[Bioperl-l] bug #1396: *.PL to *.pod conversion is broken, makebailsout

[Brian Osborne]

Hilmar,

The simple solution is to revert to plain documents, drop the *PL. files,
LocalConfig.pm. All this system does currently is put version numbers in 4
*pod files and the AUTHORS into bioperl.pod. To make it work somebody has to
enter a new version number into Makefile.PL, which writes to bioperl.conf.
The make executes the *PL which uses LocalConfig.pm which reads
bioperl.conf. It could be simpler, but this wouldn't fix your 1396 either.

Note that the maintainer of all this has not offered help during this
discussion. It's not exactly in the spirit of bioperl to remove work
contributed with the best of intentions but along with the contribution
comes a responsibility to maintain.

So: remove system? Or find workaround?

Brian O.

-----Original Message-----
From: bioperl-l-bounces at bioperl.org [mailto:bioperl-l-bounces at bioperl.org]On
Behalf Of Hilmar Lapp
Sent: Wednesday, February 26, 2003 5:02 PM
To: Brian Osborne
Cc: Bioperl
Subject: Re: [Bioperl-l] bug #1396: *.PL to *.pod conversion is broken,
makebailsout

No change, same error. The problem really is that the text is a
'here'-document enclosed between <<"THEEND" and ^THEEND. Perl doesn't
treat that as a script nor scans for documentation contained in there,
but it does substitute variables with their values. In fact, this is
the whole purpose of the .PL variant: if you look at the start of the
text, there is $Local{VERSION}, which gets converted to "1.2" in the
resulting POD.

The problem is - all the 'variables' which are merely in code snippets
for documentation purposes get substituted as well. Check what happens
to all the $something variables in the POD text -- do you still see
them in the generated POD? I just see emptiness in their place in the
POD ...

This is running perl 5.6.0 on Mac OSX, but the net effect shouldn't be
much different on any other version > 5.005 (since that's why the .PL
is there in the first place).

        -hilmar

On Wednesday, February 26, 2003, at 01:27  PM, Brian Osborne wrote:

> Hilmar,
>
> There were missing =cut delimiters in biodatabases.PL and
> biodesign.PL, I've
> put them in. Could you update and try again? Because I can't reproduce
> your
> bug I can't tell you whether 1396 is fixed or not. Adding the =cut's
> should
> enclose all the relevant text in these 2 files in proper POD, so no
> escaping
> should be necessary.
>
> Thanks again,
>
> Brian O.
>
> -----Original Message-----
> From: bioperl-l-bounces at bioperl.org
> [mailto:bioperl-l-bounces at bioperl.org]On
> Behalf Of Hilmar Lapp
> Sent: Wednesday, February 26, 2003 2:39 PM
> To: Bioperl
> Subject: [Bioperl-l] bug #1396: *.PL to *.pod conversion is broken,
> make
> bailsout
>
> At least biodatabases.PL and biodesign.PL contain @arys and $scalars
> which perl
> will (try to) substitute if unescaped. The @arys even choke the perl
> parser, but the $scalars also need to be escaped because otherwise they
> get substituted by empty values. This is a consequence of the text
> being run as a perl script first.
>
> I fixed those errors that make the parser choke, but someone needs to
> look at
> this and resolve the whole issue.
>
> donau: bioperl-live 11:22 85>make test
> /usr/bin/perl -Iblib/arch -Iblib/lib -I/System/Library/Perl/darwin
> -I/System/Library/Perl biodatabases.PL biodatabases.pod
> In string, @seqobjs now must be written as \@seqobjs at biodatabases.PL
> line 9,
> near "$seqobj)
>
> Similarly, one can query the database with id's or using query objects
> and retrieve arrays of Seq objects. For example
>
>    @seqobjs"
> Execution of biodatabases.PL aborted due to compilation errors.
> make: *** [biodatabases.pod] Error 255
>
> --
> -------------------------------------------------------------
> Hilmar Lapp                            email: lapp at gnf.org
> GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
> -------------------------------------------------------------
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at bioperl.org
> http://bioperl.org/mailman/listinfo/bioperl-l
>
>
>
--
-------------------------------------------------------------
Hilmar Lapp                            email: lapp at gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------

_______________________________________________
Bioperl-l mailing list
Bioperl-l at bioperl.org
http://bioperl.org/mailman/listinfo/bioperl-l