Posted by Steve on 09/24/07 00:12

"Shelly" <sheldonlg.news@asap-consult.com> wrote in message
news:13fcimcrurjrl0c@corp.supernews.com...
>
> "Steve" <no.one@example.com> wrote in message
> news:oyhJi.72$Qo7.69@newsfe02.lga...
>>
>> "Shelly" <sheldonlg.news@asap-consult.com> wrote in message
>> news:13fb0rcom0qur6e@corp.supernews.com...
>>> Here is a situation that I have to think out for a potential customer.
>>> Currently he receives about 150 emails a day with pdf attachments for
>>> orders. The format of the pdfs are all the same. Now he has to:
>>>
>>> 1 - look at his email
>>> 2 - open the pdf
>>> 3 - manually take the data from the pdf and enter it into an order
>>> processing mode and a database.
>>>
>>> This is taking so much of his time that he is considering hiring someone
>>> to do it.
>>>
>>> What I would like to be able to present him with is the following:
>>>
>>> 1 - Have all those emails go to a specified folder in his email
>>> 2 - Without opening the email, upload the attachment to a server
>>> 3 - Have an application that extracts the information from the pdf and
>>> then does what it has to do.
>>> 4 - Move the email to a second email folder (processed)
>>>
>>> I would like to have all this initiated with either a cron type job or
>>> via a "Go" button.
>>>
>>> Short of this ideal, I would have him look at an email in his reader
>>> and save the attachments to a directory. The "Go" button would upload
>>> button would then do the rest.
>>>
>>> There are also other compromises I can and probably will have to make.
>>> I posted the ideal.
>>> The main point is to cut the hyman time down considerably.
>>>
>>> Any suggestions?
>>
>> what os is the mail server on, and would you have access to it? if it's a
>> windows server, i'd use .net to tie into the onArrival (can't remember
>> the exact name...but...) event. from there, you can parse it for either
>> embedded content, or if it is a link, go out and grab it. if embedded,
>> it's a snap to unencode it. if the os is *nix, it depends on the mail
>> server, but usually it goes to file somewhere. a cron would work to scan
>> for new messages and kick off the grabbing of the pdf.
>
> Excelllent suggestions, Steve. I'll find out.

thanks, but unfortunately, that only covers steps 1 and 2. i've never tried
reading a pdf. ocr is pretty easy to work with (depending on who is
supplying the dll). anyway, i'm following this thread so i can mark the good
suggestions for step 3. let us know what you work out.

[Back to original message]