From c8a2a786eb3f1cda8780b348f27e2c8240621045 Mon Sep 17 00:00:00 2001
From: "Jay Freeman (saurik)" <saurik@saurik.com>
Date: Tue, 1 Dec 2015 23:32:39 -0800
Subject: [PATCH] Use start conditions to parse regular expressions.

---
 Cycript.l.in | 36 +++++++++++++++++++++++++-----------
 Parser.hpp   |  6 ++++--
 2 files changed, 29 insertions(+), 13 deletions(-)

diff --git a/Cycript.l.in b/Cycript.l.in
index 55502d4..114b19a 100644
--- a/Cycript.l.in
+++ b/Cycript.l.in
@@ -220,6 +220,9 @@ WhiteSpace [\x09\x0b\x0c\x20]|\xc2\xa0|\xef\xbb\xbf
 UnicodeEscape \\u({HexDigit}{4}|\{{HexDigit}+\})
 
 @include NotLineTerminator.l
+NoneTerminatorCharacter [^\r\n\x80-\xff]|{NotLineTerminator}
+RegExCharacter [^/[\\]{-}[\r\n\x80-\xff]|{NotLineTerminator}
+RegClsCharacter [^]\\]{-}[\r\n\x80-\xff]|{NotLineTerminator}
 CommentCharacter [^*/]{-}[\r\n\x80-\xff]|{NotLineTerminator}
 SingleCharacter [^'\\]{-}[\r\n\x80-\xff]|{NotLineTerminator}
 DoubleCharacter [^"\\]{-}[\r\n\x80-\xff]|{NotLineTerminator}
@@ -239,13 +242,8 @@ IdentifierPart {UnicodePart}|{UnicodeEscape}
 IdentifierFail {UnicodeFail}|\\(u({HexDigit}{0,3}|\{{HexDigit}*))?
 IdentifierScrap {IdentifierPart}*{IdentifierFail}?
 
-NonTerminator [^\n]
-BackslashSequence \\{NonTerminator}
-RegularExpressionFirstChar [^\n*\\/]|{BackslashSequence}
-RegularExpressionChar [^\n\\/]|{BackslashSequence}
-RegularExpressionFlags {UnicodePart}*
-RegularExpressionChars {RegularExpressionChar}*
-RegularExpressionBody {RegularExpressionFirstChar}{RegularExpressionChars}
+RegularExpressionBackslashSequence \\{NoneTerminatorCharacter}
+RegularExpressionClassChars ({RegClsCharacter}|{RegularExpressionBackslashSequence})*
 
 @begin E4X
 XMLNameStart [a-zA-Z_:]
@@ -253,6 +251,7 @@ XMLNamePart [a-zA-Z0-9.-_:]
 XMLName {XMLNameStart}{XMLNamePart}*
 @end
 
+%x RegularExpression
 %x MultiLine
 
 %x LegacySingleString
@@ -275,10 +274,25 @@ XMLName {XMLNameStart}{XMLNamePart}*
 %%
 
     /* RegEx {{{ */
-<RegExp,RegExpOrTemplateTail>{
-    \/{RegularExpressionBody}\/{RegularExpressionFlags} L I(literal, RegEx(Y), tk::RegularExpressionLiteral, hi::Constant);
-    \/{RegularExpressionBody}\/{RegularExpressionFlags}{UnicodeFail} L E("invalid flags")
-    \/{RegularExpressionBody}?\\? L E("unterminated regex")
+<RegExp,RegExpOrTemplateTail>\/ L CYLexBufferStart(RegularExpression); CYLexBufferUnit('/');
+
+<RegularExpression>{
+    \/{UnicodePart}* R CYLexBufferUnits(yytext, yyleng); CYLexBufferEnd(literal, RegEx, tk::RegularExpressionLiteral, hi::Constant);
+    \/{UnicodePart}*{UnicodeFail} R E("invalid flags")
+
+    {RegExCharacter}+ R CYLexBufferUnits(yytext, yyleng);
+    {RegExCharacter}*{UnicodeFail} R E("invalid character");
+
+    {RegularExpressionBackslashSequence} R CYLexBufferUnits(yytext, yyleng);
+    \\{UnicodeFail}? R E("invalid escape")
+
+    "["{RegularExpressionClassChars}"]" R CYLexBufferUnits(yytext, yyleng);
+    "["{RegularExpressionClassChars}\\? R E("invalid class");
+    "["{RegularExpressionClassChars}\\?{UnicodeFail} R E("invalid character");
+    "["{RegularExpressionClassChars}\\?{LineTerminatorSequence} R E("invalid newline");
+
+    (\\|{RegExCharacter}+)?{LineTerminatorSequence} R E("invalid newline");
+    <<EOF>> R E("unterminated regex")
 }
     /* }}} */
     /* Comment {{{ */
diff --git a/Parser.hpp b/Parser.hpp
index 1ea1451..02b08eb 100644
--- a/Parser.hpp
+++ b/Parser.hpp
@@ -860,9 +860,11 @@ struct CYRegEx :
     CYTrivial
 {
     const char *value_;
+    size_t size_;
 
-    CYRegEx(const char *value) :
-        value_(value)
+    CYRegEx(const char *value, size_t size) :
+        value_(value),
+        size_(size)
     {
     }
 
-- 
2.47.2