using the EmailClient as starting solution...

Want to be able to get email from account - download attachments push them to supercontainer - use scribe to extract keywords from attachments (txt, xls, pdf, doc ) then process entire record ( plain text / html ) and attachment results with some regular expressions to pull out known identifiers (dates/ emails / tracking numbers / order numbers etc. )

I have most if it working - haven't yet tried to run this server side.

End goal is the users log in and don't have to fetch mail the server will do that for them

they just process the records so they don't have to sit thru progress bars.

I have noticed the two bottlenecks is pushing to SC / but totally acceptable. The second is the processing of keywords list using scribe - It is pushing over a minute to get the text from a PDF.

With out that part I cant throw my RegEx statements at the email much of the results are dependent on the content of the PDF.

Appreciate thoughts to speed this up - and any recommendations to streamline the process.

Stephen

Are you using the latest version of scribe? It should not be taking more than a second or two to pull out the data from the PDF, unless you have a very, very large pdf document.

Could you send me one of the documents that is taking so long for you to extract text from?

just emailed you the PDF - I was comparing over 500 PDF attachments and some much larger this one only 90k.

It must be the software that they use to generate the PDF is doing something strange.