gzip decompression not always working

[Opher Shachar]

> Opher Shachar wrote:
> >> Opher Shachar wrote:
> >>> ...
> >>> But if I issue the command:
> >>>
> >>>    lp MyPostscriptFile.gz -odocument-format=text/plain
> >>>
> >>> to get the postscript source to print I get gibberish and the following in error_log:
> >> Of course you get gibberish - you are telling the system to use
> >> text/plain as the format and it just obeys.  You can use
> >> "-o document-compression=gzip" to get the file uncompressed during
> >> printing, but if you override the auto-detection you have to do it
> >> fully.
> >
> > I have both a CUPS 1.2.4 (on RH 5.2) and CUPS 1.2.12 (on Fedora). I tried
> >     lp MyPostscriptFile.gz -odocument-format=text/plain -odocument-compression=gzip
> > but the result was the same (save that 'document-compression=gzip' was now added to the fifth positional parameter).
> > Going over the code for ipp.c it seems there's no code to handle this attribute.
> >
> > Is this attribute implemented on 1.3.9? 1.4?
>
> It is implemented in CUPS 1.2, however cupsd will auto-detect the
> compression of the document when it auto-types the print file so
> you normally don't need to specify the type of compression used.
>
> If that isn't working, please file a bug:
>
>      http://www.cups.org/str.php
>

I did and you indicated that the attribute to use is 'compression'.
I tried that also but it also doesn't work. e.g.:

    lp MyPostscriptFile.gz -odocument-format=text/plain -ocompression=gzip

gives gibberish. I suspect this is because 'cupsEncodeOptions2()' encodes the 'compression' option with:

   group_tag = IPP_TAG_JOB
   value_tag = IPP_TAG_NAME

instead of:

   group_tag = IPP_TAG_OPERATION
   value_tag = IPP_TAG_KEYWORD

Should I open a new STR on this?

Regards,
Opher Shachar.