[emboss-dev] USA syntax and `%' character in sequence file names

[Nicolas Joly]

Hi,

We recently had a problem with a user that tried to use filenames that
contains `%' characters on our UNIX servers ... It does not seems to
work as expected, EMBOSS applications truncate the file name to be
used.

njoly at lanfeust [tmp/geecee]> rm -f foo%bar.data && touch foo%bar.data
njoly at lanfeust [tmp/geecee]> seqret -sequence=foo%bar.data -auto -stdout
Error: failed to open filename 'foo'
Error: Unable to read sequence 'foo%bar.data'
Died: seqret terminated: Bad value for '-sequence' with -auto defined

Removing the `%' character make this dummy test work as expected.

njoly at lanfeust [tmp/geecee]> rm -f foobar.data && touch foobar.data
njoly at lanfeust [tmp/geecee]> seqret -sequence=foobar.data -auto -stdout
Error: Unable to read sequence 'foobar.data'
Died: seqret terminated: Bad value for '-sequence' with -auto defined

I checked the USA format documentation but did not find any clue. Do
the `%' character has a special meaning in the USA syntax ?

Thanks.

-- 
Nicolas Joly

Biological Software and Databanks.
Institut Pasteur, Paris.