | 4,5 |
* | | | RRR | -043, 001, 123, 999, -999 | 4,5 |
* | | | RRRR | -0043, 0001, 2017, 9999, -9999 | 4,5 |
* | | | RRRRR | ... | 2,4,5 |
* | Extended year | 130 | u | -43, 1, 1900, 2017, 9999, -999 | 4 |
* | | | uu | -43, 01, 99, -99 | 4 |
* | | | uuu | -043, 001, 123, 999, -999 | 4 |
* | | | uuuu | -0043, 0001, 2017, 9999, -9999 | 4 |
* | | | uuuuu | ... | 2,4 |
* | Quarter (formatting) | 120 | Q | 1, 2, 3, 4 | |
* | | | Qo | 1st, 2nd, 3rd, 4th | 5 |
* | | | QQ | 01, 02, 03, 04 | |
* | | | QQQ | Q1, Q2, Q3, Q4 | |
* | | | QQQQ | 1st quarter, 2nd quarter, ... | 2 |
* | | | QQQQQ | 1, 2, 3, 4 | 4 |
* | Quarter (stand-alone) | 120 | q | 1, 2, 3, 4 | |
* | | | qo | 1st, 2nd, 3rd, 4th | 5 |
* | | | qq | 01, 02, 03, 04 | |
* | | | qqq | Q1, Q2, Q3, Q4 | |
* | | | qqqq | 1st quarter, 2nd quarter, ... | 2 |
* | | | qqqqq | 1, 2, 3, 4 | 3 |
* | Month (formatting) | 110 | M | 1, 2, ..., 12 | |
* | | | Mo | 1st, 2nd, ..., 12th | 5 |
* | | | MM | 01, 02, ..., 12 | |
* | | | MMM | Jan, Feb, ..., Dec | |
* | | | MMMM | January, February, ..., December | 2 |
* | | | MMMMM | J, F, ..., D | |
* | Month (stand-alone) | 110 | L | 1, 2, ..., 12 | |
* | | | Lo | 1st, 2nd, ..., 12th | 5 |
* | | | LL | 01, 02, ..., 12 | |
* | | | LLL | Jan, Feb, ..., Dec | |
* | | | LLLL | January, February, ..., December | 2 |
* | | | LLLLL | J, F, ..., D | |
* | Local week of year | 100 | w | 1, 2, ..., 53 | |
* | | | wo | 1st, 2nd, ..., 53th | 5 |
* | | | ww | 01, 02, ..., 53 | |
* | ISO week of year | 100 | I | 1, 2, ..., 53 | 5 |
* | | | Io | 1st, 2nd, ..., 53th | 5 |
* | | | II | 01, 02, ..., 53 | 5 |
* | Day of month | 90 | d | 1, 2, ..., 31 | |
* | | | do | 1st, 2nd, ..., 31st | 5 |
* | | | dd | 01, 02, ..., 31 | |
* | Day of year | 90 | D | 1, 2, ..., 365, 366 | 7 |
* | | | Do | 1st, 2nd, ..., 365th, 366th | 5 |
* | | | DD | 01, 02, ..., 365, 366 | 7 |
* | | | DDD | 001, 002, ..., 365, 366 | |
* | | | DDDD | ... | 2 |
* | Day of week (formatting) | 90 | E..EEE | Mon, Tue, Wed, ..., Sun | |
* | | | EEEE | Monday, Tuesday, ..., Sunday | 2 |
* | | | EEEEE | M, T, W, T, F, S, S | |
* | | | EEEEEE | Mo, Tu, We, Th, Fr, Sa, Su | |
* | ISO day of week (formatting) | 90 | i | 1, 2, 3, ..., 7 | 5 |
* | | | io | 1st, 2nd, ..., 7th | 5 |
* | | | ii | 01, 02, ..., 07 | 5 |
* | | | iii | Mon, Tue, Wed, ..., Sun | 5 |
* | | | iiii | Monday, Tuesday, ..., Sunday | 2,5 |
* | | | iiiii | M, T, W, T, F, S, S | 5 |
* | | | iiiiii | Mo, Tu, We, Th, Fr, Sa, Su | 5 |
* | Local day of week (formatting) | 90 | e | 2, 3, 4, ..., 1 | |
* | | | eo | 2nd, 3rd, ..., 1st | 5 |
* | | | ee | 02, 03, ..., 01 | |
* | | | eee | Mon, Tue, Wed, ..., Sun | |
* | | | eeee | Monday, Tuesday, ..., Sunday | 2 |
* | | | eeeee | M, T, W, T, F, S, S | |
* | | | eeeeee | Mo, Tu, We, Th, Fr, Sa, Su | |
* | Local day of week (stand-alone) | 90 | c | 2, 3, 4, ..., 1 | |
* | | | co | 2nd, 3rd, ..., 1st | 5 |
* | | | cc | 02, 03, ..., 01 | |
* | | | ccc | Mon, Tue, Wed, ..., Sun | |
* | | | cccc | Monday, Tuesday, ..., Sunday | 2 |
* | | | ccccc | M, T, W, T, F, S, S | |
* | | | cccccc | Mo, Tu, We, Th, Fr, Sa, Su | |
* | AM, PM | 80 | a..aaa | AM, PM | |
* | | | aaaa | a.m., p.m. | 2 |
* | | | aaaaa | a, p | |
* | AM, PM, noon, midnight | 80 | b..bbb | AM, PM, noon, midnight | |
* | | | bbbb | a.m., p.m., noon, midnight | 2 |
* | | | bbbbb | a, p, n, mi | |
* | Flexible day period | 80 | B..BBB | at night, in the morning, ... | |
* | | | BBBB | at night, in the morning, ... | 2 |
* | | | BBBBB | at night, in the morning, ... | |
* | Hour [1-12] | 70 | h | 1, 2, ..., 11, 12 | |
* | | | ho | 1st, 2nd, ..., 11th, 12th | 5 |
* | | | hh | 01, 02, ..., 11, 12 | |
* | Hour [0-23] | 70 | H | 0, 1, 2, ..., 23 | |
* | | | Ho | 0th, 1st, 2nd, ..., 23rd | 5 |
* | | | HH | 00, 01, 02, ..., 23 | |
* | Hour [0-11] | 70 | K | 1, 2, ..., 11, 0 | |
* | | | Ko | 1st, 2nd, ..., 11th, 0th | 5 |
* | | | KK | 01, 02, ..., 11, 00 | |
* | Hour [1-24] | 70 | k | 24, 1, 2, ..., 23 | |
* | | | ko | 24th, 1st, 2nd, ..., 23rd | 5 |
* | | | kk | 24, 01, 02, ..., 23 | |
* | Minute | 60 | m | 0, 1, ..., 59 | |
* | | | mo | 0th, 1st, ..., 59th | 5 |
* | | | mm | 00, 01, ..., 59 | |
* | Second | 50 | s | 0, 1, ..., 59 | |
* | | | so | 0th, 1st, ..., 59th | 5 |
* | | | ss | 00, 01, ..., 59 | |
* | Seconds timestamp | 40 | t | 512969520 | |
* | | | tt | ... | 2 |
* | Fraction of second | 30 | S | 0, 1, ..., 9 | |
* | | | SS | 00, 01, ..., 99 | |
* | | | SSS | 000, 001, ..., 999 | |
* | | | SSSS | ... | 2 |
* | Milliseconds timestamp | 20 | T | 512969520900 | |
* | | | TT | ... | 2 |
* | Timezone (ISO-8601 w/ Z) | 10 | X | -08, +0530, Z | |
* | | | XX | -0800, +0530, Z | |
* | | | XXX | -08:00, +05:30, Z | |
* | | | XXXX | -0800, +0530, Z, +123456 | 2 |
* | | | XXXXX | -08:00, +05:30, Z, +12:34:56 | |
* | Timezone (ISO-8601 w/o Z) | 10 | x | -08, +0530, +00 | |
* | | | xx | -0800, +0530, +0000 | |
* | | | xxx | -08:00, +05:30, +00:00 | 2 |
* | | | xxxx | -0800, +0530, +0000, +123456 | |
* | | | xxxxx | -08:00, +05:30, +00:00, +12:34:56 | |
* | Long localized date | NA | P | 05/29/1453 | 5,8 |
* | | | PP | May 29, 1453 | |
* | | | PPP | May 29th, 1453 | |
* | | | PPPP | Sunday, May 29th, 1453 | 2,5,8 |
* | Long localized time | NA | p | 12:00 AM | 5,8 |
* | | | pp | 12:00:00 AM | |
* | Combination of date and time | NA | Pp | 05/29/1453, 12:00 AM | |
* | | | PPpp | May 29, 1453, 12:00:00 AM | |
* | | | PPPpp | May 29th, 1453 at ... | |
* | | | PPPPpp | Sunday, May 29th, 1453 at ... | 2,5,8 |
* Notes:
* 1. "Formatting" units (e.g. formatting quarter) in the default en-US locale
* are the same as "stand-alone" units, but are different in some languages.
* "Formatting" units are declined according to the rules of the language
* in the context of a date. "Stand-alone" units are always nominative singular.
* In `format` function, they will produce different result:
*
* `format(new Date(2017, 10, 6), 'do LLLL', {locale: cs}) //=> '6. listopad'`
*
* `format(new Date(2017, 10, 6), 'do MMMM', {locale: cs}) //=> '6. listopadu'`
*
* `parse` will try to match both formatting and stand-alone units interchangeably.
*
* 2. Any sequence of the identical letters is a pattern, unless it is escaped by
* the single quote characters (see below).
* If the sequence is longer than listed in table:
* - for numerical units (`yyyyyyyy`) `parse` will try to match a number
* as wide as the sequence
* - for text units (`MMMMMMMM`) `parse` will try to match the widest variation of the unit.
* These variations are marked with "2" in the last column of the table.
*
* 3. `QQQQQ` and `qqqqq` could be not strictly numerical in some locales.
* These tokens represent the shortest form of the quarter.
*
* 4. The main difference between `y` and `u` patterns are B.C. years:
*
* | Year | `y` | `u` |
* |------|-----|-----|
* | AC 1 | 1 | 1 |
* | BC 1 | 1 | 0 |
* | BC 2 | 2 | -1 |
*
* Also `yy` will try to guess the century of two digit year by proximity with `referenceDate`:
*
* `parse('50', 'yy', new Date(2018, 0, 1)) //=> Sat Jan 01 2050 00:00:00`
*
* `parse('75', 'yy', new Date(2018, 0, 1)) //=> Wed Jan 01 1975 00:00:00`
*
* while `uu` will just assign the year as is:
*
* `parse('50', 'uu', new Date(2018, 0, 1)) //=> Sat Jan 01 0050 00:00:00`
*
* `parse('75', 'uu', new Date(2018, 0, 1)) //=> Tue Jan 01 0075 00:00:00`
*
* The same difference is true for local and ISO week-numbering years (`Y` and `R`),
* except local week-numbering years are dependent on `options.weekStartsOn`
* and `options.firstWeekContainsDate` (compare [setISOWeekYear](https://date-fns.org/docs/setISOWeekYear)
* and [setWeekYear](https://date-fns.org/docs/setWeekYear)).
*
* 5. These patterns are not in the Unicode Technical Standard #35:
* - `i`: ISO day of week
* - `I`: ISO week of year
* - `R`: ISO week-numbering year
* - `o`: ordinal number modifier
* - `P`: long localized date
* - `p`: long localized time
*
* 6. `YY` and `YYYY` tokens represent week-numbering years but they are often confused with years.
* You should enable `options.useAdditionalWeekYearTokens` to use them. See: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
*
* 7. `D` and `DD` tokens represent days of the year but they are often confused with days of the month.
* You should enable `options.useAdditionalDayOfYearTokens` to use them. See: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
*
* 8. `P+` tokens do not have a defined priority since they are merely aliases to other tokens based
* on the given locale.
*
* using `en-US` locale: `P` => `MM/dd/yyyy`
* using `en-US` locale: `p` => `hh:mm a`
* using `pt-BR` locale: `P` => `dd/MM/yyyy`
* using `pt-BR` locale: `p` => `HH:mm`
*
* Values will be assigned to the date in the descending order of its unit's priority.
* Units of an equal priority overwrite each other in the order of appearance.
*
* If no values of higher priority are parsed (e.g. when parsing string 'January 1st' without a year),
* the values will be taken from 3rd argument `referenceDate` which works as a context of parsing.
*
* `referenceDate` must be passed for correct work of the function.
* If you're not sure which `referenceDate` to supply, create a new instance of Date:
* `parse('02/11/2014', 'MM/dd/yyyy', new Date())`
* In this case parsing will be done in the context of the current date.
* If `referenceDate` is `Invalid Date` or a value not convertible to valid `Date`,
* then `Invalid Date` will be returned.
*
* The result may vary by locale.
*
* If `formatString` matches with `dateString` but does not provides tokens, `referenceDate` will be returned.
*
* If parsing failed, `Invalid Date` will be returned.
* Invalid Date is a Date, whose time value is NaN.
* Time value of Date: http://es5.github.io/#x15.9.1.1
*
* @typeParam DateType - The `Date` type, the function operates on. Gets inferred from passed arguments. Allows to use extensions like [`UTCDate`](https://github.com/date-fns/utc).
* @typeParam ResultDate - The result `Date` type, it is the type returned from the context function if it is passed, or inferred from the arguments.
*
* @param dateStr - The string to parse
* @param formatStr - The string of tokens
* @param referenceDate - defines values missing from the parsed dateString
* @param options - An object with options.
* see: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
* see: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
*
* @returns The parsed date
*
* @throws `options.locale` must contain `match` property
* @throws use `yyyy` instead of `YYYY` for formatting years using [format provided] to the input [input provided]; see: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
* @throws use `yy` instead of `YY` for formatting years using [format provided] to the input [input provided]; see: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
* @throws use `d` instead of `D` for formatting days of the month using [format provided] to the input [input provided]; see: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
* @throws use `dd` instead of `DD` for formatting days of the month using [format provided] to the input [input provided]; see: https://github.com/date-fns/date-fns/blob/master/docs/unicodeTokens.md
* @throws format string contains an unescaped latin alphabet character
*
* @example
* // Parse 11 February 2014 from middle-endian format:
* var result = parse('02/11/2014', 'MM/dd/yyyy', new Date())
* //=> Tue Feb 11 2014 00:00:00
*
* @example
* // Parse 28th of February in Esperanto locale in the context of 2010 year:
* import eo from 'date-fns/locale/eo'
* var result = parse('28-a de februaro', "do 'de' MMMM", new Date(2010, 0, 1), {
* locale: eo
* })
* //=> Sun Feb 28 2010 00:00:00
*/
export function parse(dateStr, formatStr, referenceDate, options) {
const invalidDate = () => constructFrom(options?.in || referenceDate, NaN);
const defaultOptions = getDefaultOptions();
const locale = options?.locale ?? defaultOptions.locale ?? defaultLocale;
const firstWeekContainsDate =
options?.firstWeekContainsDate ??
options?.locale?.options?.firstWeekContainsDate ??
defaultOptions.firstWeekContainsDate ??
defaultOptions.locale?.options?.firstWeekContainsDate ??
1;
const weekStartsOn =
options?.weekStartsOn ??
options?.locale?.options?.weekStartsOn ??
defaultOptions.weekStartsOn ??
defaultOptions.locale?.options?.weekStartsOn ??
0;
if (!formatStr)
return dateStr ? invalidDate() : toDate(referenceDate, options?.in);
const subFnOptions = {
firstWeekContainsDate,
weekStartsOn,
locale,
};
// If timezone isn't specified, it will try to use the context or
// the reference date and fallback to the system time zone.
const setters = [new DateTimezoneSetter(options?.in, referenceDate)];
const tokens = formatStr
.match(longFormattingTokensRegExp)
.map((substring) => {
const firstCharacter = substring[0];
if (firstCharacter in longFormatters) {
const longFormatter = longFormatters[firstCharacter];
return longFormatter(substring, locale.formatLong);
}
return substring;
})
.join("")
.match(formattingTokensRegExp);
const usedTokens = [];
for (let token of tokens) {
if (
!options?.useAdditionalWeekYearTokens &&
isProtectedWeekYearToken(token)
) {
warnOrThrowProtectedError(token, formatStr, dateStr);
}
if (
!options?.useAdditionalDayOfYearTokens &&
isProtectedDayOfYearToken(token)
) {
warnOrThrowProtectedError(token, formatStr, dateStr);
}
const firstCharacter = token[0];
const parser = parsers[firstCharacter];
if (parser) {
const { incompatibleTokens } = parser;
if (Array.isArray(incompatibleTokens)) {
const incompatibleToken = usedTokens.find(
(usedToken) =>
incompatibleTokens.includes(usedToken.token) ||
usedToken.token === firstCharacter,
);
if (incompatibleToken) {
throw new RangeError(
`The format string mustn't contain \`${incompatibleToken.fullToken}\` and \`${token}\` at the same time`,
);
}
} else if (parser.incompatibleTokens === "*" && usedTokens.length > 0) {
throw new RangeError(
`The format string mustn't contain \`${token}\` and any other token at the same time`,
);
}
usedTokens.push({ token: firstCharacter, fullToken: token });
const parseResult = parser.run(
dateStr,
token,
locale.match,
subFnOptions,
);
if (!parseResult) {
return invalidDate();
}
setters.push(parseResult.setter);
dateStr = parseResult.rest;
} else {
if (firstCharacter.match(unescapedLatinCharacterRegExp)) {
throw new RangeError(
"Format string contains an unescaped latin alphabet character `" +
firstCharacter +
"`",
);
}
// Replace two single quote characters with one single quote character
if (token === "''") {
token = "'";
} else if (firstCharacter === "'") {
token = cleanEscapedString(token);
}
// Cut token from string, or, if string doesn't match the token, return Invalid Date
if (dateStr.indexOf(token) === 0) {
dateStr = dateStr.slice(token.length);
} else {
return invalidDate();
}
}
}
// Check if the remaining input contains something other than whitespace
if (dateStr.length > 0 && notWhitespaceRegExp.test(dateStr)) {
return invalidDate();
}
const uniquePrioritySetters = setters
.map((setter) => setter.priority)
.sort((a, b) => b - a)
.filter((priority, index, array) => array.indexOf(priority) === index)
.map((priority) =>
setters
.filter((setter) => setter.priority === priority)
.sort((a, b) => b.subPriority - a.subPriority),
)
.map((setterArray) => setterArray[0]);
let date = toDate(referenceDate, options?.in);
if (isNaN(+date)) return invalidDate();
const flags = {};
for (const setter of uniquePrioritySetters) {
if (!setter.validate(date, subFnOptions)) {
return invalidDate();
}
const result = setter.set(date, flags, subFnOptions);
// Result is tuple (date, flags)
if (Array.isArray(result)) {
date = result[0];
Object.assign(flags, result[1]);
// Result is date
} else {
date = result;
}
}
return date;
}
function cleanEscapedString(input) {
return input.match(escapedStringRegExp)[1].replace(doubleQuoteRegExp, "'");
}
// Fallback for modularized imports:
export default parse;