herolib/aiprompts/doctree/doctree_specs.md

Doctree Export Specification

Overview

The doctree module in lib/data/doctree is responsible for processing and exporting documentation trees. This involves taking a structured representation of documentation (collections, pages, images, files) and writing it to a specified file system destination. Additionally, it leverages Redis to store metadata about the exported documentation, facilitating quick lookups and integration with other systems.

Key Components

lib/data/doctree/export.v

This file defines the main export function for the Tree object. It orchestrates the overall export process:

• Takes TreeExportArgs which includes parameters like destination, reset (to clear destination), keep_structure, exclude_errors, toreplace (for regex replacements), concurrent (for parallel processing), and redis (to control Redis metadata storage).
• Processes definitions, includes, actions, and macros within the Tree.
• Generates file paths for pages, images, and other files.
• Iterates through Collection objects within the Tree and calls their respective export methods, passing down the redis flag.

lib/data/doctree/collection/export.v

This file defines the export function for the Collection object. This is where the actual file system writing and Redis interaction for individual collections occur:

• Takes CollectionExportArgs which includes destination, file_paths, reset, keep_structure, exclude_errors, replacer, and the redis flag.
• Creates a .collection file in the destination directory with basic collection information.
• Redis Integration:
  • Obtains a Redis client using base.context().redis().
  • Stores the collection's destination path in Redis using redis.hset('doctree:path', 'collection_name', 'destination_path').
  • Calls export_pages, export_files, export_images, and export_linked_pages which all interact with Redis if the redis flag is true.
• export_pages:
  • Processes page links and handles not-found errors.
  • Writes markdown content to the destination file system.
  • Stores page metadata in Redis: redis.hset('doctree:collection_name', 'page_name', 'page_file_name.md').
• export_files and export_images:
  • Copies files and images to the destination directory (e.g., img/).
  • Stores file/image metadata in Redis: redis.hset('doctree:collection_name', 'file_name', 'img/file_name.ext').
• export_linked_pages:
  • Gathers linked pages within the collection.
  • Writes a .linkedpages file.
  • Stores linked pages file metadata in Redis: redis.hset('doctree:collection_name', 'linkedpages', 'linkedpages_file_name.md').

Link between Redis and Export

The doctree export process uses Redis as a metadata store. When the redis flag is set to true (which is the default), the export functions populate Redis with key-value pairs that map collection names, page names, file names, and image names to their respective paths and file names within the exported documentation structure.

This Redis integration serves as a quick lookup mechanism for other applications or services that might need to access or reference the exported documentation. Instead of traversing the file system, these services can query Redis to get the location of specific documentation elements.

Is Export Needed?

Yes, the export functionality is crucial for making the processed doctree content available outside the internal doctree representation.

• File System Export: The core purpose of the export is to write the documentation content (markdown files, images, other assets) to a specified directory. This is essential for serving the documentation via a web server, integrating with static site generators (like Docusaurus, as suggested by other files in the project), or simply providing a browsable version of the documentation.
• Redis Metadata: While the file system export is fundamental, the Redis metadata storage is an important complementary feature. It provides an efficient way for other systems to programmatically discover and locate documentation assets. If there are downstream applications that rely on this Redis metadata for navigation, search, or content delivery, then the Redis part of the export is indeed needed. If no such applications exist or are planned, the redis flag can be set to false to skip this step, but the file system export itself remains necessary for external consumption of the documentation.