String

NOTE

All pattern parameter in lua string library accepts number which would be converted as ASCII

Pattern Matching

Range of Occurrence

• parameters

  • pattern: string | number: pattern to find, may include capture groups
  • init?: int: the index to start searching
  • plain?: boolean: whether to match in literal

• returns

  • start: start index of the first match on the whole pattern
  • tail: start index of the first match on the whole pattern(do not use end as variable name, it's reserved keyword)
  • ...matches: arbitrary number of captured groups that can be unpacked to variables

local text = '<tag>content</tag>'

-- find range of first occurrence
-- and its captured groups, kind of useless though
local start, tail, tag, content = text:find('<(%w+)>(%w+)</%w+>')

Replace

string.gsub: replace all occurrences

• parameters

  • pattern: string | number: pattern to replace
  • repl: string | number | function | table
    • string: may use %n to reference capture groups by index n to transform from groups
    • fun(match: string): string: to replace the occurrence by tranformed string from the match(match parameter is just the first match)
    • table<string, string>: a finite pairs for match(key) and replacement(value)
  • n?: int: max count of substitutions could be made

• returns

  • text: substituted string
  • count: substitution count made

local text = 'snake_case_is_awful'

local sub, count = text:gsub('_', '-') -- snake-case-is-awful, 3

-- transform by captured groups
local sub, count = text:gsub('(%l+)_(%l+)', '%1-%2', 1) -- snake-case_is_awful, 1

-- capitalize word for 2 occurrences
local sub, count = text:gsub('%l+', function(match)
  ---@cast match string
  local cap = match:sub(1, 1):upper()
  return cap .. match:sub(2)
end, 2) -- Snake_Case_is_awful, 2

local sub, count = text:gsub('%l+', {
  snake = 'python',
  _case = '',
}) -- python_case_is_awful, 4
-- because the table were tried anyway and got nil when key is not registered

Get Captured Groups

Match for One time

• parameters

  • pattern: pattern to match that could have multiple captured groups
  • init?: int: index to start matching

• returns

  • ...matches: arbitrary number of matches that can be unpacked

local text = 'John Smith is not me'

-- only match on one time
local first_name, last_name = text:match('(%w+) (%w+)') -- John, Smith

Match for Whole String

string.gmatch is a counterpart for string.match to match all occurrence for the pattern(no optional init parameter however) as a iterator function that must access through for..in statement.

• parameters

  • pattern: pattern to match, may include captured groups

• returns

  • ...matchs: arbitrary number of captured groups

local text = 'John Smith is not me'

for first_name, last_name in text:gmatch('(%w+) (%w+)') do
  print(first_name)
  print(last_name)
end
-- John
-- Smith
-- is
-- not

CAUTION

Do not use ^ in pattern for string.gmatch, it terminates the iteration from the beginning