122047 2006-02-15 18:00:14 +0000 ∉ is displayed as ¬in; 2006-11-01 19:09:12 +0000 1 1 2 Applications konqueror khtml parsing unspecified Gentoo Packages Linux RESOLVED FIXED NOR normal --- 1 hkBst konqueror-bugs-null 0 oldest_to_newest 414454 0 hkBst 2006-02-15 18:00:14 +0000 Version: 3.5.1 (using KDE KDE 3.5.1) Installed from: Gentoo Packages Compiler: gcc-3.4.5-vanilla OS: Linux The ∉ html entity is displayed as ¬in; 481794 1 18337 kde 2006-10-31 18:03:21 +0000 Created attachment 18337 test case test case. The problem is we support a IE quirk where we parse the beginning of a entity 482168 2 kde 2006-11-01 19:09:11 +0000 SVN commit 601019 by carewolf: Dont let a IE quirk get in the way of parsing longer entities. BUG: 122047 M +19 -7 htmltokenizer.cpp M +3 -2 htmltokenizer.h --- branches/KDE/3.5/kdelibs/khtml/html/htmltokenizer.cpp #601018:601019 @@ -395,7 +395,7 @@ CachedScript* cs = 0; // forget what we just got, load from src url instead - if ( !currentScriptSrc.isEmpty() && javascript && + if ( !currentScriptSrc.isEmpty() && javascript && (cs = parser->doc()->docLoader()->requestScript(currentScriptSrc, scriptSrcCharset) )) { cachedScript.enqueue(cs); } @@ -601,6 +601,7 @@ if( start ) { cBufferPos = 0; + entityLen = 0; Entity = SearchEntity; } @@ -696,8 +697,8 @@ if ( tag == NoTag ) { const entity* e = kde_findEntity(cBuffer, cBufferPos); if ( e && e->code < 256 ) { - Entity = SearchSemicolon; - break; + EntityChar = e->code; + entityLen = cBufferPos; } } } @@ -705,8 +706,10 @@ if(Entity == SearchSemicolon) { if(cBufferPos > 1) { const entity *e = kde_findEntity(cBuffer, cBufferPos); - if(e && ( e->code < 256 || *src == ';' )) + if(e && ( e->code < 256 || *src == ';' )) { EntityChar = e->code; + entityLen = cBufferPos; + } } } break; @@ -723,7 +726,17 @@ if ( !EntityChar.isNull() ) { checkBuffer(); // Just insert it - src.push( EntityChar ); + *dest++ = EntityChar; + if (entityLen > 0 && entityLen < cBufferPos) { + int rem = cBufferPos - entityLen; + for(int i = 0; i < rem; i++) + dest[i] = cBuffer[i+entityLen]; + dest += rem; + if (pre) + prePos += rem; + } + if (pre) + prePos++; } else { #ifdef TOKEN_DEBUG kdDebug( 6036 ) << "unknown entity!" << endl; @@ -734,7 +747,6 @@ for(unsigned int i = 0; i < cBufferPos; i++) dest[i] = cBuffer[i]; dest += cBufferPos; - Entity = NoEntity; if (pre) prePos += cBufferPos+1; } @@ -1154,7 +1166,7 @@ type.compare("text/livescript") != 0 && type.compare("application/x-javascript") != 0 && type.compare("application/x-ecmascript") != 0 && - type.compare("application/javascript") != 0 && + type.compare("application/javascript") != 0 && type.compare("application/ecmascript") != 0 ) javascript = false; } else if( a ) { --- branches/KDE/3.5/kdelibs/khtml/html/htmltokenizer.h #601018:601019 @@ -133,10 +133,10 @@ void timerEvent( QTimerEvent *e ); virtual void setOnHold(bool _onHold); void abort() { m_abort = true; } - virtual void setAutoClose(bool b=true); + virtual void setAutoClose(bool b=true); virtual bool isWaitingForScripts() const; virtual bool isExecutingScript() const; - + protected: void reset(); void addPending(); @@ -342,6 +342,7 @@ #define CBUFLEN 1024 char cBuffer[CBUFLEN+2]; unsigned int cBufferPos; + unsigned int entityLen; khtml::TokenizerString src; 18337 2006-10-31 18:03:21 +0000 2006-10-31 18:03:21 +0000 test case 122047.html text/html 530 kde PGh0bWw+CjxzdHlsZT4KCXRhYmxlIHsgZm9udC1mYW1pbHk6IGZpeGVkfQo8L3N0eWxlPgo8Ym9k eT4KVGVzdCBvZiBIVE1MIGVudGl0aWVzIGluIHF1aXJreSBtb2RlOgo8dGFibGUgd2lkdGg9MjAw Pgo8dHI+PHRkPgogICAgJmFtcDthbXA7CTx0ZD4mYW1wOwo8dHI+PHRkPgogICAgJmFtcDthbXAJ PHRkPiZhbXAKPHRyPjx0ZD4KICAgICZhbXA7YW1wbGUJPHRkPiZhbXBsZQo8dHI+PHRkPgogICAg JmFtcDtub3Q7CTx0ZD4mbm90Owo8dHI+PHRkPgogICAgJmFtcDtub3QJPHRkPiZub3QKPHRyPjx0 ZD4KICAgICZhbXA7bm90YXQJPHRkPiZub3RhdAo8dHI+PHRkPgogICAgJmFtcDtub3RpbjsJPHRk PiZub3RpbjsKPHRyPjx0ZD4KICAgICZhbXA7bm90aW4JPHRkPiZub3Rpbgo8dHI+PHRkPgogICAg JmFtcDtub3RpbmEJPHRkPiZub3RpbmEKPHRyPjx0ZD4KICAgICZhbXA7Z2U7CTx0ZD4mZ2U7Cjx0 cj48dGQ+CiAgICAmYW1wO2dlCTx0ZD4mZ2UKPHRyPjx0ZD4KICAgICZhbXA7Z2VsCTx0ZD4mZ2Vs CjwvdGFibGU+CjwvYm9keT4=