Three postings on ESIS - how to learn about it via SGMLS.pm and Python. See the main ESIS entry in the SGML Web Page for more information. Ingo Macherius, T. Kurt Bond, and Jacco van Ossenbruggen.
From [email protected] Thu Apr 17 09:03:48 1997 Subject: Re: ESIS To: [email protected] Date: Thu, 17 Apr 1997 15:01:10 +0200 (MET DST) From: Ingo Macherius <[email protected]> | Choi Wonseog wrote : | | What is ESIS? | Thanks. | ... ... I personally learned much about ESIS by feeding the ESIS output produced by nsgmls to Perl scripts based on David Megginsons SGMLS.pm class library. Every ESIS event is represented by a Perl5 object there. So the documentation can be considered kind of an ESIS documentation. Please see: http://www.uottawa.ca/~dmeggins/SGMLSpm/sgmlspm.html ++im -- Snail : Ingo Macherius // L'Aigler Platz 4 // D-38678 Clausthal-Zellerfeld Mail : [email protected] WWW: http://www.tu-clausthal.de/~inim/ Information!=Knowledge!=Wisdom!=Truth!=Beauty!=Love!=Music==BEST (Frank Zappa) -- ===================================================================== Date: Thu, 17 Apr 1997 23:25:14 -0400 From: "T. Kurt Bond" <[email protected]> Subject: Re: ESIS In-reply-to: message from Svante Kleist on Thu, 17 Apr 1997 18:41:48 +0200 (MET DST) Sender: [email protected] To: [email protected] >Does anyone know if something similar to "SGMLS.pm" >has been implemented in Python? > >Svante Kleist, Stockholm, Sweden At one point I wrote a simple translation of SGMLS.pm and company in Python. It was more of a learning exercise (for both Python and SGML transforming) than a real product, so I never properly documented it, and I'm not sure it completely handled everything that SGMLS.pm did, and I never used it for any major work. It did seem to work ok on the smaller things on which I tried it. The translation from Perl was reasonably straight-forward. I'm sure a real Python coder could easily improve on it. I suppose if someone was interested in it I could package it up and put it somewhere, but it would need some hacking to be really useful. I later decided to reimplement SGMLS.pm in Objective Caml (a dialect of ML that adds object-oriented features to ML's normal static type inferencing) and found *that* a much more interesting problem; approaches that are reasonable in Perl and Python are *not* necessarily appropriate in a dialect of ML. (:-) This one I'm still working on, slowly, and I've used it in several small projects. -- T. Kurt Bond, [email protected] (and formerly [email protected]) -- ===================================================================== Date: Fri, 18 Apr 1997 11:04:16 +0200 From: JR van Ossenbruggen <[email protected]> Subject: Re: ESIS In-reply-to: Svante Kleist <[email protected]> <"Re: ESIS"@SIL.ORG> (Apr 17, 6:41pm) Sender: [email protected] To: [email protected] On Apr 17, 6:41pm, Svante Kleist wrote: >Subject: Re: ESIS > >On Thu, 17 Apr 1997, Ingo Macherius wrote: > >: I personally learned much about ESIS by feeding the ESIS output >: produced by nsgmls to Perl scripts based on David Megginsons >: SGMLS.pm class library. Every ESIS event is represented by a Perl5 >: object there. So the documentation can be considered kind of an >: ESIS documentation. > >Does anyone know if something similar to "SGMLS.pm" >has been implemented in Python? > >Svante Kleist, Stockholm, Sweden I've never seen SGMLS.pm but I did a quick and dirty hack in Python to read in an ESIS file and store it in a python object. My read prodedure reads in the complete ESIS file and returns the root element. This is a Python object which has an dictionary with (stringyfied) attributes and a dictionary with children. I used it to generate bibtex files from an SGML file. It is an undocumented Q&D hack, but if you want to, feel free to have a look and steel anything you can use... The python sources and a small example are on http://dejavu.cs.vu.nl/~jrvosse/Python/ Jacco -- Jacco van Ossenbruggen Faculty of Mathematics and Computer Science mailto:[email protected] Vrije Universiteit, de Boelelaan 1081a http://www.cs.vu.nl/~jrvosse/ 1081 HV Amsterdam (fax: +31 20 44 47653) --