Search your Paperpile library with Alfred

• Setting up
• Basic Usage
• Output
• Known Issues
• Acknowledgments
• Changelog
• Feedback

1. Download the workflow from Github and double-click to install paperpAlfred

2. Download your Paperpile library

   • in Paperpile, cogwheel > Settings > Export > Export to JSON

3. Set the PAPLIBRARY path

• Copy the library file path to the clipboard

  • In Finder, right-click the file, press option (⌥), the select Copy [FileName] as Pathname
  • Open the 'Configure Workflow' window in paperpAlfred preferences
  • set path to library as the Paperpile Library value

4. Optional: Set Paperpile Path (Path to Paperpile in Google Drive)

   • this will allow to open PDFs in the system viewer

5. Optional: change the max number of results returned (default: 99)

   • Set the MAXRESULTS value in the 'Configure Workflow' window in paperpAlfred
   • Set the Key Prefix (citekey prefix) in the 'Configure Workflow' window in paperpAlfred

6. Optional: Setup hotkeys to launch

   • main search
   • filter by label
   • filter by folder
   • filter by type

• launch paperpAlfred by typing ppp or using an optional hotkey
• words or names will be searched across all fields

• entering ppl (or optional hotkey) will show a list of labels, the number of items in each label, and the count of item types
• select the label by pressing return. PaperpAlfred will now search within that label
  • option-return (⌥⏎) will open the label in Paperpile

• entering ppf (or optional hotkey) will show a list of folders, the number of items in each folder, and the count of item types
• select the folder by pressing return. PaperpAlfred will now search within that folder.
  • option-return (⌥⏎) will open the folder in Paperpile
• Note: Added feature ✅ the Paperpile web interface does not allow to search by folder

• entering ppty will show a list of publication types and the number of items in each
• select the folder by pressing return. paperpAlfred will now search within that item type
  • shift-return (⇧⏎) will open the folder in Paperpile

• enter field:, where field is any of the fields below. Example: year:2022
  • title
  • abstract
  • citekey
  • first first author
  • last last author
  • journal
  • folder
  • label
  • pmid
  • year
  • type publication type

• Alfred returns the top 99 results, numbered. The max number of results returned can be set in Preferences (see Setting up)
• results are sorted by year (most recent first). Adding --a to the query will invert this order.
• The main text will return the title.
• the subtext will return the following:
  • record count/total result count
  • 📜if a PDF is available
  • short ID string (first-last author, journal, year)
  • associated labels

Once the right item is found, the user has seven options to act on it:

1. return (⏎) will open the PDF in the system viewer (Preview for most users), provided that the PAPPATH has been set (see Setting up). Rarely this might not work (see Known Issues)
2. shift-return (⇧⏎) will show the abstract and copy it to clipboard
3. control-return (⌃⏎) will show the complete reference and copy it to the clipboard
4. option-return (⌥⏎) will copy a short reference (First-last author, journal, year, PMID) to the clipboard
5. command-return (⌘⏎) will copy the citation key to the clipboard
6. command-shift-return (⌘⇧⏎) will open the PDF in Google drive
7. command-option-return (⌘⌥⏎) will open the record in Paperpile

• incomplete records will not be imported
• special characters (e.g. ü) will need to be entered in order to match the record
• File opening is currently using a name search. The Paperpile file naming logic is not entirely clear to me, the folder structure is deprecated and there might be a small number of cases where the PDF might not be retrievable via paperpAlfred. Google drive view should still work in these cases.
• Currently tested mainly with research papers, reviews etc. There might be untested use cases for other types of publications.
• label and folder search in the main window (i.e. using label: and folder:) will not be exact matches (e.g. AD will also return GWAS_AD). Match will be exact when starting from folder and label window.

• Dean Jackson for their incredible help on the Alfred mailing list and for creating alfred-index-demo, and other scripts used as templates for this workflow.
• Jirka from Paperpile for support on the path-to-file issue
• Alain T, StackExchange user:5237560 (https://nebularena.wordpress.com) for help with a Python script

• 12-04-2022: version 2.1 (Alfred 5)
• 03-15-2022: version 2.0 (Python3, removed dependencies)
• 03-17-2021: version 1.0