What is the best/fastest way to extract strings of text?
jimaultwins at yahoo.com
Tue Aug 2 11:33:40 EDT 2011
For Html, the returns have no meaning, so the first step I would take
replace cr with empty in textBlock
and make the text a single line...
but this may not be good, depending on the original textBlock
On Aug 2, 2011, at 1:01 AM, Keith Clarke wrote:
> The recipe I (learned here and) use with extracting specific HTML /
> XML elements is to get the specific target elements on their own
> line, remove the unwanted lines and then move the target string
> items in the remaining lines out into a separate variable -
> something like...
> 1. Get the target elements into their own line by prefixing the
> opening tag with return, using: replace "<#B>" with return & "<#B>"
> in theSource
> 2. Get the closing tag onto its own line by adding a return suffix,
> using: replace "<#E>" with "<#E>" & return in theSource
> 2. Remove the unwanted lines, (that lack the specific opening tag)
> using: filter theSource with "<#B>"
> 3. Delimit the line into items at the '>' character, using: set the
> item delimiters to numtochar(62)
> 4. Iterate through the list to extract the string, using:
> repeat for each line l in theSource
> put item 2 and return after theExtract
> end repeat
> 5. Clean-up the extract of any extra returns, using: filter
> theExtract without empty
> If (my pre-coffee brain worked) theExtract should contain the tagged
> strings in theSource.
More information about the Use-livecode