[jpstewart]

BTW: What is awf ?

It is kind of a minimalist phony roff implementation that comes with IRIX to convert unformatted man page source to plain text.  It's nowhere near as full featured as the real deal which only comes in the Documenter's Workbench add-on.  It is almost certainly installed on your IRIX boxes, so a simple 'man awf' will tell you more about it.  I think it stands for 'amazingly workable formatter' but the similarity to 'awful' is no accident, IIRC.

[jwhat]

Hi jpstewart,

thanks for that,I have Documenter's WorkBench installed so have awf.

I see this just takes troff inputs and generates (from man):

"The only output format supported by awf, in its distributed form, is that
appropriate to a dumb terminal, using overprinting for italics (via
underlining) and bold. The special characters are printed as some vague
approximation (it's sometimes very vague) to their correct appearance."

Believe it or not that is very helpful as one of the things I found was that the Italics has been converted to style == Black background/White text by aha, which I changed to Style == Bold/Underline.

As only a small number the man pages are provided as original troff files, having the ability to turn the "awf" dumb terminal output into HTML is helpful ;-)

BTW: installing Documenter's Workbench was a nightmare in terms of getting the dependencies resolved with inst. I think it is worth of its own set of IRIX wikipedia entries..

Cheers from Oz,


jwhat/John

[jwhat]

Hi SGI'ers & IRIX'er,

I have a "not half bad" working set of IRIX man page up and going.

This is currently just static pages, as I have put effort into the zcat -> html conversion, but it provides a reasonable working example.
Adding an cgi script or even PHP layer on top of converted pages would not be too hard.

Start from here: https://help.graphica.com.au/irix-6.5.30/man/1/sh
- where 1 == section sh == page

Currently this is 12,365 Man Pages of which 526 pages are in section U1 which is my bucket for pages that I did not find the page(section) header correctly.

I think I now have good strategy to deal with the final set of 526.

EDIT: Now have all pages correctly cataloged, so U1 tree is empty.

Also I have not put in the command alias links yet, but again this is pretty easy.

I have only looked a small number of the pages so far and traversed the generated cross pages links and its not looking to bad, but cannot testify to quality of all the html conversions.

Cheers from Oz,


jwhat/John

[jpstewart]

Nicely done!  I only browsed a few man pages and a couple of the indexes, but so far you're off to a great start!

[jwhat]

Thanks JPS.

I have now done all of my IRIX man files.
There are 3 types:
- catman - pre-rendered for ANSI terminal (the main set and the hard ones to render)
- man - nroff source man pages which can be rendered and passed through man2html processor (trivial)
- html - sgi delivered some pagea as HTML so just need to uncompress these (trivial)

Total unique pages is now: 12393, so not much greater than before as the nroff page set overlaps with the catman set.

Anyhow that is it, on this for a while. I have got what I wanted, which was to be able to access IRIX doco via web, without having to power up my SGI Box/es.

As icing I will have a look at what can be done by pointing search engine at the page set.

#EDIT #1 - Added indices by: Page/Section & Section/Page
#EDIT #2 - Added simple topic search and extra UNIX/BSD versions for good measure

Cheers from Oz,

jwhat/John