Re: [9fans] Announce: port

Re: [9fans] Announce: port

[forsyth]

you've put absolute pathnames in the tar file.

Re: [9fans] Announce: port

[peter a. cejchan]

forsyth@caldo.demon.co.uk wrote in message news:<20020425072800.606741998C@mail.cse.psu.edu>...
> you've put absolute pathnames in the tar file.


Fixed. Also, your mkfile included. Recompiled. Find it at
http://next.gli.cas.cz/plan9/ape/rel-1.3.1.tbz
Thank you all!!!

++pac.

[9fans] Announce: port

[Peter A. Cejchan]

[-- Attachment #1: Type: text/plain, Size: 2 bytes --]



[-- Attachment #2: ANN --]
[-- Type: text/plain, Size: 975 bytes --]

Hi, friends!

Just to announce the port of two Unix utilities:

rel-1.3.tbz	sorts input text files according to relevance to given combination of keywords
pstotext-1.8g.tbz	extracts plain ascii text from ps/pdf document

Get the ports from http://next.gli.cas.cz/plan9/ape (the server is usually up 15:30 - 03:00 GMT;  may be down outside this time span, sorry (I still switch between Plan 9 and Linux, sigh).

Both programs are APE ports, thus, they know nothing of Unicode :-(. Rel is taken from Debian sources (look at http://www.debian.org, for a link to its homepage); pstotext, now part of gsview, is was grabbed direstly from its homepage at
http://research.compaq.com/SRC/virtualpaper/pstotext.html.
I received no response from the authors, so I suppose I may redistribute the modified sw with the original licenses. Correct me if I'm wrong.

Why I didn't make native ports? I am not skilled enough :-(((. However, you are encouraged.
Enjoy!

++pac.

Re: [9fans] Announce: port

[Boyd Roberts]

"Peter A. Cejchan" wrote:
> Why I didn't make native ports?
> I am not skilled enough :-(((. However, you are encouraged.

Native is actually a lot easier.

Re: [9fans] Announce: port

[peter a. cejchan]

boyd@strakt.com (Boyd Roberts) wrote in message news:<3CC7C6AD.4FFFE07@strakt.com>...
> "Peter A. Cejchan" wrote:
> > Why I didn't make native ports?
> > I am not skilled enough :-(((. However, you are encouraged.
>
> Native is actually a lot easier.

Really? I mean 'native' should be w UTF, w/o stdio.h, etc. Or does
native stdio.h treat char as wchar? I suppose not. However, I feel
that the whole thing should better be implemented (perhaps) as an rc
script using traditional grep, etc. Or am I wrong? Perhaps someone
more skilled than me :-( could give it a challenge!
Nevertheless, thanks to all who contributed. I will include Forsyth's
manpage, and tar it using relative paths ASAP.

Cheers,
++pac.

Re: [9fans] Announce: port

[Boyd Roberts]

"peter a. cejchan" wrote:
> > Native is actually a lot easier.
>
> Really? I mean 'native' should be w UTF, w/o stdio.h, etc. Or does
> native stdio.h treat char as wchar?

stdio is horrible.  wchar's are a disaster.

use bio:

    http://plan9.bell-labs.com/magic/man2html/2/bio

The compiler already understands about UTF and runes and so does
libc (obviously):

    http://plan9.bell-labs.com/sys/doc/comp.html

Plan 9 is very close to unix, but it's not really.  A lot of grief
gets caused by ape's heroic attempt to smash ANSI/POSIX into Plan 9.

It only gets really nasty when you have to port something that uses
the tty routines, sid's/pgrp's and signals -- most programs don't.

Re: [9fans] Announce: port

[Douglas A. Gwyn]

Boyd Roberts wrote:
> stdio is horrible.  wchar's are a disaster.

Sounds like somebody who doesn't use them enough to know.
wchar_t is closely analogous to rune.
The real problem is that "char" is inadequate for encoding a character,
largely a consequence of Dennis chiming in on the sizeof(char)==1 side.

Re: [9fans] Announce: port

[peter a. cejchan]

boyd@strakt.com (Boyd Roberts) wrote in message news:<3CC93C53.2D0DEF98@strakt.com>...
> "peter a. cejchan" wrote:
> > > Native is actually a lot easier.
> >
> > Really? I mean 'native' should be w UTF, w/o stdio.h, etc. Or does
> > native stdio.h treat char as wchar?
>
> stdio is horrible.  wchar's are a disaster.
>
> use bio:
>
>     http://plan9.bell-labs.com/magic/man2html/2/bio
>
> The compiler already understands about UTF and runes and so does
> libc (obviously):
>
>     http://plan9.bell-labs.com/sys/doc/comp.html
>
> Plan 9 is very close to unix, but it's not really.  A lot of grief
> gets caused by ape's heroic attempt to smash ANSI/POSIX into Plan 9.
>
> It only gets really nasty when you have to port something that uses
> the tty routines, sid's/pgrp's and signals -- most programs don't.

Oh, beat my English.  I was just curious if a native port, I mean
*without* APE and , hopefully, *without stdio.h* would be "a lot
easier" (when it means almost complete rewrite). Nevertheless,
although the program is very helpful for me, I must admit it is not
worth a rewrite: rather a complete redesign as an rc script using grep
and other standard utils.

++pac.

Re: [9fans] Announce: port

[Boyd Roberts]

"peter a. cejchan" wrote:
> Oh, beat my English.  I was just curious if a native port, I mean
> *without* APE and , hopefully, *without stdio.h* would be "a lot
> easier" (when it means almost complete rewrite).

More likely I misread what you wrote.

Re: [9fans] Announce: port

[forsyth]

here's a mkfile for rel.

</$objtype/mkfile
CC=pcc -c
LD=pcc

OFILES=\
	rel.$O\
	postfix.$O\
	lexicon.$O\
	searchfile.$O\
	relclose.$O\
	eval.$O\
	bmhsearch.$O\
	searchpath.$O\
	translit.$O\
	qsortlist.$O\
	uppercase.$O\
	memalloc.$O\
	message.$O\
	version.$O\

TARG=rel
BIN=$home/bin/$objtype

</sys/src/cmd/mkone
CFLAGS= -DFSE_CONTINUE -D_POSIX_SOURCE -D_BSD_EXTENSION -D_Plan9_