markdown-it/lib/rules_inline/state_inline.js

// Inline parser state

'use strict';


var Token          = require('../token');
var isWhiteSpace   = require('../common/utils').isWhiteSpace;
var isPunctChar    = require('../common/utils').isPunctChar;
var isMdAsciiPunct = require('../common/utils').isMdAsciiPunct;


function StateInline(src, md, env, outTokens) {
  this.src = src;
  this.env = env;
  this.md = md;
  this.tokens = outTokens;

  this.pos = 0;
  this.posMax = this.src.length;
  this.level = 0;
  this.pending = '';
  this.pendingLevel = 0;

  this.cache = {};        // Stores { start: end } pairs. Useful for backtrack
                          // optimization of pairs parse (emphasis, strikes).

  this.delimiters = [];   // Emphasis-like delimiters
}


// Flush pending text
//
StateInline.prototype.pushPending = function () {
  var token = new Token('text', '', 0);
  token.content = this.pending;
  token.level = this.pendingLevel;
  this.tokens.push(token);
  this.pending = '';
  return token;
};


// Push new token to "stream".
// If pending text exists - flush it as text token
//
StateInline.prototype.push = function (type, tag, nesting) {
  if (this.pending) {
    this.pushPending();
  }

  var token = new Token(type, tag, nesting);

  if (nesting < 0) { this.level--; }
  token.level = this.level;
  if (nesting > 0) { this.level++; }

  this.pendingLevel = this.level;
  this.tokens.push(token);
  return token;
};


// Scan a sequence of emphasis-like markers, and determine whether
// it can start an emphasis sequence or end an emphasis sequence.
//
//  - start - position to scan from (it should point at a valid marker);
//  - canSplitWord - determine if these markers can be found inside a word
//
StateInline.prototype.scanDelims = function (start, canSplitWord) {
  var pos = start, lastChar, nextChar, count, can_open, can_close,
      isLastWhiteSpace, isLastPunctChar,
      isNextWhiteSpace, isNextPunctChar,
      left_flanking = true,
      right_flanking = true,
      max = this.posMax,
      marker = this.src.charCodeAt(start);

  // treat beginning of the line as a whitespace
  lastChar = start > 0 ? this.src.charCodeAt(start - 1) : 0x20;

  while (pos < max && this.src.charCodeAt(pos) === marker) { pos++; }

  count = pos - start;

  // treat end of the line as a whitespace
  nextChar = pos < max ? this.src.charCodeAt(pos) : 0x20;

  isLastPunctChar = isMdAsciiPunct(lastChar) || isPunctChar(String.fromCharCode(lastChar));
  isNextPunctChar = isMdAsciiPunct(nextChar) || isPunctChar(String.fromCharCode(nextChar));

  isLastWhiteSpace = isWhiteSpace(lastChar);
  isNextWhiteSpace = isWhiteSpace(nextChar);

  if (isNextWhiteSpace) {
    left_flanking = false;
  } else if (isNextPunctChar) {
    if (!(isLastWhiteSpace || isLastPunctChar)) {
      left_flanking = false;
    }
  }

  if (isLastWhiteSpace) {
    right_flanking = false;
  } else if (isLastPunctChar) {
    if (!(isNextWhiteSpace || isNextPunctChar)) {
      right_flanking = false;
    }
  }

  if (!canSplitWord) {
    can_open  = left_flanking  && (!right_flanking || isLastPunctChar);
    can_close = right_flanking && (!left_flanking  || isNextPunctChar);
  } else {
    can_open  = left_flanking;
    can_close = right_flanking;
  }

  return {
    can_open:  can_open,
    can_close: can_close,
    length:    count
  };
};


// re-export Token class to use in block rules
StateInline.prototype.Token = Token;


module.exports = StateInline;