Re: [Groff] Formatting algorithm, an experiment

[Ralph Corderoy]

Hi Steffen,

> > The closest is
> > http://www.iana.org/assignments/media-types/application/gzip which
> > leaves the tar for the user to fathom out if left to MIME types
> > alone.
>
> [http://svn.apache.org/viewvc/tika/trunk/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml]
> lists the extension tgz regulary thereunder
> 
>   <mime-type type="application/gzip">
>     <_comment>Gzip Compressed Archive</_comment>
>     <alias type="application/x-gzip"/>
>     <alias type="application/x-gunzip"/>
>     <alias type="application/gzip-compressed"/>
>     <alias type="application/gzipped"/>
>     <alias type="application/gzip-compressed"/>
>     <alias type="application/x-gzip-compressed"/>
>     <alias type="gzip/document"/>
>     <magic priority="45">
>       <match value="\037\213" type="string" offset="0" />
>       <match value="\x1f\x8b" type="string" offset="0" />
>     </magic>
>     <glob pattern="*.tgz" />
>     <glob pattern="*.gz" />
>     <glob pattern="*-gz" />
>     <glob pattern="*.emz" />
>   </mime-type>

Interesting, but the comment seems wrong, since foo.gz need not be an
archive, i.e. a collection of other things, and it's lumping foo.gz and
foo.tgz together, so no indication that one of them is a tar file.

MIME types just don't seem to allow layering.  That file also has a
couple of tar types.

    <mime-type type="application/x-tar">
        <magic priority="40">
            <!-- POSIX tar archive -->
            <match value="ustar\0" type="string" offset="257" />
        </magic>
        <glob pattern="*.tar"/>
    </mime-type>

    <mime-type type="application/x-gtar">
        <_comment>GNU tar Compressed File Archive (GNU Tape Archive)</_comment>
        <magic priority="50">
            <!-- GNU tar archive -->
            <match value="ustar  \0" type="string" offset="257" />
        </magic>
        <glob pattern="*.gtar"/>
        <sub-class-of type="application/x-tar"/>
    </mime-type>

The second's comment suggests it's compressed but then it wants `ustar'
at offset 257 which is uncompressed.

It's a mess.  HTTP can ship an application/x-tar with a Content-Encoding
of gzip, but that's not the same thing as the client should ungzip to
deliver the tar;  not what's wanted.

This seems to be getting off-groff topic.  :-)

Cheers, Ralph.