bug-gnu-pspp
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

PSPP-BUG: Failure to handle an antique SPSS file containing umlauts in v


From: Müller , Andre
Subject: PSPP-BUG: Failure to handle an antique SPSS file containing umlauts in variable names
Date: Thu, 23 Jan 2014 17:26:59 +0000

Dear all,

I stumbled over a rather esoteric problem when dealing with an ooold sav file.
(Actually, I am led to believe the file was written when the archive migrated 
off it's mainframe...)

The export with pspp-0.8.1 fails like this:
error: `source.sav' near offset 0x608: Invalid variable name `V_P?NKT'.

SavConversion/pspp-export.syntax:2: error: Stopping syntax file processing here
to avoid a cascade of dependent command failures.

The variable name actually is "V_PÜNKT", and there are more Umlauts to come.
Trying to write a testcase, I found:
- When I create a file in UNICODE locale, all is well with the export.
- When I create a file in localcode locate, the export misinterprets the 
umlauts, but otherwise does the job.
- When, using SPSS, I save a copy of the offending file, that file gets handled 
properly.
- Just when using the old file as is, the export fails as described above.

So, there is an easy workaround for this problem, and it really only concerns a 
way old .sav file version.
In other words, this is not too important and it may well not be worth the 
effort of fixing this.

But I let you decide :-)
I will open a proper bug for this on request.

Due to licensing restrictions, I cannot provide a copy of the file,
but you can get it here: 

https://dbk.gesis.org/dbksearch/download.asp?db=E&id=40582

The download requires a registration and acceptance of the TOC,
which in a nutshell requires you to use it for the purpose stated on download,
and do never ever distribute it to any third party. Or attach  (The rationale 
being that 
no old, erroneous revisions are in circulation.)

Thank you,
Andre Müller



reply via email to

[Prev in Thread] Current Thread [Next in Thread]