best approach to extract from long string tables and other info

[GS]

I need to extract sections out of a long string of about 5 to 10 KB, change
any date format of dd Mmm yyyy to yyyy-mm-dd, then further from each section
extract columns of tables.


what is the best approach in using regex for this? I can see match and
replace the dates, extract section with regex, and then for each section
extract again with the right regex the tables....

 

[rowe_newsgroups]

I need to extract sections out of a long string of about 5 to 10 KB, change
any date format of dd Mmm yyyy to yyyy-mm-dd, then further from each section
extract columns of tables.

what is the best approach in using regex for this? I can see match and
replace the dates, extract section with regex, and then for each section
extract again with the right regex the tables....

Besides from Regex, you might be able to use a LINQ query to parse the
string (if you're using the 3.0 or greater framework). I've only
started to play around with LINQ, but I know Jon Skeet has written
about how LINQ can have much better performance than Regex, but it
will depend on structure of the string and other things.

Thanks,

Seth Rowe [MVP]

 

[GS]

thanks
With linq can I still use table driven approach?

that is can LINQ query be stored in table and retrieved dynamically to use?
There is this requirement that the parse logic be table driven so that the
entire app does not need to compiled every time when parse logic change.
Consequently the app is more adaptable to different user needs and external
change without massive version of hardwired code support.

The targets are at least Athlon 64 +3800 2 GHz if not Core 2 2.2 GHz.
Users are happy to put up with 2 or 3 seconds processing time as it would
take an user quite a few minutes of copy, paste, edit, import