From e10042eecd1575d408a3dd32a9d03220d370636f Mon Sep 17 00:00:00 2001
From: Rasmus Jensen <git-email@solsort.dk>
Date: Fri, 12 Apr 2013 11:13:01 +0200
Subject: [PATCH 1/6] test case

---
 js/test/beautify-tests.js | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/js/test/beautify-tests.js b/js/test/beautify-tests.js
index d52147f3..cdbc15a4 100755
--- a/js/test/beautify-tests.js
+++ b/js/test/beautify-tests.js
@@ -78,6 +78,7 @@ function run_beautifier_tests(test_obj, Urlencoded, js_beautify)
         opts.jslint_happy      = false;
         opts.keep_array_indentation = false;
         opts.brace_style       = "collapse";
+        opts.e4x               = true;
 
 
         bt('');
@@ -849,6 +850,7 @@ function run_beautifier_tests(test_obj, Urlencoded, js_beautify)
         opts.max_preserve_newlines = 8;
         bt('a = 1;\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nb = 2;',
             'a = 1;\n\n\n\n\n\n\n\nb = 2;');
+        test_fragment('xml=<a b="c"><d/><e>\n foo</e>x</a>;', 'xml = <a b="c"><d/><e>\n foo</e>x</a>;');
 
         Urlencoded.run_tests(sanitytest);
 

From 0036967763635c546d7066041306590daeb87052 Mon Sep 17 00:00:00 2001
From: Rasmus Jensen <git-email@solsort.dk>
Date: Fri, 12 Apr 2013 11:21:44 +0200
Subject: [PATCH 2/6] option for ignoring e4x literals

---
 js/lib/beautify.js | 30 ++++++++++++++++++++++++++++++
 js/lib/cli.js      |  3 +++
 2 files changed, 33 insertions(+)

diff --git a/js/lib/beautify.js b/js/lib/beautify.js
index c5c2d53d..3dc89f4c 100644
--- a/js/lib/beautify.js
+++ b/js/lib/beautify.js
@@ -812,6 +812,36 @@
                 }
             }
 
+            if (options.e4x && c === '<') {
+                // pass e4x-literals untouched through the pretty-printing
+                var xmlRegExp = /<(\/?)([a-zA-Z:0-9]+)\s*([a-zA-Z:0-9]+="[^"]*"\s*)*(\/?)\s*>/g;
+                var xmlStr = input.slice(parser_pos - 1);
+                var match = xmlRegExp.exec(xmlStr);
+                if (match && match.index === 0) {
+                    var rootTag = match[2];
+                    var depth = 0;
+                    while (match) {
+                        var isEndTag = !! match[1];
+                        var tagName = match[2];
+                        var isSingletonTag = !! match[match.length - 1];
+                        if (tagName === rootTag && !isSingletonTag) {
+                            if (isEndTag) {
+                                --depth;
+                            } else {
+                                ++depth;
+                            }
+                        }
+                        if (depth <= 0) {
+                            break;
+                        }
+                        match = xmlRegExp.exec(xmlStr);
+                    }
+                    var xmlLength = match.index + match[0].length;
+                    parser_pos += xmlLength - 1;
+                    return [xmlStr.slice(0, xmlLength), "TK_WORD"];
+                };
+            }
+
             if (c === '<' && input.substring(parser_pos - 1, parser_pos + 3) === '<!--') {
                 parser_pos += 3;
                 c = '<!--';
diff --git a/js/lib/cli.js b/js/lib/cli.js
index 19dfad34..82e0c430 100755
--- a/js/lib/cli.js
+++ b/js/lib/cli.js
@@ -55,6 +55,7 @@ var fs = require('fs'),
         "keep_array_indentation": Boolean,
         "unescape_strings": Boolean,
         "wrap_line_length": Number,
+        "e4x": Boolean,
         // HTML-only
         "max_char": Number,
         "unformatted": [String, Array],
@@ -85,6 +86,7 @@ var fs = require('fs'),
         "k": ["--keep_array_indentation"],
         "x": ["--unescape_strings"],
         "w": ["--wrap_line_length"],
+        "X": ["--e4x"],
         // HTML-only
         "W": ["--max_char"],
         "U": ["--unformatted"],
@@ -186,6 +188,7 @@ function usage(err) {
         msg.push('  -k, --keep-array-indentation  Preserve array indentation');
         msg.push('  -x, --unescape-strings        Decode printable characters encoded in xNN notation');
         msg.push('  -w, --wrap-line-length        Wrap lines at next opportunity after N characters [0]');
+        msg.push('  -X, --e4x                     Pass E4X xml literals through untouched');
         msg.push('  --good-stuff                  Warm the cockles of Crockford\'s heart');
         break;
     case "html":

From aab53a7eb2484cc79d63dc2e1f544bdceec9292a Mon Sep 17 00:00:00 2001
From: Rasmus Jensen <git-email@solsort.dk>
Date: Thu, 18 Apr 2013 08:18:54 +0200
Subject: [PATCH 3/6] updated tests, such that e4x is off by default, and
 xml-literal is tested both with and without e4x option enabled

---
 js/test/beautify-tests.js | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/js/test/beautify-tests.js b/js/test/beautify-tests.js
index cdbc15a4..8874d932 100755
--- a/js/test/beautify-tests.js
+++ b/js/test/beautify-tests.js
@@ -78,7 +78,6 @@ function run_beautifier_tests(test_obj, Urlencoded, js_beautify)
         opts.jslint_happy      = false;
         opts.keep_array_indentation = false;
         opts.brace_style       = "collapse";
-        opts.e4x               = true;
 
 
         bt('');
@@ -850,7 +849,13 @@ function run_beautifier_tests(test_obj, Urlencoded, js_beautify)
         opts.max_preserve_newlines = 8;
         bt('a = 1;\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nb = 2;',
             'a = 1;\n\n\n\n\n\n\n\nb = 2;');
+
+        // Test that e4x literals passed through when e4x-option is enabled
+        test_fragment('xml=<a b="c"><d/><e>\n foo</e>x</a>;', 'xml = < a b = "c" > < d / > < e >\n    foo < /e>x</a > ;');
+        opts.e4x = true;
         test_fragment('xml=<a b="c"><d/><e>\n foo</e>x</a>;', 'xml = <a b="c"><d/><e>\n foo</e>x</a>;');
+        opts.e4x = false;
+
 
         Urlencoded.run_tests(sanitytest);
 

From b90b6b9bac762680e0c360f7437d875e4248bda8 Mon Sep 17 00:00:00 2001
From: Rasmus Jensen <git-email@solsort.dk>
Date: Fri, 19 Apr 2013 14:27:14 +0200
Subject: [PATCH 4/6] move xml-literal handling into TK_STRING section

---
 js/lib/beautify.js | 70 ++++++++++++++++++++++++++--------------------
 1 file changed, 39 insertions(+), 31 deletions(-)

diff --git a/js/lib/beautify.js b/js/lib/beautify.js
index 3dc89f4c..14468c7e 100644
--- a/js/lib/beautify.js
+++ b/js/lib/beautify.js
@@ -698,7 +698,14 @@
                 (last_type === 'TK_END_EXPR' && in_array(previous_flags.mode, [MODE.Conditional, MODE.ForInitializer])) ||
                 (in_array(last_type, ['TK_COMMENT', 'TK_START_EXPR', 'TK_START_BLOCK',
                     'TK_END_BLOCK', 'TK_OPERATOR', 'TK_EQUALS', 'TK_EOF', 'TK_SEMICOLON', 'TK_COMMA'
-            ]))))) { // regexp
+            ])))) || // regexp
+            (options.e4x && c ==="<" &&
+                ((last_type === 'TK_WORD' && is_special_word (flags.last_text)) ||
+                (last_type === 'TK_END_EXPR' && in_array(previous_flags.mode, [MODE.Conditional, MODE.ForInitializer])) ||
+                (in_array(last_type, ['TK_COMMENT', 'TK_START_EXPR', 'TK_START_BLOCK',
+                    'TK_END_BLOCK', 'TK_OPERATOR', 'TK_EQUALS', 'TK_EOF', 'TK_SEMICOLON', 'TK_COMMA'
+            ]))) && input.slice(parser_pos - 1).match(/^<[a-zA-Z:0-9]+\s*([a-zA-Z:0-9]+="[^"]*"\s*)*\/?\s*>/)
+            )) { // xml literal
                 var sep = c,
                     esc = false,
                     has_char_escapes = false;
@@ -731,6 +738,37 @@
                             }
                         }
 
+                    } else if (options.e4x && sep === '<') {
+                        //
+                        // handle e4x xml literals separately
+                        //
+                        var xmlRegExp = /<(\/?)([a-zA-Z:0-9]+)\s*([a-zA-Z:0-9]+="[^"]*"\s*)*(\/?)\s*>/g;
+                        var xmlStr = input.slice(parser_pos - 1);
+                        var match = xmlRegExp.exec(xmlStr);
+                        if (match && match.index === 0) {
+                            var rootTag = match[2];
+                            var depth = 0;
+                            while (match) {
+                                var isEndTag = !! match[1];
+                                var tagName = match[2];
+                                var isSingletonTag = !! match[match.length - 1];
+                                if (tagName === rootTag && !isSingletonTag) {
+                                    if (isEndTag) {
+                                        --depth;
+                                    } else {
+                                        ++depth;
+                                    }
+                                }
+                                if (depth <= 0) {
+                                    break;
+                                }
+                                match = xmlRegExp.exec(xmlStr);
+                            }
+                            var xmlLength = match.index + match[0].length;
+                            parser_pos += xmlLength - 1;
+                            return [xmlStr.slice(0, xmlLength), "TK_STRING"];
+                        };
+        
                     } else {
                         //
                         // and handle string also separately
@@ -812,36 +850,6 @@
                 }
             }
 
-            if (options.e4x && c === '<') {
-                // pass e4x-literals untouched through the pretty-printing
-                var xmlRegExp = /<(\/?)([a-zA-Z:0-9]+)\s*([a-zA-Z:0-9]+="[^"]*"\s*)*(\/?)\s*>/g;
-                var xmlStr = input.slice(parser_pos - 1);
-                var match = xmlRegExp.exec(xmlStr);
-                if (match && match.index === 0) {
-                    var rootTag = match[2];
-                    var depth = 0;
-                    while (match) {
-                        var isEndTag = !! match[1];
-                        var tagName = match[2];
-                        var isSingletonTag = !! match[match.length - 1];
-                        if (tagName === rootTag && !isSingletonTag) {
-                            if (isEndTag) {
-                                --depth;
-                            } else {
-                                ++depth;
-                            }
-                        }
-                        if (depth <= 0) {
-                            break;
-                        }
-                        match = xmlRegExp.exec(xmlStr);
-                    }
-                    var xmlLength = match.index + match[0].length;
-                    parser_pos += xmlLength - 1;
-                    return [xmlStr.slice(0, xmlLength), "TK_WORD"];
-                };
-            }
-
             if (c === '<' && input.substring(parser_pos - 1, parser_pos + 3) === '<!--') {
                 parser_pos += 3;
                 c = '<!--';

From 7dbd03a2d33606ace494978d70f5903d2a10428d Mon Sep 17 00:00:00 2001
From: Rasmus Jensen <git-email@solsort.dk>
Date: Fri, 19 Apr 2013 16:27:35 +0200
Subject: [PATCH 5/6] more tests + robustness fix for unterminated xml literal

---
 js/lib/beautify.js        |  2 +-
 js/test/beautify-tests.js | 12 ++++++++++--
 2 files changed, 11 insertions(+), 3 deletions(-)

diff --git a/js/lib/beautify.js b/js/lib/beautify.js
index 14468c7e..ec4f186c 100644
--- a/js/lib/beautify.js
+++ b/js/lib/beautify.js
@@ -764,7 +764,7 @@
                                 }
                                 match = xmlRegExp.exec(xmlStr);
                             }
-                            var xmlLength = match.index + match[0].length;
+                            var xmlLength = match ? match.index + match[0].length : xmlStr.length;
                             parser_pos += xmlLength - 1;
                             return [xmlStr.slice(0, xmlLength), "TK_STRING"];
                         };
diff --git a/js/test/beautify-tests.js b/js/test/beautify-tests.js
index 8874d932..e631fed6 100755
--- a/js/test/beautify-tests.js
+++ b/js/test/beautify-tests.js
@@ -851,9 +851,17 @@ function run_beautifier_tests(test_obj, Urlencoded, js_beautify)
             'a = 1;\n\n\n\n\n\n\n\nb = 2;');
 
         // Test that e4x literals passed through when e4x-option is enabled
-        test_fragment('xml=<a b="c"><d/><e>\n foo</e>x</a>;', 'xml = < a b = "c" > < d / > < e >\n    foo < /e>x</a > ;');
+        bt('xml=<a b="c"><d/><e>\n foo</e>x</a>;', 'xml = < a b = "c" > < d / > < e >\n    foo < /e>x</a > ;');
         opts.e4x = true;
-        test_fragment('xml=<a b="c"><d/><e>\n foo</e>x</a>;', 'xml = <a b="c"><d/><e>\n foo</e>x</a>;');
+        bt('xml=<a b="c"><d/><e>\n foo</e>x</a>;', 'xml = <a b="c"><d/><e>\n foo</e>x</a>;');
+        // Handles messed up tags, as long as it isn't the same name
+        // as the root tag. Also handles tags of same name as root tag
+        // as long as nesting matches.
+        bt('xml=<a x="jn"><c></b></f><a><d jnj="jnn"><f></a ></nj></a>;', 
+         'xml = <a x="jn"><c></b></f><a><d jnj="jnn"><f></a ></nj></a>;');
+        // If xml is not terminated, the remainder of the file is treated
+        // as part of the xml-literal (passed through unaltered)
+        test_fragment('xml=<a></b>\nc<b;', 'xml = <a></b>\nc<b;'); 
         opts.e4x = false;
 
 

From aad87666ca45284214d084a1e05f71183516f6fd Mon Sep 17 00:00:00 2001
From: Liam Newman <bitwiseman@gmail.com>
Date: Fri, 19 Apr 2013 14:46:43 -0700
Subject: [PATCH 6/6] Make e4x pay-for-play - no effect when option not enabled

---
 js/lib/beautify.js | 34 +++++++++++++++-------------------
 1 file changed, 15 insertions(+), 19 deletions(-)

diff --git a/js/lib/beautify.js b/js/lib/beautify.js
index ec4f186c..aa33c76e 100644
--- a/js/lib/beautify.js
+++ b/js/lib/beautify.js
@@ -692,20 +692,18 @@
 
             }
 
+
             if (c === "'" || c === '"' || // string
-            (c === '/' &&
-                ((last_type === 'TK_WORD' && is_special_word (flags.last_text)) ||
-                (last_type === 'TK_END_EXPR' && in_array(previous_flags.mode, [MODE.Conditional, MODE.ForInitializer])) ||
-                (in_array(last_type, ['TK_COMMENT', 'TK_START_EXPR', 'TK_START_BLOCK',
-                    'TK_END_BLOCK', 'TK_OPERATOR', 'TK_EQUALS', 'TK_EOF', 'TK_SEMICOLON', 'TK_COMMA'
-            ])))) || // regexp
-            (options.e4x && c ==="<" &&
-                ((last_type === 'TK_WORD' && is_special_word (flags.last_text)) ||
-                (last_type === 'TK_END_EXPR' && in_array(previous_flags.mode, [MODE.Conditional, MODE.ForInitializer])) ||
-                (in_array(last_type, ['TK_COMMENT', 'TK_START_EXPR', 'TK_START_BLOCK',
-                    'TK_END_BLOCK', 'TK_OPERATOR', 'TK_EQUALS', 'TK_EOF', 'TK_SEMICOLON', 'TK_COMMA'
-            ]))) && input.slice(parser_pos - 1).match(/^<[a-zA-Z:0-9]+\s*([a-zA-Z:0-9]+="[^"]*"\s*)*\/?\s*>/)
-            )) { // xml literal
+                (
+                    (c === '/') || // regexp
+                    (options.e4x && c ==="<" && input.slice(parser_pos - 1).match(/^<[a-zA-Z:0-9]+\s*([a-zA-Z:0-9]+="[^"]*"\s*)*\/?\s*>/)) // xml
+                ) && ( // regex and xml can only appear in specific locations during parsing
+                    (last_type === 'TK_WORD' && is_special_word (flags.last_text)) ||
+                    (last_type === 'TK_END_EXPR' && in_array(previous_flags.mode, [MODE.Conditional, MODE.ForInitializer])) ||
+                    (in_array(last_type, ['TK_COMMENT', 'TK_START_EXPR', 'TK_START_BLOCK',
+                    'TK_END_BLOCK', 'TK_OPERATOR', 'TK_EQUALS', 'TK_EOF', 'TK_SEMICOLON', 'TK_COMMA']))
+                )) {
+
                 var sep = c,
                     esc = false,
                     has_char_escapes = false;
@@ -715,7 +713,7 @@
                 if (parser_pos < input_length) {
                     if (sep === '/') {
                         //
-                        // handle regexp separately...
+                        // handle regexp
                         //
                         var in_char_class = false;
                         while (esc || in_char_class || input.charAt(parser_pos) !== sep) {
@@ -737,10 +735,9 @@
                                 return [resulting_string, 'TK_STRING'];
                             }
                         }
-
                     } else if (options.e4x && sep === '<') {
                         //
-                        // handle e4x xml literals separately
+                        // handle e4x xml literals
                         //
                         var xmlRegExp = /<(\/?)([a-zA-Z:0-9]+)\s*([a-zA-Z:0-9]+="[^"]*"\s*)*(\/?)\s*>/g;
                         var xmlStr = input.slice(parser_pos - 1);
@@ -767,11 +764,10 @@
                             var xmlLength = match ? match.index + match[0].length : xmlStr.length;
                             parser_pos += xmlLength - 1;
                             return [xmlStr.slice(0, xmlLength), "TK_STRING"];
-                        };
-        
+                        }
                     } else {
                         //
-                        // and handle string also separately
+                        // handle string
                         //
                         while (esc || input.charAt(parser_pos) !== sep) {
                             resulting_string += input.charAt(parser_pos);