Scanning documents from iOS (also works on Android)

[s-oliver]

Since I don't have an actual document scanner, I scan most of my day to day documents with my iPhone and like to share my workflow - maybe it helps you or maybe you have better ideas.

Step one is scanning documents as quickly as possible. iOS does have built in capabilities for this, but they are not very convenient, so I opted for a 3rd party app called "Genius Scan" (also available for Android). You do require the plus version (8,99€ at the time of writing) for all share options, but once configured it allows this very quick workflow:

• Open app and hit the large "+" button
• Scan the document
• You're done. It supports auto export to a preconfigured destination where all scanned documents are put automatically, and you can even have the documents deleted immediately after the successful export so they don't clog your phone.
  In the scanning process you MAY change every page or the entire document, though I never had to touch the automatic results so far, so this allows me to scan my documents with minimal overhead.

Unfortunately Genius Scan does not support SFTP or FTPS (only the unencrypted FTP) and otherwise almost exclusively cloud services (if I wanted my documents in the cloud, I would not self host paperless..), the last option for encrypted sharing to my paperless server is WebDAV. Of course my documents contain sensitive information, so whenever they're sent over the wire, I'd like that connection to be secure.
Getting the WebDAV service ready is as simple as dropping a docker-compose.override.yml in your paperless working directory (next to your docker-compose.yml) - or modify your existing one (I have all my modifications of the default services in that file so I don't have to reapply changes every time I upgrade paperless-ng)

version: '3.4'
services:
  webdav:
    image: bytemark/webdav
    restart: always
    environment:
      AUTH_TYPE: Basic
      USERNAME: user
      PASSWORD: a-very-long-and-secure-password
    volumes:
      - ./consume:/var/lib/dav

Now a docker-compose up -d will also start a webdav server next to the other paperless services and expose the "consume" directory. What remains is encryption (see docker hub of bytemark/webdav for their instructions). Since I am running the jwilder nginx-proxy with the let's encrypt companion anyway, I just use it to expose the webdav through a valid and free ssl certificate. So my service descriptions looks like this:

version: '3.4'
services:
  webdav:
    image: bytemark/webdav
    restart: always
    environment:
      AUTH_TYPE: Basic
      USERNAME: user
      PASSWORD: a-very-long-and-secure-password
      VIRTUAL_PORT: 80
      VIRTUAL_HOST: docdrop.my.domain.com
      LETSENCRYPT_HOST: docdrop.my.domain.com
    volumes:
      - ./consume:/var/lib/dav
    networks:
      - nginx-proxy_default
networks:
    nginx-proxy_default:
        external: true

The nginx-proxy_default network is the one that is created by my nginx-proxy service, you may have to adapt it. Having the jwilder nginx-proxy in place already, settings all this up took less than 15 minutes and that includes the download and configuration of the app.
That WebDAV is configured as auto export target in Genius Scan and works very well so far. You could also use a self signed certificate and tell Genius Scan to accept that as well.

Genius Scan also supports OCR, but I have not even tried it for two simple reasons:

• I'm happy with the results of the paperless OCR
• I'd rather have that process run async on my server than synchronously in my scanning workflow and eating up my battery

Why I don't use iOS default tools - every method as major drawbacks:

• Open notes, open any note, tap photo icon, tap "scan document". After the scan, tap to open, use the share button to send via mail / airdrop / ... => many steps and no direct way to forward into paperless, since iOS mails always use inline attachments => no go
• slightly quicker: Open files, navigate to a folder, tap "..." and "scan document", share after scanning. Same sharing problem as the "notes" method above. Also, you end up having all the documents on the phone and need to clean up every once in a while (or delete it straight after sharing)
• In the mails app you can directly scan documents as attachments, but still have to type your mail address every time and possibly a subject if your paperless email filter is configured so, and eventually you end up with the issue, that iOS attachments are always inline => also no go

How a

[jonaswinkler]

How a

Awesome write-up so far, but I feel like something is missing :)

[s-oliver]

Oh right, weird that it was cut off.

I just wanted to ask how other people ingest their paper documents when there's no scanner hardware available.

[shamoon]

Yea great write-up, hadn’t thought about implementing WebDAV. I use a mix of browser upload and email, my wife especially finds the email easy.

Btw, consensus seems iOS / macOS users would benefit from some mechanism to import inline attachments from emails so that feature is slated for a future release: #153 (comment) thanks to Jonas’ responsiveness

[s-oliver]

Definitely, with that change the additional WebDAV server becomes obsolete (Genius Scan also supports sending the document via email as automated export) IF you trust your email provider and make sure its also encrypted all the way. If you send emails across two different providers, it may be sent plain text, though this should not be much of an issue if you're emailing yourself.

I guess now that it's set up, I'll keep the WebDAV for scanning, but use inline attachment when I receive digital documents and forward them to my paperless trigger inbox - great feature indeed!
WebDAV is quite handy since it's widely supported and easy to mount on all OSes, I think it's always quicker to use than the browser upload. The main reason I don't use the browser upload so much is that I can use it only at home since my paperless instance is not exposed to the WWW - only the webdav is.

[jonaswinkler]

You know, adding documents from email attachments is already possible. It's just that some mail clients tend to not attach files the way they're supposed to do according to specification. ("inline" attachments are what you usually see when a mail has embedded images, but some clients use that for pdf documents as well instead of marking these as proper attachments.) That's what he was referring to.