home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
icon
/
newsgrp
/
group95b.txt
/
000154_icon-group-sender _Wed Sep 13 09:22:41 1995.msg
< prev
next >
Wrap
Internet Message Format
|
1995-09-18
|
2KB
Received: by cheltenham.cs.arizona.edu; Wed, 13 Sep 1995 09:02:48 MST
Date: Wed, 13 Sep 1995 09:22:41 -0400
From: motto@srd2000.srd.census.gov (Mark Otto)
Message-Id: <9509131322.AA10139@srd2000>
To: icon-group@cs.arizona.edu
Subject: Problem with decode
Content-Length: 1822
Errors-To: icon-group-errors@cs.arizona.edu
I am using encode and decode (from the IPL) to read and write records
of a database with a complicated record structure. These routines are
very useful because once I create the structures, I can decode or read
them easily and use the records in a number of different programs.
Unfortunately when I work with a large number of records (around 800),
I get run-time errors.
Run-time error 302
File codeobj.icn; Line 220
memory violation
Trace back:
derecord("10smembershiplcd...","lcdf") from line 167 in codeobj.icn
decode("10smembershiplcd...",1) from line 220 in codeobj.icn
derecord("7slmemberlcdh29R...","lcdg") from line 167 in codeobj.icn
{create ..} from line 220 in codeobj.icn
According to the Icon book, error 302 is a stack overflow, but I am
getting a memory violation. Is this a problem with the interpreter?
I am running on a Sparc 2000 running Solaris 2.3 and Icon Interpreter
Version 9.0. July 16, 1994. In Ralph Griswold's codeobj routine,
statement 220 is creating a coexpression.
218 procedure detable(s,tag)
219 local t, e
220 e := create decode(s,1) # see derecord above; here it's the default
221 # value that motivates co-expressions.
222 inlab[tag] := t := table(@e)
223 while t[@e] := @e
224 return t
225 end
I assume the coexpression stack is overflowing, but changing the
COEXPSIZE environmental variable doesn't change the error or where the
error occurs. Has anyone else had similiar experiences with these
routines? I need help diagnosing this problem.
Thanks for your help.
Mark Otto e-mail: Mark.Otto@Census.GOV (at the smtp gateway)
Census Bureau, Rm 3000-4 (SRD) otto@asgard.umd.edu
Washington, DC 20233-9100 office: 301-457-4728 fax: 301-457-2299
home: 301-431-1838
:wq