Conifg examples required for use scenario

[sbleitner]

Hi,

I'm a basic user, so I find it difficult to create a config file for anything more than just a basic use scenario.

Right now I'm trying to use this script for the following scenario:

• Use a custom search strategy to find supplier invoices by searching for the keywords {invoice, order, proforma} in the subject, body, or attachments.
• Exclude any emails that are in Trash, Drafts, Spam, or Sent folders.
• For emails with PDF attachments, save them to Google Drive, organised by the month in the email or invoice (e.g. “March 2025”, “April 2025”).
• If the attachment is in a different format (.XLSX, .DOCX, etc.), convert it to PDF before saving it to the folder.
• If there’s no attachment, then export the email body as a PDF and save it in the correct folder.
• Once an email has been processed, apply the Gmail label: "invoices processed" and archive the thread.

If anyone has a similar use scenario, would you kindly share this with the community so I can tweak it to my own requirements?

TIA
S

[ahochsteger]

Hi @sbleitner,
I'm not sure, if all of your scenario es exactly possible as you described it but here is my feedback about what might be possible and what not:

• Searching for keywords in the subject or body of an email or the filename of an attachment is possible using the query ({invoice order proforma} OR filename:{invoice order proforma}) (see https://support.google.com/mail/answer/7190?hl=en for syntax details) but I'm not aware that you can search for emails containing certain words within the attachment content, if that's what you meant.
• Excluding folders is possible with -in:trash -in:drafts -in:spam -in:sent in the thread query (globally or as part of every thread query)
• I'd define different thread queries for each of the attachment types:
  • PDF attachments: (has:attachment AND filename:pdf)
  • Excel/Word documents: (has:attachment AND filename:{docx xlsx})
  • No attachment: -has:attachment
• As long as every thread consists of just a single email then mark processed emails with the label invoices processed is possible. If multiple emails may exist per thread than you may run into issues since it's not possible to attach labels to individual emails but just to the whole thread.

Here's a raw draft of the config you will have to refine and test to fully match you complex scenario:

{
  "description": "Example for https://github.com/ahochsteger/gmail-processor/discussions/513",
  "settings": {
    "markProcessedMethod": "add-label",
    "markProcessedLabel": "invoices processed"
  },
  "global": {
    "thread": {
      "match": {
        "query": "-in:trash -in:drafts -in:spam -in:sent -label:\"invoices processed\" ({invoice order proforma} OR filename:{invoice order proforma})"
      }
    }
  },
  "threads": [
    {
      "description": "Store PDF attachments",
      "match": {
        "query": "(has:attachment AND filename:pdf)"
      },
      "messages": [
        {
          "attachments": [
            {
              "match": {
                "name": "\\.pdf$"
              },
              "actions": [
                {
                  "name": "attachment.store",
                  "args": {
                    "location": "{{message.date|formatDate('MMMM yyyy')}}/{{attachment.name}}",
                    "conflictStrategy": "keep"
                  }
                }
              ]
            }
          ]
        }
      ]
    },
    {
      "description": "Store non-PDF attachments converted to PDF",
      "match": {
        "query": "(has:attachment AND filename:{docx xlsx})"
      },
      "messages": [
        {
          "attachments": [
            {
              "match": {
                "name": "\\.(docx|xlsx)$"
              },
              "actions": [
                {
                  "name": "attachment.store",
                  "args": {
                    "location": "{{message.date|formatDate('MMMM yyyy')}}/{{attachment.name}}",
                    "conflictStrategy": "keep"
                  }
                }
              ]
            }
          ]
        }
      ]
    },
    {
      "description": "Export emails without attachments to PDF",
      "match": {
        "query": "-has:attachment"
      },
      "messages": [
        {
          "actions": [
            {
              "name": "message.exportAsPdf",
              "args": {
                "location": "{{message.date|formatDate('MMMM yyyy')}}/{{message.subject}}.pdf",
                "conflictStrategy": "keep"
              }
            }
          ]
        }
      ]
    }
  ]
}

You can paste this config into the playground to further refine it. Afterwards you can copy the full JavaScript code using the JS icon and paste it into your Google Apps Script file.
If you have not already done so - make sure to have done the Getting Started Guide to initially setup Gmail Processor, in case it's the first time using it.
Let me know, if this helps or if you need further assistance.

[sbleitner]

Hi,

Thanks so much for this - I’ve already put it to good use with only a few minor hiccups!

The first issue is that it also picks up and saves promotional emails that happen to contain the word order or invoice in the body. To try and filter these out, I added AND category:primary to the search query, will see how it goes.

The next issue I’ve encountered is with threads. Even though my Gmail is not set to conversation view, if an email matches the search query and would normally have been part of a thread, it seems the script treats it as a thread anyway. As a result, the entire thread is saved again in Drive every time a new related email arrives. I’m not sure how best to address this.

Lastly, I noticed that attachments are only saved if the file extension is in lowercase. If the file extension contains uppercase letters, the file isn’t saved to Drive - although the label is still correctly applied.

Thanks again for all your help - it’s been a massive help getting me this far!

Best,
S

[ahochsteger]

@sbleitner glad to see that it worked for you. :-)

The fix for promotional emails seems a good fit since you can build on the trainable auto-categorization features of Gmail.

Here are some advices to your remaining issues:

To correctly process emails that are attached to a thread you need the possibility to mark individual emails as processed, which is not possible with labels, since they are attached on the thread-level. A solution would be to use the read status that can be set for each email individually. Therefore you have to switch the markProcessedMethod from add-label to mark-read and remove -label:\"invoices processed\" from the global query (which was redundant anyway, since it had been added by setting the markProcessedMethod to add-label anyway).
Here's a simple example that shows how to use it:
https://ahochsteger.github.io/gmail-processor/docs/examples/basics/simple

Concerning the unprocessed uppercase letters in file extensions I suggest to adjust the regex with the file names (e.g. "\\.(docx|xlsx)$") to be prefixed with (?i) to do a case-insensitive match so that the full regex reads "(?i)\\.(docx|xlsx)$".
Here's an example that shows how to use flags with regular expressions:
https://ahochsteger.github.io/gmail-processor/docs/examples/advanced/regularExpressions

I hope this helps you to fix your remaining issues.
Andreas