add new .build addon (thanks to @LeaVerou)

README.md

@@ -97,7 +97,7 @@ unicodeWord.test('日本語'); // -> true
 unicodeWord.test('العربية'); // -> true
 ~~~
 
-The base script adds `\p{L}` (and its alias, `\p{Letter}`), but other Unicode categories, scripts, blocks, and properties require addon packages. Try these next examples after additionally including `unicode-scripts.js`:
+The base script adds `\p{Letter}` and its alias `\p{L}`, but other Unicode categories, scripts, blocks, and properties require addon packages. Try these next examples after additionally including `unicode-scripts.js`:
 
 ~~~ js
 XRegExp('^\\p{Hiragana}+$').test('ひらがな'); // -> true
@@ -109,9 +109,9 @@ XRegExp uses the Unicode 6.1 character database (released January 2012).
 More details: [Addons: Unicode](http://xregexp.com/plugins/#unicode).
 
 
-## XRegExp Match Recursive
+## XRegExp.matchRecursive
 
-First include the Match Recursive script:
+First include the script:
 
 ~~~ html
 <script src="xregexp.js"></script>
@@ -156,6 +156,30 @@ XRegExp.matchRecursive(str, '<', '>', 'gy');
 More details: [Addons: Match Recursive](http://xregexp.com/plugins/#matchRecursive).
 
 
+## XRegExp.build
+
+First include the script:
+
+~~~ html
+<script src="xregexp.js"></script>
+<script src="addons/build.js"></script>
+~~~
+
+You can then build complex regular expressions that use named subpatterns for readability and code reuse:
+
+~~~ js
+var color = XRegExp.build("{{keyword}}|{{func}}|{{hex}}", {
+    keyword: /^(?:red|tan|[a-z]{4,20})$/,
+    func: XRegExp.build("^(?:rgb|hsl)a?\\((?:\\s*{{number}}%?\\s*,?\\s*){3,4}\\)$", {
+        number: /^-?\d+(?:\.\d+)?$/
+    }),
+    hex: /^#(?:[0-9a-f]{1,2}){3}$/
+});
+~~~
+
+The `{{…}}` syntax works only for regexes created by `XRegExp.build`.
+
+
 ## How to run server-side tests
 
 ~~~ bash
@@ -176,6 +200,8 @@ XRegExp and addons copyright 2007-2012 by [Steven Levithan](http://stevenlevitha
 
 Tools: Unicode range generators by [Mathias Bynens](http://mathiasbynens.be/). Source file concatenator by [Bjarke Walling](http://twitter.com/walling).
 
+Thanks to [Lea Verou](http://lea.verou.me/) for the [inspiration](http://lea.verou.me/2011/03/create-complex-regexps-more-easily/) behind XRegExp.build.
+
 All code released under the [MIT License](http://opensource.org/licenses/mit-license.php).
 
 Fork me to show support, fix, and extend.

src/addons/build.js

@@ -0,0 +1,66 @@
+/*!
+ * XRegExp.build v0.1.0-beta
+ * Copyright 2012 Steven Levithan <http://xregexp.com/>
+ * Available under the MIT License
+ * Based on RegExp.create by Lea Verou <http://lea.verou.me/2011/03/create-complex-regexps-more-easily/>
+ */
+
+(function (XRegExp) {
+    "use strict";
+
+    var data = null;
+
+    XRegExp.install("extensibility");
+
+/**
+ * Builds complex regular expressions using named subpatterns for readability and code reuse.
+ * @memberOf XRegExp
+ * @param {String} pattern XRegExp pattern using `{{..}}` for embedded subpatterns.
+ * @param {Object} subs Named subpatterns as strings or regexes. Leading ^ and trailing $ are
+ *   stripped from subpatterns provided as regex objects.
+ * @param {String} [flags] Any combination of XRegExp flags.
+ * @returns {RegExp} Extended regular expression object.
+ * @example
+ *
+ * var color = XRegExp.build("{{keyword}}|{{func}}|{{hex}}", {
+ *   keyword: /^(?:red|tan|[a-z]{4,20})$/,
+ *   func: XRegExp.build("^(?:rgb|hsl)a?\\((?:\\s*{{number}}%?\\s*,?\\s*){3,4}\\)$", {
+ *     number: /^-?\d+(?:\.\d+)?$/
+ *   }),
+ *   hex: /^#(?:[0-9a-f]{1,2}){3}$/
+ * });
+ */
+    XRegExp.build = function (pattern, subs, flags) {
+        var p, regex;
+        data = {};
+        for (p in subs) {
+            data[p] = XRegExp.isRegExp(subs[p]) ? subs[p].source.replace(/^\^|\$$/g, "") : subs[p];
+        }
+        try {
+            regex = XRegExp(pattern, flags);
+        } catch (err) {
+            throw err;
+        } finally {
+            data = null; // Reset
+        }
+        return regex;
+    };
+
+    XRegExp.addToken(
+        /{{([\w$]+)}}/,
+        function (match) {
+            if (data[match[1]] === undefined) {
+                throw new ReferenceError("unknown property: " + match[1]);
+            }
+            return "(?:" + data[match[1]] + ")";
+        },
+        {
+            trigger: function () {
+                return !!data;
+            },
+            scope: "all"
+        }
+    );
+
+}(XRegExp));
+

src/addons/matchrecursive.js

@@ -1,5 +1,5 @@
 /*!
- * XRegExp Match Recursive v0.2.0-beta
+ * XRegExp.matchRecursive v0.2.0-beta
  * Copyright 2009-2012 Steven Levithan <http://xregexp.com/>
  * Available under the MIT License
  */
@@ -12,8 +12,8 @@
  * position data. An error is thrown if delimiters are unbalanced within the data.
  * @memberOf XRegExp
  * @param {String} str String to search.
- * @param {String} left Left delimiter as an XRegExp pattern string.
- * @param {String} right Right delimiter as an XRegExp pattern string.
+ * @param {String} left Left delimiter as an XRegExp pattern.
+ * @param {String} right Right delimiter as an XRegExp pattern.
  * @param {String} [flags] Flags for the left and right delimiters. Use any of: `gimnsxy`.
  * @param {Object} [options] Lets you specify `valueNames` and `escapeChar` options.
  * @returns {Array} Array of matches, or an empty array.

src/addons/unicode/unicode-base.js

@@ -15,8 +15,7 @@
 (function (XRegExp) {
     "use strict";
 
-    var unicode = {},
-        extensible = XRegExp.isInstalled("extensibility");
+    var unicode = {};
 
 /*--------------------------------------
  *  Private helper functions
@@ -83,9 +82,7 @@
  *  Core functionality
  *------------------------------------*/
 
-    if (!extensible) {
-        XRegExp.install("extensibility"); // Temporarily install
-    }
+    XRegExp.install("extensibility");
 
 // Adds Unicode token syntax to XRegExp: \p{..} or \P{..}
     XRegExp.addToken(
@@ -116,9 +113,9 @@
  * @example
  *
  * XRegExp.addUnicodePackage({
- *   xdigit: '0030-00390041-00460061-0066' // 0-9A-Fa-f
+ *   XDigit: '0030-00390041-00460061-0066' // 0-9A-Fa-f
  * }, {
- *   xdigit: 'Hexadecimal'
+ *   XDigit: 'Hexadecimal'
  * });
  */
     XRegExp.addUnicodePackage = function (pack, aliases) {
@@ -149,9 +146,5 @@
         L: "Letter"
     });
 
-    if (!extensible) {
-        XRegExp.uninstall("extensibility"); // Revert to previous state
-    }
-
 }(XRegExp));
 

src/addons/unicode/unicode-blocks.js

@@ -18,11 +18,7 @@
         throw new ReferenceError("Unicode Base must be loaded before Unicode Blocks");
     }
 
-    var extensible = XRegExp.isInstalled("extensibility");
-
-    if (!extensible) {
-        XRegExp.install("extensibility"); // Temporarily install
-    }
+    XRegExp.install("extensibility");
 
     XRegExp.addUnicodePackage({
         InBasic_Latin: "0000-007F",
@@ -183,9 +179,5 @@
         InSpecials: "FFF0-FFFF"
     });
 
-    if (!extensible) {
-        XRegExp.uninstall("extensibility"); // Revert to previous state
-    }
-
 }(XRegExp));
 

src/addons/unicode/unicode-scripts.js

@@ -18,11 +18,7 @@
         throw new ReferenceError("Unicode Base must be loaded before Unicode Scripts");
     }
 
-    var extensible = XRegExp.isInstalled("extensibility");
-
-    if (!extensible) {
-        XRegExp.install("extensibility"); // Temporarily install
-    }
+    XRegExp.install("extensibility");
 
     XRegExp.addUnicodePackage({
         Arabic: "0600-06040606-060B060D-061A061E0620-063F0641-064A0656-065E066A-066F0671-06DC06DE-06FF0750-077F08A008A2-08AC08E4-08FEFB50-FBC1FBD3-FD3DFD50-FD8FFD92-FDC7FDF0-FDFCFE70-FE74FE76-FEFC",
@@ -98,9 +94,5 @@
         Yi: "A000-A48CA490-A4C6"
     });
 
-    if (!extensible) {
-        XRegExp.uninstall("extensibility"); // Revert to previous state
-    }
-
 }(XRegExp));
 

tools/concatenate-source-files.sh

@@ -12,6 +12,7 @@ source_files="
     ../src/addons/unicode/unicode-blocks.js
     ../src/addons/unicode/unicode-properties.js
     ../src/addons/matchrecursive.js
+    ../src/addons/build.js
 "
 
 # Filename of concatenated package