[pcre-dev] PCRE UTF8 umlauts CASELESS matching

Hi there,
I'm trying to use libpcre3 for pattern matching. ;-) My problem is, that
I can't achieve that the german umlaut "Ä" (Ä) matches "ä" (ä)
when using CASELESS option for pattern matching. Since my project uses
unicode I tried it with UTF8. But using a character table didn't work
for me, too.
My project uses Qt4 and to create UTF8 strings I used some Qt code in my
minimal test code. Perhaps you could convert it, if you don't have qt
development files installed.

First I only had the mentioned problem (in the project code), that
Ä didn't match ä.
Now (in my example code) Ä even doesn't match Ä and I don't
know why. I tried quite a few things.
Minimal example code is attached.

My output is as follows:
codec: "UTF-8"
UNICODE_PROPERTIES: 1
match: -1

I believe I'm doing something stupid, so if you could show me an example
where I can caseless match "Ä" (Ä) with "ä" (ä), I would
really appreciate that.

Sincerely,
Enno Gröper

#############################################################################

# Makefile for building: cdcat
# Generated by qmake (2.01a) (Qt 4.2.3) on: Fr Okt 26 22:54:54 2007
# Project:  cdcat.pro
# Template: app
# Command: /usr/bin/qmake-qt4 -unix -o Makefile cdcat.pro
#############################################################################

####### Compiler, tools and options

CC            = gcc
CXX           = g++
LEX           = flex
YACC          = yacc
DEFINES       = -DQT_QT3SUPPORT_LIB -DQT3_SUPPORT -DQT_GUI_LIB -DQT_CORE_LIB
CFLAGS        = -pipe -g -D_REENTRANT -Wall -W  $(DEFINES)
CXXFLAGS      = -pipe -g -D_REENTRANT -Wall -W  $(DEFINES)
LEXFLAGS      = 
YACCFLAGS     = -d
INCPATH       = -I/usr/share/qt4/mkspecs/linux-g++ -I. -I/usr/include/qt4/QtCore -I/usr/include/qt4/QtCore -I/usr/include/qt4/QtGui -I/usr/include/qt4/QtGui -I/usr/include/qt4/Qt3Support -I/usr/include/qt4/Qt3Support -I/usr/include/qt4 -Imoc_files -I.
LINK          = g++
LFLAGS        = 
LIBS          = $(SUBLIBS)  -L/usr/lib -lpcre -lQtCore
TARGET        = pcre_tst

####### Files

SOURCES       = pcre_tst.cpp
OBJECTS       = pcre_tst.o
first: all
####### Implicit rules

.SUFFIXES: .o .c .cpp .cc .cxx .C

.cpp.o:
    $(CXX) -c $(CXXFLAGS) $(INCPATH) -o "$@" "$<"

.cc.o:
    $(CXX) -c $(CXXFLAGS) $(INCPATH) -o "$@" "$<"

.cxx.o:
    $(CXX) -c $(CXXFLAGS) $(INCPATH) -o "$@" "$<"

.C.o:
    $(CXX) -c $(CXXFLAGS) $(INCPATH) -o "$@" "$<"

.c.o:
    $(CC) -c $(CFLAGS) $(INCPATH) -o "$@" "$<"

####### Build rules

all: $(TARGET)

$(TARGET):  $(OBJECTS)  
    $(LINK) $(LFLAGS) -o $(TARGET) $(OBJECTS) $(OBJCOMP) $(LIBS)

clean: 
    -rm -f $(OBJECTS)
    -rm -f $(TARGET)

####### Compile

pcre_tst.o: pcre_tst.cpp
    $(CXX) -c $(CXXFLAGS) $(INCPATH) -o pcre_tst.o pcre_tst.cpp

#include <QString>

#include <QTextCodec>
#include <pcre.h>
#include <string.h>
#include <ctype.h>
#include <QtDebug>

int main(int argc, char * argv[]) {
    QString patt;
    QString txt;
    pcre *re;
    int pcre_opt = PCRE_UTF8;
    pcre_extra *hints;
       const char *error;
       int errptr;
    int  offsets[99];
    int match;
    int utf8prop;
    QTextCodec * codec;

    QTextCodec::setCodecForCStrings(QTextCodec::codecForName("UTF-8"));
    codec = QTextCodec::codecForCStrings();
    if (codec == NULL)
        qDebug() << "standard codec (latin-1)";
    else
        qDebug() << "codec:" << codec->name();

    pcre_opt |= PCRE_CASELESS;
    patt = "/Ä/";
    QByteArray encPatt = patt.toUtf8();

    pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &utf8prop);
    qDebug() << "UNICODE_PROPERTIES:" << utf8prop;

    //re = pcre_compile((const char*)patt.data(),pcre_opt,&error,&errptr,NULL);
    re = pcre_compile(encPatt.constData(),pcre_opt,&error,&errptr,NULL);

    if(re == NULL)
       {
        qDebug() << "(comp) Error in the pattern:" << error << ": " << errptr;
        return 1;
       }

      hints = pcre_study(re,0,&error);

      if(error != NULL)
       {
        qDebug() << "(study) Error in the pattern:" << error;
        return 1;
       }
    txt = "Ä";
    QByteArray enc = txt.toUtf8();
    //match   = pcre_exec(re,hints,(const char*)&txt,txt.size(),0,0,offsets,99);
    match   = pcre_exec(re,hints,enc.constData(),enc.size(),0,0,offsets,99);
    qDebug() << "match:" << match;

    return 0;
}