std.strings#

mars

takepkg std.strings;

func m{
    words = strings.split_whitespace("  red green  blue ");
    out(strings.join(words, ","));          // red,green,blue
    out(strings.pad_start("7", 3, "0"));    // 007
}

Positions and widths count characters (grapheme clusters), the same unit as len() and lenslice(). Searching, splitting, and replacing match whole characters only: a match must start and end at character boundaries, so a position from find always works with lenslice. A combining accent inside a character is not found on its own, and "\r\n" is one character, so split(text, "\n") does not split Windows line endings; use lines(text) for those. The package is named strings so that it does not hide the built-in string(value) conversion.

Function	Result
`split(text, separator)`	Array of the parts between separators; an empty separator raises `RangeError`
`split_whitespace(text)`	Array of the words between runs of whitespace
`lines(text)`	Array of lines, splitting at `\n` or `\r\n`
`join(items, separator)`	One string from an array of strings
`trim(text)` / `trim_start(text)` / `trim_end(text)`	Text without surrounding whitespace
`starts_with(text, prefix)` / `ends_with(text, suffix)` / `contains(text, part)`	Boolean
`find(text, part)` / `rfind(text, part)`	Character position of the first/last occurrence, or -1
`replace(text, old, new)`	Every occurrence replaced; an empty `old` raises `RangeError`
`upper(text)` / `lower(text)`	Unicode case conversion (`straße` becomes `STRASSE`)
`repeated(text, count)`	`text` repeated `count` times
`pad_start(text, width, fill)` / `pad_end(text, width, fill)`	Padded to `width` characters with a one-character `fill`; a fill that merges with its neighbours (a lone combining accent, a flag letter) raises `RangeError`

The Marslang API is in std/strings.mars; the text operations are native (std/rs/string.rs).