Announcement

**Paul Franks** · 7 Jun 2008, 09:24 PM

Originally posted by paul d purvis View Post

it would be nice have something that would spit out text from html but i just do not see anything.

Well, you could start by automating IE, using the "SaveAs" method, which can save as text.

There are dozens of "html2txt" converters out there, just Google for them. Most use regular expressions to keep the coding to a minimum (hint, PB has some nice regex functions).

If your html input is valid XML (it ought to be in the year 2008), a simple XSL transform will do the job.

There's even an online converter, courtesy of www.w3.org:

html2txt: online service to convert web pages to plain text

http://cgi.w3.org/cgi-bin/html2txt

Here's a pretty good list of converters to several other formats:

Converting HTML to other formats

http://www.w3.org/Tools/html2things.html

I compiled one of them (C source, using gcc under MSYS). It's a filter, so gets its input from STDIN and its output goes to STDOUT.

To use, simply

C:> type mypage.html | html2setext.exe > mypage.txt

Attached Files

html2setext.zip (8.2 KB, 28 views)

**Donald Darden** · 8 Jun 2008, 03:41 PM

A very simple filter would look for opening tag flags (<) and trailing tag flags (>) and simply ignore everything in between. However, quoted text (surrounded by double quote symbols (") are generally not to be interpreted, so if you find a "<" or ">" inside a quoted section, you would ignore them.

the code to do both these things is very simple. We define two variables which
we call tag and quote to track the current status of these marks

Code:

tag=0
quote=0
FOR a=1 TO LEN(httpstr$)
  aa$=MID$(httpstr$,a,1)   'deal with one character at a time
  SELECT CASE aa$
  CASE CHR$(34)    'double quote
     IF NOT tag THEN quote = NOT quote  'ignore insides of tags
  CASE "<"     'possible start of tag
     IF NOT quote THEN INCR tag  'this allows for embedded tags
   CASE ">"
     IF NOT quote AND tab>0 THEN DECR tag  'again, allows for embedded tags
   END SELECT
   IF NOT tag THEN PRINT aa$;  'anything not in a tag is fair game
NEXT

This is certainly not going to deal with every situation by any means.
If you want to recover embedded URLs, you have to look for certain types of
tags and process them. If the HTTP page has links to other documents, lists,
tables, or whatever, you have to search them out, retrieve them, process them, and embed them in the right place. But in its simplest form, this is how
you can remove http tags from a straightforward document and retain just the text.

Announcement

html to text

html to text

Comment

Comment