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=-1.4 required=5.0 tests=BAYES_00,FREEMAIL_FROM,
	SUBJ_ALL_CAPS autolearn=no autolearn_force=no version=3.4.4
X-Google-Thread: 103376,ade59281d0eea302
X-Google-Attributes: gid103376,public
X-Google-Language: ENGLISH,ASCII-7-bit
Path: 
 g2news1.google.com!news2.google.com!news.glorb.com!newsgate.cistron.nl!xs4all!news.tele.dk!news.tele.dk!small.news.tele.dk!lnewsinpeer00.lnd.ops.eu.uu.net!emea.uu.net!peer-uk.news.demon.net!kibo.news.demon.net!mutlu.news.demon.net!news.demon.co.uk!demon!not-for-mail
From: Simon Wright <simon.j.wright@mac.com>
Newsgroups: comp.lang.ada
Subject: Re: STORAGE_ERROR : EXCEPTION_STACK_OVERFLOW
Date: Wed, 11 Apr 2007 22:31:09 +0100
Organization: Pushface
Message-ID: <m2d52afvj6.fsf@mac.com>
References: <1175494388.509572.267790@l77g2000hsb.googlegroups.com>
 <1176259776.630617.50120@e65g2000hsc.googlegroups.com>
NNTP-Posting-Host: pogner.demon.co.uk
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Trace: news.demon.co.uk 1176327069 3017 62.49.19.209 (11 Apr 2007 21:31:09
 GMT)
X-Complaints-To: abuse@demon.net
NNTP-Posting-Date: Wed, 11 Apr 2007 21:31:09 +0000 (UTC)
Cancel-Lock: sha1:RiFH3X2AcSiHVe0zxcFFOA3U6xE=
User-Agent: Gnus/5.11 (Gnus v5.11) Emacs/22.0.95 (darwin)
Xref: g2news1.google.com comp.lang.ada:14918
Date: 2007-04-11T22:31:09+01:00
List-Id: <comp.lang.ada>

Looking over your first complete posted code, I see that the procedure
schema_types.createtable has an unusual approach when writing the
schema to the file.

Normally you rely on the compiler to support streaming through the
inbuilt 'Output, 'Input, 'Wrtte, 'Read operations. The only place
where this falls down in your code is when outputting the contents of
attributes (an array of classwide pointers to attributes).

The approach I'd start with is to override the default 'Output and
'Input for attribute_types.attribute_ptr so that what you write to
disk is the thing designated by the pointer, and when you read back
you allocate the appropriate memory and return the pointer
.. something like

  return new attribute'class (attribute'class'input (fin));

This would completely remove all the hand-written stuff about
outputting and inputting the different attribute records.


I noticed when trying with T3 that the file's index (p fin.all in the
debugger) was 104 which is not in any sensible position in the
file. Something has got confused (I restored the commented-out line in
schema_types.loadtable that actually reads attribute info back, no
idea if it's getting called ... no, seems not)