parseutils

This module contains helpers for parsing tokens, numbers, integers, floats, identifiers, etc.

To unpack raw bytes look at the streams module.

import parseutils

let logs = @["2019-01-10: OK_", "2019-01-11: FAIL_", "2019-01: aaaa"]

for log in logs:
  var res: string
  if parseUntil(log, res, ':') == 10: # YYYY-MM-DD == 10
    echo res & " - " & captureBetween(log, ' ', '_')
    # => 2019-01-10 - OK

import parseutils
from strutils import Digits, parseInt

let input1 = "2019 school start"
let input2 = "3 years back"

let startYear = input1[0..skipWhile(input1, Digits)-1] # 2019
let yearsBack = input2[0..skipWhile(input2, Digits)-1] # 3

echo "Examination is in " & $(parseInt(startYear) + parseInt(yearsBack))
# => Examination is in 2022

See also:

strutils module for combined and identical parsing proc's
json module for a JSON parser
parsecfg module for a configuration file parser
parsecsv module for a simple CSV (comma separated value) parser
parseopt module for a command line parser
parsexml module for a XML / HTML parser
other parsers for other parsers

Imports

Types

InterpolatedKind = enum ikStr, ## ``str`` part of the interpolated string ikDollar, ## escaped ``$`` part of the interpolated string ikVar, ## ``var`` part of the interpolated string ikExpr ## ``expr`` part of the interpolated string: Describes for interpolatedFragments which part of the interpolated string is yielded; for example in "str$$$var${expr}" Source Edit

Procs

proc parseBin[T: SomeInteger](s: string; number: var T; start = 0; maxLen = 0): int {...}{. noSideEffect.}

Parses a binary number and stores its value in number.

Returns the number of the parsed characters or 0 in case of an error. If error, the value of number is not changed.

If maxLen == 0, the parsing continues until the first non-bin character or to the end of the string. Otherwise, no more than maxLen characters are parsed starting from the start position.

It does not check for overflow. If the value represented by the string is too big to fit into number, only the value of last fitting characters will be stored in number without producing an error.

Examples:

var num: int
doAssert parseBin("0100_1110_0110_1001_1110_1101", num) == 29
doAssert num == 5138925
doAssert parseBin("3", num) == 0
var num8: int8
doAssert parseBin("0b_0100_1110_0110_1001_1110_1101", num8) == 32
doAssert num8 == 0b11101101'i8
doAssert parseBin("0b_0100_1110_0110_1001_1110_1101", num8, 3, 9) == 9
doAssert num8 == 0b01001110'i8
var num8u: uint8
doAssert parseBin("0b_0100_1110_0110_1001_1110_1101", num8u) == 32
doAssert num8u == 237
var num64: int64
doAssert parseBin("0100111001101001111011010100111001101001", num64) == 40
doAssert num64 == 336784608873'i64

proc parseOct[T: SomeInteger](s: string; number: var T; start = 0; maxLen = 0): int {...}{. noSideEffect.}

Parses an octal number and stores its value in number.

Returns the number of the parsed characters or 0 in case of an error. If error, the value of number is not changed.

If maxLen == 0, the parsing continues until the first non-oct character or to the end of the string. Otherwise, no more than maxLen characters are parsed starting from the start position.

It does not check for overflow. If the value represented by the string is too big to fit into number, only the value of last fitting characters will be stored in number without producing an error.

Examples:

var num: int
doAssert parseOct("0o23464755", num) == 10
doAssert num == 5138925
doAssert parseOct("8", num) == 0
var num8: int8
doAssert parseOct("0o_1464_755", num8) == 11
doAssert num8 == -19
doAssert parseOct("0o_1464_755", num8, 3, 3) == 3
doAssert num8 == 102
var num8u: uint8
doAssert parseOct("1464755", num8u) == 7
doAssert num8u == 237
var num64: int64
doAssert parseOct("2346475523464755", num64) == 16
doAssert num64 == 86216859871725'i64

proc parseHex[T: SomeInteger](s: string; number: var T; start = 0; maxLen = 0): int {...}{. noSideEffect.}

Parses a hexadecimal number and stores its value in number.

Returns the number of the parsed characters or 0 in case of an error. If error, the value of number is not changed.

If maxLen == 0, the parsing continues until the first non-hex character or to the end of the string. Otherwise, no more than maxLen characters are parsed starting from the start position.

It does not check for overflow. If the value represented by the string is too big to fit into number, only the value of last fitting characters will be stored in number without producing an error.

Examples:

var num: int
doAssert parseHex("4E_69_ED", num) == 8
doAssert num == 5138925
doAssert parseHex("X", num) == 0
doAssert parseHex("#ABC", num) == 4
var num8: int8
doAssert parseHex("0x_4E_69_ED", num8) == 11
doAssert num8 == 0xED'i8
doAssert parseHex("0x_4E_69_ED", num8, 3, 2) == 2
doAssert num8 == 0x4E'i8
var num8u: uint8
doAssert parseHex("0x_4E_69_ED", num8u) == 11
doAssert num8u == 237
var num64: int64
doAssert parseHex("4E69ED4E69ED", num64) == 12
doAssert num64 == 86216859871725'i64

proc parseIdent(s: string; ident: var string; start = 0): int {...}{.raises: [], tags: [].}

Parses an identifier and stores it in ident. Returns the number of the parsed characters or 0 in case of an error.

Examples:

var res: string
doAssert parseIdent("Hello World", res, 0) == 5
doAssert res == "Hello"
doAssert parseIdent("Hello World", res, 1) == 4
doAssert res == "ello"
doAssert parseIdent("Hello World", res, 6) == 5
doAssert res == "World"

proc parseIdent(s: string; start = 0): string {...}{.raises: [], tags: [].}

Parses an identifier and returns it or an empty string in case of an error.

Examples:

doAssert parseIdent("Hello World", 0) == "Hello"
doAssert parseIdent("Hello World", 1) == "ello"
doAssert parseIdent("Hello World", 5) == ""
doAssert parseIdent("Hello World", 6) == "World"

proc skipWhitespace(s: string; start = 0): int {...}{.inline, raises: [], tags: [].}

Skips the whitespace starting at s[start]. Returns the number of skipped characters.

Examples:

doAssert skipWhitespace("Hello World", 0) == 0
doAssert skipWhitespace(" Hello World", 0) == 1
doAssert skipWhitespace("Hello World", 5) == 1
doAssert skipWhitespace("Hello  World", 5) == 2

proc skip(s, token: string; start = 0): int {...}{.inline, raises: [], tags: [].}

Skips the token starting at s[start]. Returns the length of token or 0 if there was no token at s[start].

Examples:

doAssert skip("2019-01-22", "2019", 0) == 4
doAssert skip("2019-01-22", "19", 0) == 0
doAssert skip("2019-01-22", "19", 2) == 2
doAssert skip("CAPlow", "CAP", 0) == 3
doAssert skip("CAPlow", "cap", 0) == 0

proc skipIgnoreCase(s, token: string; start = 0): int {...}{.raises: [], tags: [].}

Same as skip but case is ignored for token matching.

Examples:

doAssert skipIgnoreCase("CAPlow", "CAP", 0) == 3
doAssert skipIgnoreCase("CAPlow", "cap", 0) == 3

proc skipUntil(s: string; until: set[char]; start = 0): int {...}{.inline, raises: [], tags: [].}

Skips all characters until one char from the set until is found or the end is reached. Returns number of characters skipped.

Examples:

doAssert skipUntil("Hello World", {'W', 'e'}, 0) == 1
doAssert skipUntil("Hello World", {'W'}, 0) == 6
doAssert skipUntil("Hello World", {'W', 'd'}, 0) == 6

proc skipUntil(s: string; until: char; start = 0): int {...}{.inline, raises: [], tags: [].}

Skips all characters until the char until is found or the end is reached. Returns number of characters skipped.

Examples:

doAssert skipUntil("Hello World", 'o', 0) == 4
doAssert skipUntil("Hello World", 'o', 4) == 0
doAssert skipUntil("Hello World", 'W', 0) == 6
doAssert skipUntil("Hello World", 'w', 0) == 11

proc skipWhile(s: string; toSkip: set[char]; start = 0): int {...}{.inline, raises: [], tags: [].}

Skips all characters while one char from the set token is found. Returns number of characters skipped.

Examples:

doAssert skipWhile("Hello World", {'H', 'e'}) == 2
doAssert skipWhile("Hello World", {'e'}) == 0
doAssert skipWhile("Hello World", {'W', 'o', 'r'}, 6) == 3

proc parseUntil(s: string; token: var string; until: set[char]; start = 0): int {...}{.inline, raises: [], tags: [].}

Parses a token and stores it in token. Returns the number of the parsed characters or 0 in case of an error. A token consists of the characters notin until.

Examples:

var myToken: string
doAssert parseUntil("Hello World", myToken, {'W', 'o', 'r'}) == 4
doAssert myToken == "Hell"
doAssert parseUntil("Hello World", myToken, {'W', 'r'}) == 6
doAssert myToken == "Hello "
doAssert parseUntil("Hello World", myToken, {'W', 'r'}, 3) == 3
doAssert myToken == "lo "

proc parseUntil(s: string; token: var string; until: char; start = 0): int {...}{.inline, raises: [], tags: [].}

Parses a token and stores it in token. Returns the number of the parsed characters or 0 in case of an error. A token consists of any character that is not the until character.

Examples:

var myToken: string
doAssert parseUntil("Hello World", myToken, 'W') == 6
doAssert myToken == "Hello "
doAssert parseUntil("Hello World", myToken, 'o') == 4
doAssert myToken == "Hell"
doAssert parseUntil("Hello World", myToken, 'o', 2) == 2
doAssert myToken == "ll"

proc parseUntil(s: string; token: var string; until: string; start = 0): int {...}{.inline, raises: [], tags: [].}

Parses a token and stores it in token. Returns the number of the parsed characters or 0 in case of an error. A token consists of any character that comes before the until token.

Examples:

var myToken: string
doAssert parseUntil("Hello World", myToken, "Wor") == 6
doAssert myToken == "Hello "
doAssert parseUntil("Hello World", myToken, "Wor", 2) == 4
doAssert myToken == "llo "

proc parseWhile(s: string; token: var string; validChars: set[char]; start = 0): int {...}{. inline, raises: [], tags: [].}

Parses a token and stores it in token. Returns the number of the parsed characters or 0 in case of an error. A token consists of the characters in validChars.

Examples:

var myToken: string
doAssert parseWhile("Hello World", myToken, {'W', 'o', 'r'}, 0) == 0
doAssert myToken.len() == 0
doAssert parseWhile("Hello World", myToken, {'W', 'o', 'r'}, 6) == 3
doAssert myToken == "Wor"

proc captureBetween(s: string; first: char; second = '\x00'; start = 0): string {...}{. raises: [], tags: [].}

Finds the first occurrence of first, then returns everything from there up to second (if second is '0', then first is used).

Examples:

doAssert captureBetween("Hello World", 'e') == "llo World"
doAssert captureBetween("Hello World", 'e', 'r') == "llo Wo"
doAssert captureBetween("Hello World", 'l', start = 6) == "d"

proc parseBiggestInt(s: string; number: var BiggestInt; start = 0): int {...}{.gcsafe, extern: "npuParseBiggestInt", noSideEffect, raises: [ValueError], tags: [].}

Parses an integer starting at start and stores the value into number. Result is the number of processed chars or 0 if there is no integer. ValueError is raised if the parsed integer is out of the valid range.

Examples:

var res: BiggestInt
doAssert parseBiggestInt("9223372036854775807", res, 0) == 19
doAssert res == 9223372036854775807'i64

proc parseInt(s: string; number: var int; start = 0): int {...}{.gcsafe, extern: "npuParseInt", noSideEffect, raises: [ValueError], tags: [].}

Parses an integer starting at start and stores the value into number. Result is the number of processed chars or 0 if there is no integer. ValueError is raised if the parsed integer is out of the valid range.

Examples:

var res: int
doAssert parseInt("2019", res, 0) == 4
doAssert res == 2019
doAssert parseInt("2019", res, 2) == 2
doAssert res == 19

proc parseSaturatedNatural(s: string; b: var int; start = 0): int {...}{.raises: [], tags: [].}

Parses a natural number into b. This cannot raise an overflow error. high(int) is returned for an overflow. The number of processed character is returned. This is usually what you really want to use instead of parseInt.

Examples:

var res = 0
discard parseSaturatedNatural("848", res)
doAssert res == 848

proc parseBiggestUInt(s: string; number: var BiggestUInt; start = 0): int {...}{.gcsafe, extern: "npuParseBiggestUInt", noSideEffect, raises: [ValueError], tags: [].}

Parses an unsigned integer starting at start and stores the value into number. ValueError is raised if the parsed integer is out of the valid range.

Examples:

var res: BiggestUInt
doAssert parseBiggestUInt("12", res, 0) == 2
doAssert res == 12
doAssert parseBiggestUInt("1111111111111111111", res, 0) == 19
doAssert res == 1111111111111111111'u64

proc parseUInt(s: string; number: var uint; start = 0): int {...}{.gcsafe, extern: "npuParseUInt", noSideEffect, raises: [ValueError], tags: [].}

Parses an unsigned integer starting at start and stores the value into number. ValueError is raised if the parsed integer is out of the valid range.

Examples:

var res: uint
doAssert parseUInt("3450", res) == 4
doAssert res == 3450
doAssert parseUInt("3450", res, 2) == 2
doAssert res == 50

proc parseBiggestFloat(s: string; number: var BiggestFloat; start = 0): int {...}{. magic: "ParseBiggestFloat", importc: "nimParseBiggestFloat", noSideEffect.}

Parses a float starting at start and stores the value into number. Result is the number of processed chars or 0 if a parsing error occurred. Source Edit

proc parseFloat(s: string; number: var float; start = 0): int {...}{.gcsafe, extern: "npuParseFloat", noSideEffect, raises: [], tags: [].}

Parses a float starting at start and stores the value into number. Result is the number of processed chars or 0 if there occurred a parsing error.

Examples:

var res: float
doAssert parseFloat("32", res, 0) == 2
doAssert res == 32.0
doAssert parseFloat("32.57", res, 0) == 5
doAssert res == 32.57
doAssert parseFloat("32.57", res, 3) == 2
doAssert res == 57.0

Iterators

iterator interpolatedFragments(s: string): tuple[kind: InterpolatedKind, value: string] {...}{.raises: [ValueError], tags: [].}

Tokenizes the string s into substrings for interpolation purposes.

Example:

for k, v in interpolatedFragments("  $this is ${an  example}  $$"):
  echo "(", k, ", \"", v, "\")"

Results in:

(ikString, "  ")
(ikExpr, "this")
(ikString, " is ")
(ikExpr, "an  example")
(ikString, "  ")
(ikDollar, "$")