No, do the following:


  2    while (inbytesleft &&
  3        iconv(cd,
  4 #ifdef UTF8_ICONV_CAST_OK
  5        (char**)(&src),
  6 #else
  7        (char**)(&src),  <================= HERE IS THE CHANGE
  8 #endif
  9       &inbytesleft, &outbuf, &outbytesleft) != -1);
  10    iconv_close (cd);
  11    return SMALL_BUFFER_SIZE - outbytesleft;
  12 #endif
  13 }


Lou Spironello wrote:

> Thank you.
> 
> Just to clarify things for me.
> This is the area of the offending code.  I'm assuming that line 9 should read:
> 
>        char(**)&inbytesleft, &outbuf, &outbytesleft) != -1);
> 
> instead of
> 
>        &inbytesleft, &outbuf, &outbytesleft) != -1);
> 
> 
> 
> 1    // but RH7 problem probably persists, will have to fork some more if so
> 2    while (inbytesleft &&
> 3        iconv(cd,
> 4 #ifdef UTF8_ICONV_CAST_OK
> 5        (char**)(&src),
> 6 #else
> 7        &src,
> 8 #endif
> 9       &inbytesleft, &outbuf, &outbytesleft) != -1);
> 10    iconv_close (cd);
> 11    return SMALL_BUFFER_SIZE - outbytesleft;
> 12 #endif
> 13 }
> 
> 
> ----- Original Message -----
> From: "Pavel Hlavnicka" <[EMAIL PROTECTED]>
> To: "Sablotron Mailing List" <[EMAIL PROTECTED]>
> Sent: Wednesday, January 10, 2001 8:47 AM
> Subject: Re: [Sab] Make problems. :-(
> 
> 
> 
>> It is the well known problem, already fixed (but not published).
>> 
>> Now you can fix this problem, if you add (char**) before the second
>> param of iconv() call.
>> 
>> Pavel
>> 
>> Lou Spironello wrote:
>> 
>> 
>>> Thank you Pavel.  That seemed to correct that compile problem.  Now I'm
>> 
> getting
> 
>>> another one in utf8.cpp.  :-(
>>> 
>>> This is the output from make:
>>> 
>>>   mv -f .libs/uri.lo uri.lo
>>> /bin/sh ../../libtool --mode=compile
>>> -DHAVE_CONFIG_H -I. -I. -I../../autocfg     -g -O2 -c utf8.cpp
>>> rm -f .libs/utf8.lo
>>> c++ -DHAVE_CONFIG_H -I. -I. -I../../autocfg -g -O2 -c
>> 
> utf8.cpp  -fPIC -DPIC -o
> 
>>> .libs/utf8.lo
>>> utf8.cpp: In function `int utf8Recode (char *, const char *,
>>> Encoding)':
>>> utf8.cpp:274: cannot convert `const char **' to `char **' for argument
>>> `2' to `iconv (void *, char **, size_t *, char **, size_t *)'
>>> make[2]: *** [utf8.lo] Error 1
>>> make[2]: Leaving directory `/home/spiro/xml/Sablot-0.50/Sablot/engine'
>>> make[1]: *** [all-recursive] Error 1
>>> make[1]: Leaving directory `/home/spiro/xml/Sablot-0.50/Sablot'
>>> make: *** [all-recursive] Error 1
>>> [root@spiro1 Sablot-0.50]#
>>> 
>>> 
>>> Thank you.
>>> Lou Spironello
>>> 
>>> ----- Original Message -----
>>> From: "Pavel Hlavnicka" <[EMAIL PROTECTED]>
>>> To: "Sablotron Mailing List" <[EMAIL PROTECTED]>
>>> Sent: Wednesday, January 10, 2001 7:24 AM
>>> Subject: Re: [Sab] Make problems. :-(
>>> 
>>> 
>>>> Lou Spironello wrote:
>>>> 
>>>> 
>>>>> That's what I thought also.  So, I recomplied the latest version of expat
>>>>> without errors.
>>>>> I also set the LD... and CPLUS... environment variables to /usr/local/lib
>>>> 
>>> and
>>> 
>>> 
>>>>> /usr/local/include repspectively.
>>>>> 
>>>>> I still get the following
>>>>> 
>>>> 
>>>> This line in the ouput of configure script says, that expat header is
>>>> not found. Note, that this value is cached. If you have installed new
>>>> wersion of expat after running the configure the first time, you have to
>>>> delete config.* files from the Sablotron source tree to force configure
>>>> re-run the tests.
>>>> 
>>>> Also try to set CPPFLAGS instead of C_INCLUDE_DIR & Co.
>>>> 
>>>> 
>>>> 
>>>>>>> checking for expat.h... (cached) no
>>>>>> 
>>>> --
>>>> Pavel Hlavnicka
>>>> Ginger Alliance Ltd.
>>>> Prague; Czech Republic
>>>> 
>>>> 
>>> 
>> 
>> --
>> Pavel Hlavnicka
>> Ginger Alliance Ltd.
>> Prague; Czech Republic
>> 
>> 


-- 
Pavel Hlavnicka
Ginger Alliance Ltd.
Prague; Czech Republic

Reply via email to