[PyKDE] QSocket.readBlock/QSocket.getch() difference

[Michael Ferraro]

thanks for the fixes --

since I had gotten PyQt3.0 as a tar I pulled the cvs tree down into a
new directory an *copied* the qsocket.sip file into my original source
tree.  Typed make. Make indicated that nothing need be done for 'all' in all
the subdirectories (including the PyQt-3.0/sip where I put qsocket.sip)
I checked the qsocket.sip date and file and the date is correct and the
file contains the fixes.

I then tried a make rebuild-src which reported that
    sip: Unable to find file "sip/qt.sip"

I tried to run sip on the qsocket.sip manually with
    sip -I sip -c qtnetwork -m qt/Makefile.am sip/qsocket.sip
and got:
    sip: sip/qsocket.sip:59: Lower bound is not a primary version in the
module

then went back to the version I checked out from cvs and did a make build,
followed
by a make which complained about a lot of things being not found
    copypyuic[23]: tmake: not found
    sed: can't read PyQt/pyuic/Makefile.win32-msvc: No such file or directory

    cp: PyQt/t1.html: No such file or directory
    /usr/local/src/PyQt-cvs/PyQt
    configtools[4]: aclocal: not found
    configtools[5]: autoheader: not found
    configtools[6]: automake not found
    configtools[7]: libtoolize: not found
    configtools[8]: autoconf: not found

How should I apply the fixes ??

Michael


Phil Thompson wrote:

> Michael Ferraro wrote:
> >
> >  hello
> >
> > in response to a readyRead() signal i check if a socket.bytesAvailable()
> >
> > is non-zero at which point I do a socket.readBlock(maxbytes).  The
> > buffer returned is zero bytes long and  doesn't have any data in it.
> >
> > As a test I followed the readBlock with a loop that makes calls to
> > socket.getch() which returns  valid data for the number of
> > bytesAvailable.
> >
> > I can use getch()s if I have to, but it seems that readBlock should work
> >
> > in here. What might I be doing wrong this time, I ask.
> >
> > def HandShake (self):
> >     socket = self.sender()
> >     n = socket.bytesAvailable()
> >     if n > 0:
> >         hail = socket.readBlock(32)
> >         print "readBlock(): ", len(hail), hail        # -- hail is empty
> >
> >         for i in range (n):
> >             print "getch(): ", socket.getch ()          # -- prints
> > valid data
>
> I think it's a bug in QSocket.readBlock() (and QSocket.readLine()).
>
> Should now be fixed in CVS.
>
> Phil
>
> _______________________________________________
> PyKDE mailing list    PyKDE at mats.gmd.de
> http://mats.gmd.de/mailman/listinfo/pykde