hashirama/goldendict-ng
hashirama/goldendict-ng
0
0
Fork 0
mirror of https://github.com/xiaoyifang/goldendict-ng.git synced 2024-11-30 21:34:07 +00:00
Code Issues Actions Projects Releases Wiki Activity

*! Do proper '<charset c=t>..</charset>' replacements with a regexp.

Browse source
This commit is contained in:
Konstantin Isakov 2009-05-07 16:14:56 +00:00
parent 44b5524a91
commit 484c3e356b
1 changed files with 20 additions and 88 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

104
src/bgl.cc
View file

@ -25,6 +25,8 @@
#include <QThreadPool> #include <QThreadPool>
#include <QAtomicInt> #include <QAtomicInt>
#include <QRegExp>
namespace Bgl { namespace Bgl {
using std::map; using std::map;
@ -826,103 +828,33 @@ sptr< Dictionary::DataRequest > BglDictionary::getResource( string const & name
/// Replaces <CHARSET c="t">1234;</CHARSET> occurences with &#x1234; /// Replaces <CHARSET c="t">1234;</CHARSET> occurences with &#x1234;
void BglDictionary::replaceCharsetEntities( string & text ) void BglDictionary::replaceCharsetEntities( string & text )
{ {
string lowercased = text; QRegExp charsetExp( "<\\s*charset\\s+c\\s*=\\s*[\"']?t[\"']?\\s*>((?:\\s*[0-9a-fA-F]+\\s*;\\s*)*)<\\s*/\\s*charset\\s*>",
Qt::CaseInsensitive );
// Make a lowercased version of text, used for searching only. Only touch charsetExp.setMinimal( true );
// symbols < 0x80 to avoid any weird results.
for( unsigned x = lowercased.size(); x--; )
if ( (unsigned char )lowercased[ x ] < 0x80 )
lowercased[ x ] = tolower( lowercased[ x ] );
size_t prevPos = 0; QRegExp oneValueExp( "\\s*([0-9a-fA-F]+)\\s*;" );
for( ;; ) QString str = QString::fromUtf8( text.c_str() );
for( int pos = 0; ( pos = charsetExp.indexIn( str, pos ) ) != -1; )
{ {
size_t pos = lowercased.find( "<charset c=\"t\">", prevPos ); //printf( "Match: %s\n", str.mid( pos, charsetExp.matchedLength() ).toUtf8().data() );
if ( pos == string::npos ) QString out;
break;
if ( lowercased.size() - pos < 30 ) for( int p = 0; ( p = oneValueExp.indexIn( charsetExp.cap( 1 ), p ) ) != -1; )
{ {
// This is not right, the string is too short, leave it alone //printf( "Cap: %s\n", oneValueExp.cap( 1 ).toUtf8().data() );
break; out += "&#x" + oneValueExp.cap( 1 ) + ";";
p += oneValueExp.matchedLength();
} }
prevPos = pos + 1; str.replace( pos, charsetExp.matchedLength(), out );
if ( lowercased.substr( pos + 15 + 4, 11 ) != ";</charset>" )
{
// The ending doesn't match
printf( "!!!!!!ending mismatch\n" );
continue;
} }
// Check if digits are all hex text = str.toUtf8().data();
if ( !isxdigit( lowercased[ pos + 15 ] ) ||
!isxdigit( lowercased[ pos + 16 ] ) ||
!isxdigit( lowercased[ pos + 17 ] ) ||
!isxdigit( lowercased[ pos + 18 ] ) )
{
printf( "!!!!!!!!not hex digits\n" );
continue;
}
// Ok, replace now.
lowercased.replace( pos, 15, "&#x" );
lowercased.erase( pos + 8, 10 );
text.replace( pos, 15, "&#x" );
text.erase( pos + 8, 10 );
}
prevPos = 0;
// Copy-pasted version for <charset c=t>. This should all be replaced
// by regexps.
for( ;; )
{
size_t pos = lowercased.find( "<charset c=t>", prevPos );
if ( pos == string::npos )
break;
if ( lowercased.size() - pos < 28 )
{
// This is not right, the string is too short, leave it alone
break;
}
prevPos = pos + 1;
if ( lowercased.substr( pos + 13 + 4, 11 ) != ";</charset>" )
{
// The ending doesn't match
printf( "!!!!!!ending mismatch\n" );
continue;
}
// Check if digits are all hex
if ( !isxdigit( lowercased[ pos + 13 ] ) ||
!isxdigit( lowercased[ pos + 14 ] ) ||
!isxdigit( lowercased[ pos + 15 ] ) ||
!isxdigit( lowercased[ pos + 16 ] ) )
{
printf( "!!!!!!!!not hex digits\n" );
continue;
}
// Ok, replace now.
lowercased.replace( pos, 13, "&#x" );
lowercased.erase( pos + 8, 10 );
text.replace( pos, 13, "&#x" );
text.erase( pos + 8, 10 );
}
} }
class ResourceHandler: public Babylon::ResourceHandler class ResourceHandler: public Babylon::ResourceHandler