flow like the river

BACK_BACK/node_modules/grapheme-breaker/src/GraphemeBreaker.coffee

@@ -0,0 +1,149 @@
+{CR,LF,Control,Extend,Regional_Indicator,SpacingMark,L,V,T,LV,LVT} = require './classes.json'
+UnicodeTrie = require 'unicode-trie'
+fs = require 'fs'
+classTrie = new UnicodeTrie fs.readFileSync __dirname + '/classes.trie'
+
+# Gets a code point from a UTF-16 string
+# handling surrogate pairs appropriately
+codePointAt = (str, idx) ->
+  idx = idx or 0
+  code = str.charCodeAt(idx)
+
+  # High surrogate
+  if 0xD800 <= code <= 0xDBFF
+    hi = code
+    low = str.charCodeAt(idx + 1)    
+    if 0xDC00 <= low <= 0xDFFF
+      return ((hi - 0xD800) * 0x400) + (low - 0xDC00) + 0x10000
+      
+    return hi
+  
+  # Low surrogate
+  if 0xDC00 <= code <= 0xDFFF
+    hi = str.charCodeAt(idx - 1)
+    low = code
+    if 0xD800 <= hi <= 0xDBFF
+      return ((hi - 0xD800) * 0x400) + (low - 0xDC00) + 0x10000
+      
+    return low
+    
+  return code
+  
+# Returns whether a break is allowed between the 
+# two given grapheme breaking classes
+shouldBreak = (previous, current) ->  
+  # GB3. CR X LF
+  if previous is CR and current is LF
+    return false
+    
+  # GB4. (Control|CR|LF) ÷
+  else if previous in [Control, CR, LF]
+    return true
+    
+  # GB5. ÷ (Control|CR|LF)
+  else if current in [Control, CR, LF]
+    return true
+    
+  # GB6. L X (L|V|LV|LVT)
+  else if previous is L and current in [L, V, LV, LVT]
+    return false
+    
+  # GB7. (LV|V) X (V|T)
+  else if previous in [LV, V] and current in [V, T]
+    return false
+    
+  # GB8. (LVT|T) X (T)
+  else if previous in [LVT, T] and current is T
+    return false
+    
+  # GB8a. Regional_Indicator X Regional_Indicator
+  else if previous is Regional_Indicator and current is Regional_Indicator
+    return false
+    
+  # GB9. X Extend
+  else if current is Extend
+    return false
+    
+  # GB9a. X SpacingMark
+  else if current is SpacingMark
+    return false
+    
+  # GB9b. Prepend X (there are currently no characters with this class)
+  # else if previous is Prepend
+  #   return false
+    
+  # GB10. Any ÷ Any
+  return true
+
+# Returns the next grapheme break in the string after the given index
+exports.nextBreak = (string, index = 0) ->
+  if index < 0
+    return 0
+    
+  if index >= string.length - 1
+    return string.length
+    
+  prev = classTrie.get codePointAt(string, index)
+  for i in [index + 1...string.length] by 1
+    # check for already processed low surrogates
+    continue if 0xd800 <= string.charCodeAt(i - 1) <= 0xdbff and 
+                0xdc00 <= string.charCodeAt(i)     <= 0xdfff
+    
+    next = classTrie.get codePointAt(string, i)
+    if shouldBreak prev, next
+      return i
+      
+    prev = next
+      
+  return string.length
+  
+# Returns the next grapheme break in the string before the given index
+exports.previousBreak = (string, index = string.length) ->
+  if index > string.length
+    return string.length
+  
+  if index <= 1
+    return 0
+  
+  index--
+  next = classTrie.get codePointAt(string, index)
+  for i in [index - 1..0] by -1
+    # check for already processed high surrogates
+    continue if 0xd800 <= string.charCodeAt(i)     <= 0xdbff and 
+                0xdc00 <= string.charCodeAt(i + 1) <= 0xdfff
+    
+    prev = classTrie.get codePointAt(string, i)
+    if shouldBreak prev, next
+      return i + 1
+      
+    next = prev
+      
+  return 0
+  
+# Breaks the given string into an array of grapheme cluster strings
+exports.break = (str) ->
+  res = []
+  index = 0
+  
+  while (brk = exports.nextBreak(str, index)) < str.length
+    res.push str.slice(index, brk)
+    index = brk
+    
+  if index < str.length
+    res.push str.slice(index)
+    
+  return res
+
+# Returns the number of grapheme clusters there are in the given string
+exports.countBreaks = (str) ->
+  count = 0
+  index = 0
+  
+  while (brk = exports.nextBreak(str, index)) < str.length
+    index = brk
+    count++
+    
+  if index < str.length
+    count++
+      
+  return count

BACK_BACK/node_modules/grapheme-breaker/src/GraphemeBreaker.js

@@ -0,0 +1,138 @@
+// Generated by CoffeeScript 1.8.0
+(function() {
+  var CR, Control, Extend, L, LF, LV, LVT, Regional_Indicator, SpacingMark, T, UnicodeTrie, V, classTrie, codePointAt, fs, shouldBreak, _ref;
+
+  _ref = require('./classes.json'), CR = _ref.CR, LF = _ref.LF, Control = _ref.Control, Extend = _ref.Extend, Regional_Indicator = _ref.Regional_Indicator, SpacingMark = _ref.SpacingMark, L = _ref.L, V = _ref.V, T = _ref.T, LV = _ref.LV, LVT = _ref.LVT;
+
+  UnicodeTrie = require('unicode-trie');
+
+  fs = require('fs');
+
+  classTrie = new UnicodeTrie(fs.readFileSync(__dirname + '/classes.trie'));
+
+  codePointAt = function(str, idx) {
+    var code, hi, low;
+    idx = idx || 0;
+    code = str.charCodeAt(idx);
+    if ((0xD800 <= code && code <= 0xDBFF)) {
+      hi = code;
+      low = str.charCodeAt(idx + 1);
+      if ((0xDC00 <= low && low <= 0xDFFF)) {
+        return ((hi - 0xD800) * 0x400) + (low - 0xDC00) + 0x10000;
+      }
+      return hi;
+    }
+    if ((0xDC00 <= code && code <= 0xDFFF)) {
+      hi = str.charCodeAt(idx - 1);
+      low = code;
+      if ((0xD800 <= hi && hi <= 0xDBFF)) {
+        return ((hi - 0xD800) * 0x400) + (low - 0xDC00) + 0x10000;
+      }
+      return low;
+    }
+    return code;
+  };
+
+  shouldBreak = function(previous, current) {
+    if (previous === CR && current === LF) {
+      return false;
+    } else if (previous === Control || previous === CR || previous === LF) {
+      return true;
+    } else if (current === Control || current === CR || current === LF) {
+      return true;
+    } else if (previous === L && (current === L || current === V || current === LV || current === LVT)) {
+      return false;
+    } else if ((previous === LV || previous === V) && (current === V || current === T)) {
+      return false;
+    } else if ((previous === LVT || previous === T) && current === T) {
+      return false;
+    } else if (previous === Regional_Indicator && current === Regional_Indicator) {
+      return false;
+    } else if (current === Extend) {
+      return false;
+    } else if (current === SpacingMark) {
+      return false;
+    }
+    return true;
+  };
+
+  exports.nextBreak = function(string, index) {
+    var i, next, prev, _i, _ref1, _ref2, _ref3, _ref4;
+    if (index == null) {
+      index = 0;
+    }
+    if (index < 0) {
+      return 0;
+    }
+    if (index >= string.length - 1) {
+      return string.length;
+    }
+    prev = classTrie.get(codePointAt(string, index));
+    for (i = _i = _ref1 = index + 1, _ref2 = string.length; _i < _ref2; i = _i += 1) {
+      if ((0xd800 <= (_ref3 = string.charCodeAt(i - 1)) && _ref3 <= 0xdbff) && (0xdc00 <= (_ref4 = string.charCodeAt(i)) && _ref4 <= 0xdfff)) {
+        continue;
+      }
+      next = classTrie.get(codePointAt(string, i));
+      if (shouldBreak(prev, next)) {
+        return i;
+      }
+      prev = next;
+    }
+    return string.length;
+  };
+
+  exports.previousBreak = function(string, index) {
+    var i, next, prev, _i, _ref1, _ref2, _ref3;
+    if (index == null) {
+      index = string.length;
+    }
+    if (index > string.length) {
+      return string.length;
+    }
+    if (index <= 1) {
+      return 0;
+    }
+    index--;
+    next = classTrie.get(codePointAt(string, index));
+    for (i = _i = _ref1 = index - 1; _i >= 0; i = _i += -1) {
+      if ((0xd800 <= (_ref2 = string.charCodeAt(i)) && _ref2 <= 0xdbff) && (0xdc00 <= (_ref3 = string.charCodeAt(i + 1)) && _ref3 <= 0xdfff)) {
+        continue;
+      }
+      prev = classTrie.get(codePointAt(string, i));
+      if (shouldBreak(prev, next)) {
+        return i + 1;
+      }
+      next = prev;
+    }
+    return 0;
+  };
+
+  exports["break"] = function(str) {
+    var brk, index, res;
+    res = [];
+    index = 0;
+    while ((brk = exports.nextBreak(str, index)) < str.length) {
+      res.push(str.slice(index, brk));
+      index = brk;
+    }
+    if (index < str.length) {
+      res.push(str.slice(index));
+    }
+    return res;
+  };
+
+  exports.countBreaks = function(str) {
+    var brk, count, index;
+    count = 0;
+    index = 0;
+    while ((brk = exports.nextBreak(str, index)) < str.length) {
+      index = brk;
+      count++;
+    }
+    if (index < str.length) {
+      count++;
+    }
+    return count;
+  };
+
+}).call(this);

BACK_BACK/node_modules/grapheme-breaker/src/classes.json

@@ -0,0 +1 @@
+{"Other":0,"CR":1,"LF":2,"Control":3,"Extend":4,"Regional_Indicator":5,"SpacingMark":6,"L":7,"V":8,"T":9,"LV":10,"LVT":11}

BACK_BACK/node_modules/grapheme-breaker/src/generate_data.coffee

@@ -0,0 +1,33 @@
+request = require 'request'
+fs = require 'fs'
+UnicodeTrieBuilder = require 'unicode-trie/builder'
+
+UNICODE_VERSION = '8.0.0'
+BASE_URL = "http://www.unicode.org/Public/#{UNICODE_VERSION}/ucd"
+
+# this loads the GraphemeBreakProperty.txt file for Unicode 8.0.0 and parses it to
+# combine ranges and generate CoffeeScript
+request "#{BASE_URL}/auxiliary/GraphemeBreakProperty.txt", (err, res, data) ->
+  re = /^([0-9A-F]+)(?:\.\.([0-9A-F]+))?\s*;\s*([A-Za-z_]+)/gm
+  nextClass = 1
+  classes = 
+    Other: 0
+  
+  trie = new UnicodeTrieBuilder classes.Other
+
+  # collect entries in the table into ranges
+  # to keep things smaller.
+  while match = re.exec(data)
+    start = match[1]
+    end = match[2] ? start  
+    type = match[3]
+    unless classes[type]?
+      classes[type] = nextClass++
+        
+    trie.setRange parseInt(start, 16), parseInt(end, 16), classes[type]
+
+  # write the trie to a file
+  fs.writeFile __dirname + '/classes.trie', trie.toBuffer()
+  
+  # write classes to a file
+  fs.writeFile __dirname + '/classes.json', JSON.stringify classes

BACK_BACK/node_modules/grapheme-breaker/src/generate_data.js

@@ -0,0 +1,36 @@
+// Generated by CoffeeScript 1.8.0
+(function() {
+  var BASE_URL, UNICODE_VERSION, UnicodeTrieBuilder, fs, request;
+
+  request = require('request');
+
+  fs = require('fs');
+
+  UnicodeTrieBuilder = require('unicode-trie/builder');
+
+  UNICODE_VERSION = '8.0.0';
+
+  BASE_URL = "http://www.unicode.org/Public/" + UNICODE_VERSION + "/ucd";
+
+  request("" + BASE_URL + "/auxiliary/GraphemeBreakProperty.txt", function(err, res, data) {
+    var classes, end, match, nextClass, re, start, trie, type, _ref;
+    re = /^([0-9A-F]+)(?:\.\.([0-9A-F]+))?\s*;\s*([A-Za-z_]+)/gm;
+    nextClass = 1;
+    classes = {
+      Other: 0
+    };
+    trie = new UnicodeTrieBuilder(classes.Other);
+    while (match = re.exec(data)) {
+      start = match[1];
+      end = (_ref = match[2]) != null ? _ref : start;
+      type = match[3];
+      if (classes[type] == null) {
+        classes[type] = nextClass++;
+      }
+      trie.setRange(parseInt(start, 16), parseInt(end, 16), classes[type]);
+    }
+    fs.writeFile(__dirname + '/classes.trie', trie.toBuffer());
+    return fs.writeFile(__dirname + '/classes.json', JSON.stringify(classes));
+  });
+
+}).call(this);