Hello trid users,
some days weeks i must migrate to Windows 10. During that process i
lost some Virtual Box machines. So i must recreate my machines. Often
i use VirtualBox Extension Packs. These are typically stored as an
CD-ROM image with name like VBoxGuestAdditions.iso. Sometimes i have
difficulties to install some extensions. So i looked at the files on
that iso image. There appear a file with name TRANS.TBL inside the
directories.
So i run trid utility on such TBL files. All samples are described as
"Unknown!" (See appended output/trid-v-old.txt).
For comparison reason i also run the file format identification
utility DROID ( See
https://sourceforge.net/projects/droid/). Here the
examples are described wrong as "Pagemaker TableEditor Graphics" by
PUID x-fmt/198 because file name Extension is used as recognition
method.
For comparison reason i also run file command (version 5.44) on such
samples. Here all samples are described generic as "ASCII text". Some
samples are described with additional phrase "with very long lines".
It also shows generic mime type text/plain (see appended
output/file-i-5.44.txt) and no file name suffix (see appended
output/file-ext-5.44.txt).
Luckily i found a page about TRANS.TBL on file formats archive team.
So i use this as reference URL. That is expressed by line like:
<RefURL>
http://fileformats.archiveteam.org/wiki/TRANS.TBL</RefURL>
I could also a similar page on Wikipedia that directs to section Other
extensions on ISO 9660 page. But on the used side i also found links
with sample files to download and the Wikipedia is there also
mentioned.
So now i approximately know what patterns should appear in TrID
definition. Then i run tridscan on such TBL samples to generate trid
definition. Then i look inside definition for patterns and try to
understand why things happens. In Front Block with only few samples i
get some patterns.That looks like:
<Pattern>
<Bytes>2020202020202020202020202020202020</Bytes>
<Pos>65</Pos>
</Pattern>
<Pattern>
<Bytes>20</Bytes>
<Pos>75</Pos>
</Pattern>
<Pattern>
<Bytes>2020</Bytes>
<Pos>80</Pos>
</Pattern>
<Pattern>
<Bytes>2020202020202020</Bytes>
<Pos>117</Pos>
</Pattern>
<Pattern>
<Bytes>20202020202020</Bytes>
<Pos>164</Pos>
</Pattern>
That is not surprising because the "short" and "long" file names are
separated by some white space. So i create some ISO images where
content contains files with longer names. So the white space parts
shrink or move to different positions. So the above XML constructs
vanish and only one pattern survives. That looks like:
<Bytes>4620</Bytes>
<ASCII> F</ASCII>
<Pos>0</Pos>
The upper case letter F at the beginning means first entry is file.
Then comes the separating space. Unfortunately there exist no
precisely specification, but in my examples it was exactly one space
character. Afterward comes the "short" ISO 99960 name and in some
samples the trailing ;1 is missing. That is not explicitly written in
the mentioned documentation. Afterwards comes next separating space.
Here i found samples where this consist of just 1 tabular character
and in others these are a mixture of some spaces and tabulator
character. Then comes the "long" name which can also contain
spaces. In my samples the lines are terminated with a line feed
character.
In my generated examples the name was always TRANS.TBL, but i also
download one example with names 00_TRANS.TBL. So that is expressed
inside definition by line like:
<Ext>TBL</Ext>
The suffix TBL is apparently an abbreviation for table. Unfortunately
there exist also other table formats which also use this suffix. Even
the TrID database contains some of such examples. This is listed in
the following table:
TrID definition mime type description
tbl-aegis-impact.trid.xml x Aegis Impact! Table
tbl-amn.trid.xml text/xml AM-Notebook Spreadsheet/Table
tbl-corum.trid.xml x Corum VIP Table data
tbl-dxb.trid.xml text/plain Duxbury conversion Table
tbl-pm5.trid.xml x PageMaker Table
tbl-st6.trid.xml text/plain ST-6 color Table
tbl-unicode.trid.xml x Binary Unicode conversion Table
My inspected TBL samples are related with CD-ROM images which are
described by iso-9660-image.trid.xml as "ISO 9660 CD image". So i
choose tbl-iso9660.trid.xml as trid definition name. The description
is expressed by line like:
<FileType>ISO 9660 names translation table</FileType>
On Windows systems the file type is determined by file name
extension. In this case this is TBL. If you have installed seven of
the above mentioned applications and you click inside your file
manager on a PageMaker Table then by lucky circumstances the PageMaker
program starts if this is the standard application for TBL suffix, but
without luck one of the other six applications will start. That is bad
or inefficient!
When you apply this Microsoft behavior in the real world, then a woman
with names Snyder handling with needles and pins which marries a man
with name smith now must be hammering on iron. That is stupid, because
name is now correlated with person type.
But instead of improving or total replacing that system in general
Microsoft is trying to fix the symptoms. So in the Explorer by
default the file name suffix is hidden. So malware writer send
executable like CatPicture.jpg.exe to non-experienced users where this
look for such people as normal JPEG picture CatPicture.jpg.
Even worse as you tuber nick also complains the bad designs of
Microsoft are poisoning Linux. Many distribution say and think we do
it the same way as Microsoft because of their leadership in desktop
computers. Luckily on Linux system you can use thee mime type system
to determine the file type. But like in the energy twist in Germany
you can stop or reduce the current systems when the alternatives are
not well approved and working. When we look at current mime types for
TBL then the mime type gives no bargain at the moment, because in
principal it offers only 2 choices. One isapplication/octet-stream.
Normally such samples are opened inside a hex-editor. The other type
is text/plain. So such samples are normally opened in a text editor,
where you can make your wanted changes. The TBL samples are just text
files. So opening in text editor is principal OK, but only in viewing
mode, because you can not save changes done in editor, because CD-ROM
images are only readable. The names and then corresponding names and
relationships are recreated by you burning software. So the first
action for such TBL samples is just text viewing. When you transfer
the contents of an ISO9660 CD-ROM for example on an USB stick with
VFAT file system, then you still get "long" file names and "short"
file names are done by the FAT file system itself. So the TBL are not
needed any more here, use just disc space and is confusing normal
users. So the only useful second action for such samples on such file
systems is deleting. To trigger such actions of course an unique and
useful mime type must be provided. So i choose an user defined
one. That is expressed by line like:
<Mime>text/x-iso9660</Mime>
With the new trid definition such TBL examples are now described (see
appended output/trid-v-new.txt). TrID definitions, and output are
stored in archive tbl_.zip. I hope that my definition can be used i
future version of triddefs.
According to documentation there should exist samples where first entry
starts with directory. Then upper case letter D occurs at the
beginning in TBL. Unfortunately i do not found such samples and i also
could not create such samples. If such samples exist then of course
these are not described by my created definition.
With best wishes
Jörg Jenderek
