Fixed catastrophic regex backtracking in implementation of the quoted string expressions (dblQuotedString, sglQuotedString, and quotedString)

1 files changed, 5 insertions, 4 deletions

diff --git a/src/pyparsing.py b/src/pyparsing.py
index 5cc4727..fc6a439 100644
--- a/src/pyparsing.py
+++ b/src/pyparsing.py
@@ -58,7 +58,7 @@ The pyparsing module handles some of the problems that are typically vexing when
 """

 

 __version__ = "2.1.2"

-__versionTime__ = "22 Apr 2016 09:25 UTC"

+__versionTime__ = "28 Apr 2016 22:32 UTC"

 __author__ = "Paul McGuire <ptmcg@users.sourceforge.net>"

 

 import string

@@ -3693,9 +3693,10 @@ def infixNotation( baseExpr, opList, lpar=Suppress('('), rpar=Suppress(')') ):
     return ret

 operatorPrecedence = infixNotation

 

-dblQuotedString = Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*"').setName("string enclosed in double quotes")

-sglQuotedString = Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*'").setName("string enclosed in single quotes")

-quotedString = Regex(r'''(?:"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*")|(?:'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*')''').setName("quotedString using single or double quotes")

+dblQuotedString = Combine(Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*')+'"').setName("string enclosed in double quotes")

+sglQuotedString = Combine(Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*")+"'").setName("string enclosed in single quotes")

+quotedString = Combine(Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*')+'"'|

+                       Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*")+"'").setName("quotedString using single or double quotes")

 unicodeString = Combine(_L('u') + quotedString.copy()).setName("unicode string literal")

 

 def nestedExpr(opener="(", closer=")", content=None, ignoreExpr=quotedString.copy()):