Hi Donald,
attached is the html-to-text engine I used in my %browser.r,
it's a little rough at some edges, but might keep you going ...
kind regards,
Ingo
Once upon a time Donald Dalley spoketh thus:
>
> Hi, Bo:
>
> I have a question about this e-mail. As you can see below, the line breaks are
> not right. The lines that have a single quote (") on it probably had a NEWLINE
> of some sort on the previous line. This causes an error when run, so what is
> really supposed to be between the two quote marks?
>
> I figured out how to get it to run OK, but, for my purposes, the retained text
> needs much better formatting - close to how the author wanted to format it.
> Using an optional arg for the max-linelength would also help. If someone knows
> how and is willing to do this, I would dearly like to see an improved version.
> Anyone who has ever used a good HTML stripper, such as the excellent HTTX
> (Amiga), knows how useful they can be, when called by other programs.
>
> Thanks for the seed, Bo!
>
-- Attached file included as plaintext by Listar --
REBOL [
Title: "HTML to Text Converter"
File: %html-to-text.r
Date: 2000-06-10
Author: "Ingo Hohmann"
Email: [EMAIL PROTECTED]
Site: http://www.2b1.de/
Rights: "(c) Ingo Hohmann"
Purpose: {Create text from html}
Comments: {
extracted from my browser.r
(which should be updated to current /View, btw.)
}
]
html: make object! [
help: {A html parser}
evaluate: false
read-error: none
skip: false
spaces: charset " ^-^/"
non-spaces: complement spaces
delimiters: charset { ^-^/="}
non-delimiters: complement delimiters
html-source: copy ""
get-html: func [][ return html-source ]
find-base: func [ url [url! file!] /local u2][
if #"/" = last url [ return url ]
if exists? u2: to-url rejoin [ url "/" ] [
return u2
]
first split-path url
]
conv-list: [ "&" "&" "<" "<" ">" ">" """ {"}
"ä" "�" "Ä" "�" "ö" "�" "Ö" "�" "ü" "�"
"Ü" "�" "ß" "�" " " " "]
clean: func [
{Converts html-entities to special-characters}
text [string!]
/local special entity
] [
foreach [special entity] conv-list [
replace/all text special entity
]
text
]
parse-tag: func [
{parses a tag, returns block of tag-name arguments}
tag /local tag-name tag-params] [
name-rule: [ some non-delimiters ]
param-rule: [
any spaces [
copy param-name name-rule (append tag-params param-name)
any spaces [
"=" any spaces [
{"} copy param-val to {"} skip |
{'} copy param-val to {'} skip |
copy param-val some non-delimiters skip
]
(append tag-params param-val) | (append tag-params true)
]
]
]
tag-params: copy []
parse/all tag [ copy tag-name name-rule any param-rule ]
compose [ (tag-name) (tag-params) ]
]
read: func [
{read url and return the page as ...}
url [url! file!]
/html "html source" /text "text" /links "link-list"
/local data txt lnk return-block
] [
return-block: copy []
either error? err: try [data: read url] [
read-error: disarm err
] [
read-error: none
if html [ append return-block data ]
if any [ txt links ] [
set [txt lnk] to-text url data
if text [ append return-block txt ]
if links [ append/only return-block lnk ]
]
print dir? url
]
]
to-text: func [
{Convert html to text, url is needed for handling of relative urls}
url [url! file!] html [string!]
/local elem txt links link lfd link-blk pos end-pos the-script script-funcs
] [
script-funcs: make object! [
print: func [val][ insert pos load/markup form join val newline ]
prin: func [val][ insert pos load/markup form val ]
]
url: find-base url
links: copy []
link-blk: copy []
html-source: copy html
html: load/markup html
txt: make string! 500
lfd: 0
parse html [
some [
set elem string! (
if not skip [
if 0 < length? trim/lines elem [
append txt rejoin [ elem " "]
]
]
) |
pos: set elem tag! (
elem: parse-tag elem
switch first elem [
"a" [
lfd: lfd + 1
append txt rejoin [ "(" lfd ")" ]
elem: select elem "href"
if elem [
if all [ not find elem "://" not find elem "mailto:"] [ ;
<a name="top"> ???
elem: rejoin [ url elem ]
]
append links compose [ (lfd) (elem)]
]
]
"img" [
either elem: select elem "alt" [
append txt rejoin [ "[" elem "]" ]
] [
append txt "[graphic]"
]
]
"p" [append txt "^/^/"]
"br" [append txt newline]
"hr" [append txt "^/------------------------------------^/" ]
"li" [append txt "^/* "]
"ul" [append txt newline]
"/ul" [append txt newline]
"ol" [append txt newline]
"/ol" [append txt newline]
"div" [append txt newline]
"/div" [append txt newline]
"blockquote" [append txt newline]
"/blockquote" [append txt newline]
"style" [skip: true]
"/style" [skip: false]
"pre" [
end-pos: find pos </pre>
append txt rejoin [ newline copy/part next pos end-pos ]
pos: end-pos
]
"script" [
either all [ evaluate "rebol" = select elem "language" ] [
end-pos: find pos </script>
the-script: copy/part next pos end-pos
remove/part pos next end-pos
if error? err: try [
do bind load rejoin the-script in script-funcs 'print
] [
;inform layout [
; subtitle red "Error in script !"
; text mold disarm err
;]
print ["Error in Script: ^/" mold disarm err]
]
pos: back pos
][
skip: true
]
] ; do bind load rejoin n in t 'print
"/script" [skip: false]
"/title" [append txt "^/^/"]
]
pos: next pos
)
:pos
]
]
txt: clean txt
append txt "^/^/^/The links:^/-------^/^/"
foreach [lfd link] links [ append txt rejoin [ lfd " " link newline]]
foreach [lfd link] links [ if not find link "mailto:" [append link-blk rejoin
[ lfd " " link]]]
return compose/deep [ (txt) [(link-blk)]]
]
]
--
To unsubscribe from this list, please send an email to
[EMAIL PROTECTED] with "unsubscribe" in the
subject, without the quotes.
