lexers/xml.lua

   1 -- Copyright 2006-2016 Mitchell mitchell.att.foicica.com. See LICENSE.
   2 -- XML LPeg lexer.
   3
   4 local l = require('lexer')
   5 local token, word_match = l.token, l.word_match
   6 local P, R, S, V = lpeg.P, lpeg.R, lpeg.S, lpeg.V
   7
   8 local M = {_NAME = 'xml'}
   9
  10 -- Whitespace.
  11 local ws = token(l.WHITESPACE, l.space^1)
  12
  13 -- Comments and CDATA.
  14 local comment = token(l.COMMENT, '<!--' * (l.any - '-->')^0 * P('-->')^-1)
  15 local cdata = token('cdata', '<![CDATA[' * (l.any - ']]>')^0 * P(']]>')^-1)
  16
  17 -- Strings.
  18 local sq_str = l.delimited_range("'", false, true)
  19 local dq_str = l.delimited_range('"', false, true)
  20 local string = #S('\'"') * l.last_char_includes('=') *
  21                token(l.STRING, sq_str + dq_str)
  22
  23 local in_tag = P(function(input, index)
  24   local before = input:sub(1, index - 1)
  25   local s, e = before:find('<[^>]-$'), before:find('>[^<]-$')
  26   if s and e then return s > e and index or nil end
  27   if s then return index end
  28   return input:find('^[^<]->', index) and index or nil
  29 end)
  30
  31 -- Numbers.
  32 local number = #l.digit * l.last_char_includes('=') *
  33                token(l.NUMBER, l.digit^1 * P('%')^-1) * in_tag
  34
  35 local alpha = R('az', 'AZ', '\127\255')
  36 local word_char = l.alnum + S('_-:.??')
  37 local identifier = (l.alpha + S('_-:.??')) * word_char^0
  38 local namespace = token(l.OPERATOR, ':') * token('namespace', identifier)
  39
  40 -- Elements.
  41 local element = token('element', '<' * P('/')^-1 * identifier) * namespace^-1
  42
  43 -- Attributes.
  44 local attribute = token('attribute', identifier) * namespace^-1 *
  45                   #(l.space^0 * '=')
  46
  47 -- Closing tags.
  48 local close_tag = token('element', P('/')^-1 * '>')
  49
  50 -- Equals.
  51 local equals = token(l.OPERATOR, '=') * in_tag
  52
  53 -- Entities.
  54 local entity = token('entity', '&' * word_match{
  55   'lt', 'gt', 'amp', 'apos', 'quot'
  56 } * ';')
  57
  58 -- Doctypes and other markup tags.
  59 local doctype = token('doctype', P('<!DOCTYPE')) * ws *
  60                 token('doctype', identifier) * (ws * identifier)^-1 *
  61                 (1 - P('>'))^0 * token('doctype', '>')
  62
  63 -- Processing instructions.
  64 local proc_insn = token('proc_insn', P('<?') * (1 - P('?>'))^0 * P('?>')^-1)
  65
  66 M._rules = {
  67   {'whitespace', ws},
  68   {'comment', comment},
  69   {'cdata', cdata},
  70   {'doctype', doctype},
  71   {'proc_insn', proc_insn},
  72   {'element', element},
  73   {'close_tag', close_tag},
  74   {'attribute', attribute},
  75   {'equals', equals},
  76   {'string', string},
  77   {'number', number},
  78   {'entity', entity},
  79 }
  80
  81 M._tokenstyles = {
  82   element = l.STYLE_KEYWORD,
  83   namespace = l.STYLE_CLASS,
  84   attribute = l.STYLE_TYPE,
  85   cdata = l.STYLE_COMMENT,
  86   entity = l.STYLE_OPERATOR,
  87   doctype = l.STYLE_COMMENT,
  88   proc_insn = l.STYLE_COMMENT,
  89   --markup = l.STYLE_COMMENT
  90 }
  91
  92 M._foldsymbols = {
  93   _patterns = {'</?', '/>', '<!%-%-', '%-%->', '<!%[CDATA%[', '%]%]>'},
  94   element = {['<'] = 1, ['/>'] = -1, ['</'] = -1},
  95   [l.COMMENT] = {['<!--'] = 1, ['-->'] = -1},
  96   cdata = {['<![CDATA['] = 1, [']]>'] = -1}
  97 }
  98
  99 return M