From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00
	autolearn=unavailable autolearn_force=no version=3.4.4
X-Received: by 2002:a6b:6315:: with SMTP id
 p21-v6mr169479iog.43.1530001736456;
        Tue, 26 Jun 2018 01:28:56 -0700 (PDT)
X-Received: by 2002:aca:c6ca:: with SMTP id w193-v6mr74454oif.1.1530001736350;
 Tue, 26 Jun 2018 01:28:56 -0700 (PDT)
Path: 
 eternal-september.org!reader01.eternal-september.org!reader02.eternal-september.org!feeder.eternal-september.org!news.uzoreto.com!weretis.net!feeder6.news.weretis.net!feeder.usenetexpress.com!feeder-in1.iad1.usenetexpress.com!border1.nntp.dca1.giganews.com!border2.nntp.dca1.giganews.com!nntp.giganews.com!u78-v6no543430itb.0!news-out.google.com!p13-v6ni607itf.0!nntp.google.com!u78-v6no543429itb.0!postnews.google.com!glegroupsg2000goo.googlegroups.com!not-for-mail
Newsgroups: comp.lang.ada
Date: Tue, 26 Jun 2018 01:28:56 -0700 (PDT)
Complaints-To: groups-abuse@google.com
Injection-Info: glegroupsg2000goo.googlegroups.com;
 posting-host=124.171.158.113;
 posting-account=d51RWwoAAADvR-x0zYAtT9z3CRxT1eXo
NNTP-Posting-Host: 124.171.158.113
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <f4c544c8-bbdf-48da-a27a-7611b880537d@googlegroups.com>
Subject: Problem Building Ada Interface To ImageMagick 7
From: Roger <rogermc@grapevine.com.au>
Injection-Date: Tue, 26 Jun 2018 08:28:56 +0000
Content-Type: text/plain; charset="UTF-8"
Xref: reader02.eternal-september.org comp.lang.ada:53318
Date: 2018-06-26T01:28:56-07:00
List-Id: <comp.lang.ada>

OSX 10.13.5
GNAT community 2018
XCode 9.4.1

ImageMagic Libraries:
libMagick++-7.Q16HDRI
libMagickCore-7.Q16HDRI

I have been developing some Ada Interfaces To ImageMagick 7 but have come to a problem that I need help with.
I built the ImageMagick libraries with:
CC=clang CXX="clang++ -std=c++11 -stdlib=libc++ -g -O0"  CFLAGS="-g -O0" CXXFLAGS="-g -O0" ./configure --prefix=/opt --with-quantum-depth=16 --disable-dependency-tracking --without-perl
I enabled debugging (-g) to allow me to track through the ImageMagick code , C++ in particular which I find almost impossible to reverse engineer from the source alone.
My C++ capabilities are very limited.

The following code works OK under XCode but crashes (segmentation fault) under GNAT.

#include <string>
#include <stdio.h>
#include "/opt/include/ImageMagick-7/Magick++.h"
#include "Image.h"

using namespace Magick;

int main()
{
  Image   anImage;

  ssize_t  max_length = 0;
  ssize_t  foo_len;
  ssize_t  copied_str_len = 0;
  ssize_t  buff_len = 4096;
  size_t   rows = 0;
  std::string foo("12345678901234567890\n12345678901234567890");
  std::string bf = "";
  char bar[buff_len];

  max_length=sizeof(foo) - 1;
  foo_len=foo.length();
  printf ("foo length %zd bytes.\n", foo_len);

  printf ( "Maximum length: %zd\n", max_length);
  printf ("Rows: %zd\n", rows);
  try {bf = anImage.baseFilename();}
  catch (Error& my_error) {printf ("%s", my_error.what());}
  catch (...) {printf ("Unhandled error\n");}
  printf ("Rows: %zd\n", rows);
    printf ("baseFilename: %s\n", anImage.baseFilename().c_str());

  printf ("Copied string length: %zd\n", copied_str_len);
  printf ("bar: \n%s\n", bar);
}

Under GNAT GPS it produces:
foo length 41 bytes.
Maximum length: 7
Rows: 0
[2018-06-26 18:02:10] process terminated successfully, elapsed time: 00.73s
So it appears to terminate after the first  printf ("Rows: %zd\n", rows);
That is, it terminates during the call to  anImage.baseFilename();

When I run the executable from the terminal:
foo length 41 bytes.
Maximum length: 7
Rows: 0
Segmentation fault: 11

Running the same code under XCode produces.

foo length 41 bytes.
Maximum length: 23
Rows: 0
Rows: 0
baseFilename: 
Copied string length: 0
bar: 

Program ended with exit code: 0
 The code above is a cut down version of my test code.
The complete version works under XCode.

 The problem seems to concern C++ string processing.
A problem that one of my build attempts reported  appeared to concern the string Swap function being unable to handle null parameters.
Unfortunately, I can't recall how I got that information.

I have tried numerous methods to solve this problem but am unable to. It seems that GNAT is using a different C++ compiler to XCode but I've had no success trying to remedy that idea.

Positive advice on how to proceed will be greatly appreciated.