Ideas.org

Automated cluing

Implementation 1

1. Provide an option to create an answer index.
   • two files: across.idx and down.idx
   • each has the form: clue-number <tab> answer
2. Create a program that, given an idx file, calls out to various sites, collects their output and stores it in a file named for the answer (e.g., across/23_aloe)
   • http://crosswordtracker.com/answer/aloe/?search_redirect=True
   • https://www.merriam-webster.com/dictionary/aloe
   • https://www.google.com/search?q=define+aloe
   • https://translate.google.com/?source=osdd#auto|auto|aloe
3. Create MacOS automation routines to copy the selected text and insert it into Phil (???)

Implementation 2

• pop up windows with this info

Highlight similar words

• iterate through all answers. If share the same root, assign them the same class (e.g., { common_root₁, common_root₂, … common_root₁₀ } )
• rationale: this will point out where the same (or substantially similar) words are used, e.g.,
  • “car” and “car” or
  • “car” and “cars” or
  • “car” and “cartoon” or
  • “art” and “cartoon”
• This will be helpful to point out where the same word was used or a trivially different form.
• This may be intentional or unintentional but it’s good to inform the user.