|Parsing fixed-layout screens firstname.lastname@example.org (1995-07-26)|
|Re: Parsing fixed-layout screens email@example.com (1995-07-31)|
|From:||firstname.lastname@example.org (Roland Paterson-Jones)|
|Keywords:||parse, question, comment|
|Organization:||University of Cape Town|
|Date:||Wed, 26 Jul 1995 13:04:44 GMT|
I can't find any newsgroups dedicated to parsing, so that's why this is here.
Does anyone know of any tools or techniques for parsing forms-type
documents/screens, where certain elements at specific locations are required
(to be parsed)?
This might have the same answer: Are there any established techniques for
parsing column-based languages such as fortran/pascal?
We want to do screen-scraping of data from text terminals in as intelligent a
manner as possible. Hand-parsing would not be too difficult, but it would be
neater to have a higher-level specification tool. BNF appears to be no good,
since there is no concept of layout, and it can easily become confused by
missing columns on a line.
Thanks to all bites
[Fortran compilers handle the fixed column aspects of Fortran in an utterly
ad-hoc way. I'd think it'd be possible to do some sort of table-driven
thing that turns stuff at known positions into lexemes which you could then
treat as a data stream to parse. -John]
Return to the
Search the comp.compilers archives again.