jjst/songbook-generator

AI Summary: Implement a GitHub Action to automatically build and deploy a preview version of the Songbook Generator's frontend to GitHub Pages for each pull request, creating a unique URL (e.g., `/preview/pr-<PR_NUMBER>/`) and posting a comment with the link. The action should also clean up old preview deployments after a PR is merged.

AI Summary: Refactor the existing synchronous songbook generation backend (Google Cloud Function) into an asynchronous architecture using Cloud Run, Pub/Sub, and Firestore. This involves creating a job submission endpoint, a background worker to process generation tasks, and a data store for tracking job status and storing generated PDFs in Google Cloud Storage. The goal is to handle long-running songbook generation jobs without timeouts, providing a better user experience with status updates.

AI Summary: Migrate the infrastructure setup of the Songbook Generator application from manual `gcloud` commands to an Infrastructure-as-Code (IaC) approach, likely using Terraform, to improve reproducibility and deployment across different environments. This involves defining the Google Cloud Platform resources (e.g., Cloud Functions, storage) using declarative configuration files.

AI Summary: Implement functionality in the Songbook Generator web application to allow users to click on a song title within the generated PDF (likely using a PDF viewer library) and be redirected to the corresponding entry in the Table of Contents. This enhancement improves user experience, particularly for tablet users.

AI Summary: Implement proper table of contents (TOC) support in the songbook generator by leveraging the `set_toc` function from the PyMuPDF library. This involves determining the page number for each song and using this information to create a correctly indexed TOC within the generated PDF.

AI Summary: This task requires modifying the frontend of the Songbook Generator web application to accept a list of Google Drive folder IDs as input for song sheet sources. This involves updating the `ui/index.html` file to handle multiple folder selections. Additionally, it requires obtaining read-only access for the Songbook Generator service account (`993670465212-compute@developer.gserviceaccount.com`) to the relevant Google Drive folders containing the 'Ready to Play' song sheets.

AI Summary: Implement clickable links in the table of contents (TOC) of the generated songbook PDF. This involves using the PyMuPDF library to add links to each song title in the TOC, directing the user to the corresponding page within the PDF.