zorba-coders team mailing list archive

["Paul J. Lucas" <paul@xxxxxxxxxxxxx>]

Paul J. Lucas has proposed merging lp:~paul-lucas/zorba/bug-855715 into lp:zorba.

Requested reviews:
  Paul J. Lucas (paul-lucas)
  Matthias Brantner (matthias-brantner)

For more details, see:
https://code.launchpad.net/~paul-lucas/zorba/bug-855715/+merge/79768

Now checking for invalid regex escape sequences.
-- 
https://code.launchpad.net/~paul-lucas/zorba/bug-855715/+merge/79768
Your team Zorba Coders is subscribed to branch lp:zorba.

=== modified file 'ChangeLog'
--- ChangeLog	2011-10-12 21:30:46 +0000
+++ ChangeLog	2011-10-19 05:12:24 +0000
@@ -52,6 +52,7 @@
   * Added undo for node revalidation
   * Fixed bug #872796  (validate-in-place can interfere with other update primitives)
   * Fixed bug #872799 (validate-in-place can set incorrect types)
+  * Fixed bug #855715 (Invalid escaped characters in regex not caught)
 
 version 2.0.1
 

=== modified file 'modules/com/zorba-xquery/www/modules/xqdoc2xhtml/index.xq'
--- modules/com/zorba-xquery/www/modules/xqdoc2xhtml/index.xq	2011-10-07 08:28:43 +0000
+++ modules/com/zorba-xquery/www/modules/xqdoc2xhtml/index.xq	2011-10-19 05:12:24 +0000
@@ -1364,7 +1364,7 @@
                                fn:concat(fn:substring-before($description,"."),".") else ""
       order by $name, $param-number
       return
-        let $type := replace(normalize-space(substring-after(substring-before($signature, "function"), "declare")),"\%",""),
+        let $type := replace(normalize-space(substring-after(substring-before($signature, "function"), "declare")),"%",""),
             $isExternal := ends-with($signature, "external"),
             $paramsAndReturn := substring-after($signature,concat(':',$name)),
             $external := if(ends-with($signature,"external")) then "external" else ""

=== modified file 'modules/org/expath/ns/file.xq'
--- modules/org/expath/ns/file.xq	2011-10-17 11:49:38 +0000
+++ modules/org/expath/ns/file.xq	2011-10-19 05:12:24 +0000
@@ -624,7 +624,7 @@
 declare function file:glob-to-regex(
   $pattern as xs:string
 ) {
-  let $pattern := fn:replace($pattern, '(\.|\[|\]|\\|\/|\||\-|\^|\$|\?|\*|\+|\{|\}|\(|\))','\\$1')
+  let $pattern := fn:replace($pattern, '(\.|\[|\]|\\|/|\||\-|\^|\$|\?|\*|\+|\{|\}|\(|\))','\\$1')
   let $pattern := fn:replace($pattern, '\\\?', '.')
   let $pattern := fn:replace($pattern, '\\\*', '.*')
   return

=== modified file 'src/diagnostics/diagnostic_en.xml'
--- src/diagnostics/diagnostic_en.xml	2011-10-05 18:52:55 +0000
+++ src/diagnostics/diagnostic_en.xml	2011-10-19 05:12:24 +0000
@@ -2393,6 +2393,10 @@
       <value>invalid library module</value>
     </entry>
 
+    <entry key="BadRegexEscape_3">
+      <value>"$3": illegal escape character</value>
+    </entry>
+
     <entry key="BadPath">
       <value>invalid path</value>
     </entry>

=== modified file 'src/diagnostics/pregenerated/dict_en.cpp'
--- src/diagnostics/pregenerated/dict_en.cpp	2011-10-05 17:49:48 +0000
+++ src/diagnostics/pregenerated/dict_en.cpp	2011-10-19 05:12:24 +0000
@@ -424,6 +424,7 @@
   { "~BadIterator", "invalid iterator" },
   { "~BadLibraryModule", "invalid library module" },
   { "~BadPath", "invalid path" },
+  { "~BadRegexEscape_3", "\"$3\": illegal escape character" },
   { "~BadStreamState", "bad I/O stream state" },
   { "~BadTokenInBraces_3", "\"$3\": illegal token within { }" },
   { "~BadTraceStream", "trace stream not retrievable using SerializationCallback" },

=== modified file 'src/util/regex.cpp'
--- src/util/regex.cpp	2011-07-18 14:25:21 +0000
+++ src/util/regex.cpp	2011-10-19 05:12:24 +0000
@@ -128,33 +128,69 @@
         case 'c': // NameChar
           *icu_re += "[" bs_c "]";
           continue;
-        case 'C': // ^\c
+        case 'C': // [^\c]
           *icu_re += "[^" bs_c "]";
           continue;
         case 'i': // initial NameChar
           *icu_re += "[" bs_i "]";
           continue;
-        case 'I': // ^\i
+        case 'I': // [^\i]
           *icu_re += "[^" bs_i "]";
           continue;
-        default:
-          if ( ascii::is_digit( *xq_c ) ) {
-            backref_no = *xq_c - '0';
-            if ( !backref_no )          // \0 is illegal
-              throw INVALID_RE_EXCEPTION( xq_re, ZED( BackRef0Illegal ) );
-            if ( in_char_class ) {
-              //
-              // XQuery 3.0 F&O 5.6.1: Within a character class expression,
-              // \ followed by a digit is invalid.
-              //
-              throw INVALID_RE_EXCEPTION(
-                xq_re, ZED( BackRefIllegalInCharClass )
-              );
-            }
-            in_backref = true;
+        case '0':
+        case '1':
+        case '2':
+        case '3':
+        case '4':
+        case '5':
+        case '6':
+        case '7':
+        case '8':
+        case '9':
+          backref_no = *xq_c - '0';
+          if ( !backref_no )          // \0 is illegal
+            throw INVALID_RE_EXCEPTION( xq_re, ZED( BackRef0Illegal ) );
+          if ( in_char_class ) {
+            //
+            // XQuery 3.0 F&O 5.6.1: Within a character class expression,
+            // \ followed by a digit is invalid.
+            //
+            throw INVALID_RE_EXCEPTION(
+              xq_re, ZED( BackRefIllegalInCharClass )
+            );
           }
+          in_backref = true;
+          // no break;
+        case '$':
+        case '(':
+        case ')':
+        case '*':
+        case '+':
+        case '-':
+        case '.':
+        case '?':
+        case 'd': // [0-9]
+        case 'D': // [^\d]
+        case 'n': // newline
+        case 'p': // category escape
+        case 'P': // [^\p]
+        case 'r': // carriage return
+        case 's': // whitespace
+        case 'S': // [^\s]
+        case 't': // tab
+        case 'w': // word char
+        case 'W': // [^\w]
+        case '[':
+        case '\\':
+        case ']':
+        case '^':
+        case '{':
+        case '|':
+        case '}':
           *icu_re += '\\';
           break;
+        default:
+          throw INVALID_RE_EXCEPTION( xq_re, ZED( BadRegexEscape_3 ), *xq_c );
       }
     } else {
       if ( in_backref ) {

=== modified file 'test/rbkt/Queries/zorba/file/files_pattern_rec1.xq'
--- test/rbkt/Queries/zorba/file/files_pattern_rec1.xq	2011-04-29 22:29:04 +0000
+++ test/rbkt/Queries/zorba/file/files_pattern_rec1.xq	2011-10-19 05:12:24 +0000
@@ -2,9 +2,15 @@
 
 declare variable $path as xs:string external;
 
+let $sep := file:directory-separator()
+let $escaped_sep :=
+  if ($sep eq "\") then
+    fn:concat("\", $sep)
+  else
+    $sep
 let $files := file:list($path, fn:true(), "*dir*")
 for $file in $files
 where fn:not(fn:matches($file, "\.svn"))
-let $f := fn:replace($file, fn:concat("\", file:directory-separator()), "/")
+let $f := fn:replace($file, $escaped_sep, "/")
 order by $f
 return <result>{$f}</result>

=== modified file 'test/rbkt/Queries/zorba/file/files_pattern_rec2.xq'
--- test/rbkt/Queries/zorba/file/files_pattern_rec2.xq	2011-04-29 22:29:04 +0000
+++ test/rbkt/Queries/zorba/file/files_pattern_rec2.xq	2011-10-19 05:12:24 +0000
@@ -2,9 +2,15 @@
 
 declare variable $path as xs:string external;
 
+let $sep := file:directory-separator()
+let $escaped_sep :=
+  if ($sep eq "\") then
+    fn:concat("\", $sep)
+  else
+    $sep
 let $files := file:list($path, fn:true(), "*.txt")
 for $file in $files
 where fn:not(fn:matches($file, "\.svn"))
-let $f := fn:replace($file, fn:concat("\", file:directory-separator()), "/")
+let $f := fn:replace($file, $escaped_sep, "/")
 order by $f
 return <result>{$f}</result>