unicode

This module provides support to handle the Unicode UTF-8 encoding.

There are no specialized insert, delete, add and contains procedures for seq[Rune] in this module because the generic variants of these procedures in the system module already work with it.

The current version is compatible with Unicode v12.0.0.

See also:

Types

Rune = distinct RuneImpl

Type that can hold a single Unicode code point.

A Rune may be composed with other Runes to a character on the screen. RuneImpl is the underlying type used to store Runes, currently int32.

Source Edit

Procs

proc runeLen(s: string): int {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}
Returns the number of runes of the string s.

Example:

let a = "añyóng"
doAssert a.runeLen == 6
## note: a.len == 8
Source Edit
proc runeLenAt(s: string; i: Natural): int {...}{.raises: [], tags: [].}

Returns the number of bytes the rune starting at s[i] takes.

See also:

Example:

let a = "añyóng"
doAssert a.runeLenAt(0) == 1
doAssert a.runeLenAt(1) == 2
Source Edit
proc runeAt(s: string; i: Natural): Rune {...}{.raises: [], tags: [].}

Returns the rune in s at byte index i.

See also:

Example:

let a = "añyóng"
doAssert a.runeAt(1) == "ñ".runeAt(0)
doAssert a.runeAt(2) == "ñ".runeAt(1)
doAssert a.runeAt(3) == "y".runeAt(0)
Source Edit
proc validateUtf8(s: string): int {...}{.raises: [], tags: [].}

Returns the position of the invalid byte in s if the string s does not hold valid UTF-8 data. Otherwise -1 is returned.

See also:

Source Edit
proc toUTF8(c: Rune): string {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Converts a rune into its UTF-8 representation.

See also:

Example:

let a = "añyóng"
doAssert a.runeAt(1).toUTF8 == "ñ"
Source Edit
proc add(s: var string; c: Rune) {...}{.raises: [], tags: [].}
Adds a rune c to a string s.

Example:

var s = "abc"
let c = "ä".runeAt(0)
s.add(c)
doAssert s == "abcä"
Source Edit
proc `$`(rune: Rune): string {...}{.raises: [], tags: [].}

An alias for toUTF8.

See also:

Source Edit
proc `$`(runes: seq[Rune]): string {...}{.raises: [], tags: [].}

Converts a sequence of Runes to a string.

See also:

Example:

let
  someString = "öÑ"
  someRunes = toRunes(someString)
doAssert $someRunes == someString
Source Edit
proc runeOffset(s: string; pos: Natural; start: Natural = 0): int {...}{.raises: [],
    tags: [].}

Returns the byte position of rune at position pos in s with an optional start byte position. Returns the special value -1 if it runs out of the string.

Beware: This can lead to unoptimized code and slow execution! Most problems can be solved more efficiently by using an iterator or conversion to a seq of Rune.

See also:

Example:

let a = "añyóng"
doAssert a.runeOffset(1) == 1
doAssert a.runeOffset(3) == 4
doAssert a.runeOffset(4) == 6
Source Edit
proc runeReverseOffset(s: string; rev: Positive): (int, int) {...}{.raises: [],
    tags: [].}

Returns a tuple with the byte offset of the rune at position rev in s, counting from the end (starting with 1) and the total number of runes in the string.

Returns a negative value for offset if there are to few runes in the string to satisfy the request.

Beware: This can lead to unoptimized code and slow execution! Most problems can be solved more efficiently by using an iterator or conversion to a seq of Rune.

See also:

Source Edit
proc runeAtPos(s: string; pos: int): Rune {...}{.raises: [], tags: [].}

Returns the rune at position pos.

Beware: This can lead to unoptimized code and slow execution! Most problems can be solved more efficiently by using an iterator or conversion to a seq of Rune.

See also:

Source Edit
proc runeStrAtPos(s: string; pos: Natural): string {...}{.raises: [], tags: [].}

Returns the rune at position pos as UTF8 String.

Beware: This can lead to unoptimized code and slow execution! Most problems can be solved more efficiently by using an iterator or conversion to a seq of Rune.

See also:

Source Edit
proc runeSubStr(s: string; pos: int; len: int = int.high): string {...}{.raises: [],
    tags: [].}

Returns the UTF-8 substring starting at code point pos with len code points.

If pos or len is negative they count from the end of the string. If len is not given it means the longest possible string.

Example:

let s = "Hänsel  ««: 10,00€"
doAssert(runeSubStr(s, 0, 2) == "Hä")
doAssert(runeSubStr(s, 10, 1) == ":")
doAssert(runeSubStr(s, -6) == "10,00€")
doAssert(runeSubStr(s, 10) == ": 10,00€")
doAssert(runeSubStr(s, 12, 5) == "10,00")
doAssert(runeSubStr(s, -6, 3) == "10,")
Source Edit
proc `<=%`(a, b: Rune): bool {...}{.raises: [], tags: [].}
Checks if code point of a is smaller or equal to code point of b.

Example:

let
  a = "ú".runeAt(0)
  b = "ü".runeAt(0)
doAssert a <=% b
Source Edit
proc `<%`(a, b: Rune): bool {...}{.raises: [], tags: [].}
Checks if code point of a is smaller than code point of b.

Example:

let
  a = "ú".runeAt(0)
  b = "ü".runeAt(0)
doAssert a <% b
Source Edit
proc `==`(a, b: Rune): bool {...}{.raises: [], tags: [].}
Checks if two runes are equal. Source Edit
proc toLower(c: Rune): Rune {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Converts c into lower case. This works for any rune.

If possible, prefer toLower over toUpper.

See also:

Source Edit
proc toUpper(c: Rune): Rune {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Converts c into upper case. This works for any rune.

If possible, prefer toLower over toUpper.

See also:

Source Edit
proc toTitle(c: Rune): Rune {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Converts c to title case.

See also:

Source Edit
proc isLower(c: Rune): bool {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Returns true if c is a lower case rune.

If possible, prefer isLower over isUpper.

See also:

Source Edit
proc isUpper(c: Rune): bool {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Returns true if c is a upper case rune.

If possible, prefer isLower over isUpper.

See also:

Source Edit
proc isAlpha(c: Rune): bool {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Returns true if c is an alpha rune (i.e., a letter).

See also:

Source Edit
proc isTitle(c: Rune): bool {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Returns true if c is a Unicode titlecase code point.

See also:

Source Edit
proc isWhiteSpace(c: Rune): bool {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Returns true if c is a Unicode whitespace code point.

See also:

Source Edit
proc isCombining(c: Rune): bool {...}{.gcsafe, extern: "nuc$1", raises: [], tags: [].}

Returns true if c is a Unicode combining code unit.

See also:

Source Edit
proc isAlpha(s: string): bool {...}{.noSideEffect, gcsafe, extern: "nuc$1Str",
                                raises: [], tags: [].}
Returns true if s contains all alphabetic runes.

Example:

let a = "añyóng"
doAssert a.isAlpha
Source Edit
proc isSpace(s: string): bool {...}{.noSideEffect, gcsafe, extern: "nuc$1Str",
                                raises: [], tags: [].}
Returns true if s contains all whitespace runes.

Example:

let a = "\t\l \v\r\f"
doAssert a.isSpace
Source Edit
proc toUpper(s: string): string {...}{.noSideEffect, gcsafe, extern: "nuc$1Str",
                                  raises: [], tags: [].}
Converts s into upper-case runes.

Example:

doAssert toUpper("abγ") == "ABΓ"
Source Edit
proc toLower(s: string): string {...}{.noSideEffect, gcsafe, extern: "nuc$1Str",
                                  raises: [], tags: [].}
Converts s into lower-case runes.

Example:

doAssert toLower("ABΓ") == "abγ"
Source Edit
proc swapCase(s: string): string {...}{.noSideEffect, gcsafe, extern: "nuc$1",
                                   raises: [], tags: [].}

Swaps the case of runes in s.

Returns a new string such that the cases of all runes are swapped if possible.

Example:

doAssert swapCase("Αlpha Βeta Γamma") == "αLPHA βETA γAMMA"
Source Edit
proc capitalize(s: string): string {...}{.noSideEffect, gcsafe, extern: "nuc$1",
                                     raises: [], tags: [].}
Converts the first character of s into an upper-case rune.

Example:

doAssert capitalize("βeta") == "Βeta"
Source Edit
proc translate(s: string; replacements: proc (key: string): string): string {...}{.
    gcsafe, extern: "nuc$1", raises: [], tags: [].}

Translates words in a string using the replacements proc to substitute words inside s with their replacements.

replacements is any proc that takes a word and returns a new word to fill it's place.

Example:

proc wordToNumber(s: string): string =
  case s
  of "one": "1"
  of "two": "2"
  else: s
let a = "one two three four"
doAssert a.translate(wordToNumber) == "1 2 three four"
Source Edit
proc title(s: string): string {...}{.noSideEffect, gcsafe, extern: "nuc$1",
                                raises: [], tags: [].}

Converts s to a unicode title.

Returns a new string such that the first character in each word inside s is capitalized.

Example:

doAssert title("αlpha βeta γamma") == "Αlpha Βeta Γamma"
Source Edit
proc toRunes(s: string): seq[Rune] {...}{.raises: [], tags: [].}

Obtains a sequence containing the Runes in s.

See also:

  • $ proc for a reverse operation

Example:

let a = toRunes("aáä")
doAssert a == @["a".runeAt(0), "á".runeAt(0), "ä".runeAt(0)]
Source Edit
proc cmpRunesIgnoreCase(a, b: string): int {...}{.gcsafe, extern: "nuc$1",
    raises: [], tags: [].}
Compares two UTF-8 strings and ignores the case. Returns:

0 if a == b
< 0 if a < b
> 0 if a > b

Source Edit
proc reversed(s: string): string {...}{.raises: [], tags: [].}

Returns the reverse of s, interpreting it as runes.

Unicode combining characters are correctly interpreted as well.

Example:

assert reversed("Reverse this!") == "!siht esreveR"
assert reversed("先秦兩漢") == "漢兩秦先"
assert reversed("as⃝df̅") == "f̅ds⃝a"
assert reversed("a⃞b⃞c⃞") == "c⃞b⃞a⃞"
Source Edit
proc graphemeLen(s: string; i: Natural): Natural {...}{.raises: [], tags: [].}
The number of bytes belonging to byte index s[i], including following combining code unit.

Example:

let a = "añyóng"
doAssert a.graphemeLen(1) == 2 ## ñ
doAssert a.graphemeLen(2) == 1
doAssert a.graphemeLen(4) == 2 ## ó
Source Edit
proc lastRune(s: string; last: int): (Rune, int) {...}{.raises: [], tags: [].}
Length of the last rune in s[0..last]. Returns the rune and its length in bytes. Source Edit
proc size(r: Rune): int {...}{.noSideEffect, raises: [], tags: [].}
Returns the number of bytes the rune r takes.

Example:

let a = toRunes "aá"
doAssert size(a[0]) == 1
doAssert size(a[1]) == 2
Source Edit
proc splitWhitespace(s: string): seq[string] {...}{.noSideEffect, gcsafe,
    extern: "ncuSplitWhitespace", raises: [], tags: [].}
The same as the splitWhitespace iterator, but is a proc that returns a sequence of substrings. Source Edit
proc split(s: string; seps: openArray[Rune] = unicodeSpaces; maxsplit: int = -1): seq[
    string] {...}{.noSideEffect, gcsafe, extern: "nucSplitRunes", raises: [],
              tags: [].}
The same as the split iterator, but is a proc that returns a sequence of substrings. Source Edit
proc split(s: string; sep: Rune; maxsplit: int = -1): seq[string] {...}{.
    noSideEffect, gcsafe, extern: "nucSplitRune", raises: [], tags: [].}
The same as the split iterator, but is a proc that returns a sequence of substrings. Source Edit
proc strip(s: string; leading = true; trailing = true;
           runes: openArray[Rune] = unicodeSpaces): string {...}{.noSideEffect,
    gcsafe, extern: "nucStrip", raises: [], tags: [].}

Strips leading or trailing runes from s and returns the resulting string.

If leading is true (default), leading runes are stripped. If trailing is true (default), trailing runes are stripped. If both are false, the string is returned unchanged.

Example:

let a = "\táñyóng   "
doAssert a.strip == "áñyóng"
doAssert a.strip(leading = false) == "\táñyóng"
doAssert a.strip(trailing = false) == "áñyóng   "
Source Edit
proc repeat(c: Rune; count: Natural): string {...}{.noSideEffect, gcsafe,
    extern: "nucRepeatRune", raises: [], tags: [].}

Returns a string of count Runes c.

The returned string will have a rune-length of count.

Example:

let a = "ñ".runeAt(0)
doAssert a.repeat(5) == "ñññññ"
Source Edit
proc align(s: string; count: Natural; padding = ' '.Rune): string {...}{.
    noSideEffect, gcsafe, extern: "nucAlignString", raises: [], tags: [].}

Aligns a unicode string s with padding, so that it has a rune-length of count.

padding characters (by default spaces) are added before s resulting in right alignment. If s.runelen >= count, no spaces are added and s is returned unchanged. If you need to left align a string use the alignLeft proc.

Example:

assert align("abc", 4) == " abc"
assert align("a", 0) == "a"
assert align("1232", 6) == "  1232"
assert align("1232", 6, '#'.Rune) == "##1232"
assert align("Åge", 5) == "  Åge"
assert align("×", 4, '_'.Rune) == "___×"
Source Edit
proc alignLeft(s: string; count: Natural; padding = ' '.Rune): string {...}{.
    noSideEffect, raises: [], tags: [].}

Left-aligns a unicode string s with padding, so that it has a rune-length of count.

padding characters (by default spaces) are added after s resulting in left alignment. If s.runelen >= count, no spaces are added and s is returned unchanged. If you need to right align a string use the align proc.

Example:

assert alignLeft("abc", 4) == "abc "
assert alignLeft("a", 0) == "a"
assert alignLeft("1232", 6) == "1232  "
assert alignLeft("1232", 6, '#'.Rune) == "1232##"
assert alignLeft("Åge", 5) == "Åge  "
assert alignLeft("×", 4, '_'.Rune) == "×___"
Source Edit

Iterators

iterator runes(s: string): Rune {...}{.raises: [], tags: [].}
Iterates over any rune of the string s returning runes. Source Edit
iterator utf8(s: string): string {...}{.raises: [], tags: [].}

Iterates over any rune of the string s returning utf8 values.

See also:

Source Edit
iterator split(s: string; seps: openArray[Rune] = unicodeSpaces;
               maxsplit: int = -1): string {...}{.raises: [], tags: [].}

Splits the unicode string s into substrings using a group of separators.

Substrings are separated by a substring containing only seps.

for word in split("this\lis an\texample"):
  writeLine(stdout, word)

...generates this output:

"this"
"is"
"an"
"example"

And the following code:

for word in split("this:is;an$example", {';', ':', '$'}):
  writeLine(stdout, word)

...produces the same output as the first example. The code:

let date = "2012-11-20T22:08:08.398990"
let separators = {' ', '-', ':', 'T'}
for number in split(date, separators):
  writeLine(stdout, number)

...results in:

"2012"
"11"
"20"
"22"
"08"
"08.398990"
Source Edit
iterator splitWhitespace(s: string): string {...}{.raises: [], tags: [].}
Splits a unicode string at whitespace runes. Source Edit
iterator split(s: string; sep: Rune; maxsplit: int = -1): string {...}{.raises: [],
    tags: [].}

Splits the unicode string s into substrings using a single separator.

Substrings are separated by the rune sep. The code:

for word in split(";;this;is;an;;example;;;", ';'):
  writeLine(stdout, word)

Results in:

""
""
"this"
"is"
"an"
""
"example"
""
""
""
Source Edit

Templates

template fastRuneAt(s: string; i: int; result: untyped; doInc = true)

Returns the rune s[i] in result.

If doInc == true (default), i is incremented by the number of bytes that have been processed.

Source Edit
template fastToUTF8Copy(c: Rune; s: var string; pos: int; doInc = true)

Copies UTF-8 representation of c into the preallocated string s starting at position pos.

If doInc == true (default), pos is incremented by the number of bytes that have been processed.

To be the most efficient, make sure s is preallocated with an additional amount equal to the byte length of c.

See also:

Source Edit

© 2006–2021 Andreas Rumpf
Licensed under the MIT License.
https://nim-lang.org/docs/unicode.html