Hi Andreas, many thanks for your useful advice! I had wondered if there was a standard destination for help output. The latest version (67) in the svn repository incorporates your comments.
By the way, lastal has two new options, -F and -G, which are undocumented in the manual and half-implemented. I hope to finish this within a week or so. And thanks for Debianizing LAST! Martin Frith http://www.cbrc.jp/~martin/ On Wed, Sep 30, 2009 at 09:59:28AM +0200, Andreas Tille wrote: > Hi, > > when preparing the latest version of last as Debian package (here named > last-align) I tried to provide man pages for all tools provided in the > package to comply with Debian policy. The easiest way to do this is > simply using the help2man tool which produces more or less reasonable > manpages from the output you give when using the help option > (-h/--help). > > When doing so I stumbled upon the fact that not all tools direct the > help output to stdout but rather stderr. This is not conform to GNU > coding standards and thus intentionally not supported by help2man[1]. > So I did a small patch (attached) to enable output to stdout (and > defining a -h option for lastdb which was missing). Please feel free > to apply this patch in your next version. > > I would also be really happy if > last-map-probs, last-reduce-alignments, last-remove-dominated, > maf-join, maf2tab > would get a -/--help option as well to be able to provide rudimentary > manpages also for these tools. > > IMHO this would be in the interest of your users - properly maintained > manpages with real text would be even better. > > Hope this helps a bit enhancing your fine software. > > Kind regards and thanks for maintaining last > > Andreas. > > [1] http://bugs.debian.org/138752 > > -- > http://fam-tille.de > Author: Andreas Tille <ti...@debian.org> > Description: According to GNU coding standards help messages should go to > stdout > not to stderr. THis patch ensures that help2man will work correctly > --- last-62.orig/src/LastalArguments.cc > +++ last-62/src/LastalArguments.cc > @@ -4,6 +4,8 @@ > #include "stringify.hh" > #include <unistd.h> // getopt > #include <sstream> > +#include <cstdlib> > +#include <iostream> > #include <vector> > #include <stdexcept> > #include <cstring> // strtok > @@ -108,7 +110,8 @@ > != -1 ){ > switch(c){ > case 'h': > - throw std::runtime_error(help); > + std::cout << help ; > + exit(1); > case 'o': > outFile = optarg; > break; > --- last-62.orig/src/LastdbArguments.cc > +++ last-62/src/LastdbArguments.cc > @@ -2,6 +2,8 @@ > > #include "LastdbArguments.hh" > #include "stringify.hh" > +#include <cstdlib> > +#include <iostream> > #include <unistd.h> // getopt > #include <stdexcept> > > @@ -27,6 +29,7 @@ > usage: lastdb [options] output-name fasta-sequence-file(s)\n\ > \n\ > Main Options (default settings):\n\ > +-h: show all options and their default settings\n\ > -p: interpret the sequences as proteins\n\ > -c: read the sequences case-sensitively\n\ > -m: periodic spaced-seed pattern (" + maskPattern + ")\n\ > @@ -40,8 +43,11 @@ > "; > > int c; > - while( (c = getopt(argc, argv, "pcm:w:s:a:b:v")) != -1 ) { > + while( (c = getopt(argc, argv, "hpcm:w:s:a:b:v")) != -1 ) { > switch(c){ > + case 'h': > + std::cout << usage; > + exit(1); > case 'p': > isProtein = true; > break; -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
