OCR, S3 backup questions.

Discussion in 'General Discussion' started by Jason, Sep 10, 2009.

  1. qilin

    qilin Member

    Is this fax inbox the same as the watch folder? I only had 4 PDF files. Also, as I am running the sync-faxboxes, S3 back up stopped.
  2. Graham

    Graham Developer Staff Member

    I wonder if this is a limitation of using XP instead of 2003 as the server OS. It sounds as though you are not able to sustain more than 2-3 obdc connections at the same time.

    Have you considered installing a trial version of Windows Server 2008 ?

    Also check the faxinq table using IBOconsole and see if there are any files present.
  3. qilin

    qilin Member

    I stopped sync-faxboxes to let S3 back up finish 1500 files left.

    faxinq table is empty.

    Why does sync-faxboxes OCR all the exiting 15,000 files? I only need it to work on new scan/faxes.
  4. Graham

    Graham Developer Staff Member

    That's what I need .. a single app that does ocr of results and faxes.

    Guess I could create a flag that stops that behavior.

    Try running sync-faxboxes again to see if it picks up the files. You can stop it after 10 mins or so.

  5. qilin

    qilin Member

    Add an option whether to just ocr the new scans/faxes or include the exiting files?
    No, still not picking up. Do I have to run client side at the same timefor it to pick up?
  6. Graham

    Graham Developer Staff Member

    FaxImport scans the faxes folder and uploads them to S3 as encrypted files.

    It then sends a message after each file to the SQS queue you specified. These messages sit there in the queue for up to 4 days before they are automatically deleted.

    Sync-faxboxes reads the queue messages, OCRs these files as necessary, and then deletes the messages.

    Both programs operate asynchronously.

    If it is not working, then

    1. the queue does not exist ( check the server logs to see if messages are being picked up )

    2. the messages have already expired

    3. some other error

    I'd suggest moving the files out of exported back into your fax directory, and run FaxImport again to recreate the messages. Once it has sent all the messages, wait 15 mins, and then start up the Sync-faxboxes. And look at the logs .

  7. Graham

    Graham Developer Staff Member

    It can't ocr the latest as the timestamp is not in the files database.

    But i will pause it after 10 files and so it will then rescan for messages in the queue after that.
  8. qilin

    qilin Member

    Worked this time! Guess the previous ones expired. This is very cool.

    Will let my secretary try it later and see if she likes it.
  9. Graham

    Graham Developer Staff Member

    I think the workflow is

    1. you scan stuff to a folder and direct incoming faxes to that folder

    2. your secretary looks in the fax inbox. She allocates the fax/scan to a patient, and sets the type of document.

    3. she decides if it needs to be seen by you and forwards it to your inbox. Oherwise she just files it eg. office documents.

    4. and if she can't cope, time to get another secretary.
  10. qilin

    qilin Member

    I now have snappyfax print to paperport image printer using searchable PDF, so all the scans and faxes are OCR'ed PDF files except the intake sheets and rx copies which are hand filled ones. For these latter files, will have to use [view fax] to see who they belong to. I wonder if it'll help to display the originalfile name in the fax list so if I name the file with the patient's name like "adam smith intake", I don't need to open the file to see what it is since OCR doesn't help in this type of files. I do see you assign new internal names to the scanes/faxes.
    Edit: Nevermind, guess click [view fax] is faster than renaming a file.

    Can you make [To:] default to the patient's PHCP once a patient is selected?

    Again, have to figure out how to have S3 backup work at the same time. Maybe I just have to shut down faximport over the weekend and let S3 backup run once a week?
  11. Graham

    Graham Developer Staff Member

    Didn't know paperport had OCR .. interesting.

    Very odd that you can't run both apps at once .. as I mentioned, it may be an XP limitation.
  12. Jason

    Jason Developer / Handyman Staff Member

    When you add OmniPage to PaperPort, you get fairly robust OCR.
  13. qilin

    qilin Member

    S3 backup stopped at 34 files left, tried restarting the server and PC, still not going any further
  14. Graham

    Graham Developer Staff Member

    that likely means that you have 34 files missing from the cache-listener.

    delete the S3 log, start it up and then examine the logs .. it will tell you which ones are missing.
  15. qilin

    qilin Member

    Ok, these are miss-filed files that got lost during filing process, meaning they are uploaded to the server but did not get assigned to the patient, they have files names start with "-", not "patient's ID-". So I removed these files manually form cache-listener. Should I just ignore this? Or should I fix this?
  16. Graham

    Graham Developer Staff Member

    Try copying a few of these files to your watch directory and see if FaxImport will fix the issue for you.

    Hopefully it will save them to S3, put them into your inbox and you can allocate them correctly.
  17. Graham

    Graham Developer Staff Member

    The new fax inbox handling means that I can file a result using the mouse only.

    1. Double click on the surname in the OCR, then click on the first "?". This brings up the patient or you can use the mouse to select the correct one.

    2. The physician auto-selects

    3. Click on the beginning of the test date in the ocr, and then click on the second "?". This brings up the date requester generally on the correct date. Click on it to accept.

    4. Click on the drop down for test type

    5. Then file or send it to the inbox.
  18. qilin

    qilin Member

    These files no longer exist. When I realized a mis-filed scan, I went to the cache-listener and move the file to re-file. I thengot a sql error since the server thought the file was already there. I just cropped the scan a bit in paperport and uploaded againto server. So can I just delete the related item from the database?
    Excellent idea.
  19. Jason

    Jason Developer / Handyman Staff Member

    [quote user="Graham"]

    The new fax inbox handling means that I can file a result using the mouse only.


    Mouse only or Keyboard only is a good thing.

    Sounds pretty quick !

    Doing a good job of this makes such good business sense for Synapse users.

    My scanner girl has moved on to Nursing School.

    We are overloaded.

  20. Graham

    Graham Developer Staff Member

    For purposes of referential integrity you should not unless you can also delete all references in the results table to these files.

Share This Page