[Pkg-javascript-commits] [uglifyjs] 78/491: improve `RegExp` handling (#1959)

[Jonas Smedegaard]

This is an automated email from the git hooks/post-receive script.

js pushed a commit to annotated tag debian/3.3.10-1
in repository uglifyjs.

commit 569c21e952c99340db73d6ad1c7102500cc1b341
Author: Alex Lam S.L <alexlamsl at gmail.com>
Date:   Wed May 17 20:10:50 2017 +0800

    improve `RegExp` handling (#1959)
    
    - remove `options.output.unescape_regexps`
    - preserve original pattern whenever possible
    
    fixes #54
    fixes #1929
---
 lib/output.js                   | 42 +++++------------------------------------
 lib/parse.js                    | 16 +++++++++-------
 test/compress/string-literal.js |  9 +++++++++
 3 files changed, 23 insertions(+), 44 deletions(-)

diff --git a/lib/output.js b/lib/output.js
index b4d31b5..4080055 100644
--- a/lib/output.js
+++ b/lib/output.js
@@ -70,7 +70,6 @@ function OutputStream(options) {
         semicolons       : true,
         shebang          : true,
         source_map       : null,
-        unescape_regexps : false,
         width            : 80,
         wrap_iife        : false,
     }, true);
@@ -1260,45 +1259,14 @@ function OutputStream(options) {
         }
     });
 
-    function regexp_safe_literal(code) {
-        return [
-            0x5c   , // \
-            0x2f   , // /
-            0x2e   , // .
-            0x2b   , // +
-            0x2a   , // *
-            0x3f   , // ?
-            0x28   , // (
-            0x29   , // )
-            0x5b   , // [
-            0x5d   , // ]
-            0x7b   , // {
-            0x7d   , // }
-            0x24   , // $
-            0x5e   , // ^
-            0x3a   , // :
-            0x7c   , // |
-            0x21   , // !
-            0x0a   , // \n
-            0x0d   , // \r
-            0x00   , // \0
-            0xfeff , // Unicode BOM
-            0x2028 , // unicode "line separator"
-            0x2029 , // unicode "paragraph separator"
-        ].indexOf(code) < 0;
-    };
-
     DEFPRINT(AST_RegExp, function(self, output){
-        var str = self.getValue().toString();
+        var regexp = self.getValue();
+        var str = regexp.toString();
+        if (regexp.raw_source) {
+            str = "/" + regexp.raw_source + str.slice(str.lastIndexOf("/"));
+        }
         if (output.option("ascii_only")) {
             str = output.to_ascii(str);
-        } else if (output.option("unescape_regexps")) {
-            str = str.split("\\\\").map(function(str){
-                return str.replace(/\\u[0-9a-fA-F]{4}|\\x[0-9a-fA-F]{2}/g, function(s){
-                    var code = parseInt(s.substr(2), 16);
-                    return regexp_safe_literal(code) ? String.fromCharCode(code) : s;
-                });
-            }).join("\\\\");
         }
         output.print(str);
         var p = output.parent();
diff --git a/lib/parse.js b/lib/parse.js
index 97dd6d4..c960b49 100644
--- a/lib/parse.js
+++ b/lib/parse.js
@@ -477,31 +477,33 @@ function tokenizer($TEXT, filename, html5_comments, shebang) {
         return name;
     };
 
-    var read_regexp = with_eof_error("Unterminated regular expression", function(regexp){
+    var read_regexp = with_eof_error("Unterminated regular expression", function(source) {
         var prev_backslash = false, ch, in_class = false;
         while ((ch = next(true))) if (NEWLINE_CHARS(ch)) {
             parse_error("Unexpected line terminator");
         } else if (prev_backslash) {
-            regexp += "\\" + ch;
+            source += "\\" + ch;
             prev_backslash = false;
         } else if (ch == "[") {
             in_class = true;
-            regexp += ch;
+            source += ch;
         } else if (ch == "]" && in_class) {
             in_class = false;
-            regexp += ch;
+            source += ch;
         } else if (ch == "/" && !in_class) {
             break;
         } else if (ch == "\\") {
             prev_backslash = true;
         } else {
-            regexp += ch;
+            source += ch;
         }
         var mods = read_name();
         try {
-          return token("regexp", new RegExp(regexp, mods));
+            var regexp = new RegExp(source, mods);
+            regexp.raw_source = source;
+            return token("regexp", regexp);
         } catch(e) {
-          parse_error(e.message);
+            parse_error(e.message);
         }
     });
 
diff --git a/test/compress/string-literal.js b/test/compress/string-literal.js
index 8b93961..1138fed 100644
--- a/test/compress/string-literal.js
+++ b/test/compress/string-literal.js
@@ -8,3 +8,12 @@ octal_escape_sequence: {
         var border_check = "\x20\x30\x38\x30\x00\x30\xc0\x30";
     }
 }
+
+issue_1929: {
+    input: {
+        function f(s) {
+            return s.split(/[\\/]/);
+        }
+    }
+    expect_exact: "function f(s){return s.split(/[\\\\/]/)}"
+}

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/pkg-javascript/uglifyjs.git