editorial-system

Editorial System Documentation

Purpose

This application gives the editorial team a single-page operations console for harvesting outside sources, enriching them with keywords, preventing duplicates, and mirroring select content from WordPress. The stack is intentionally minimal: plain PHP 8 for every server concern and a vanilla JavaScript frontend served from index.html.

Quick Facts

• Authentication: HTTP Basic credentials from Config::getAdminLogin() protect every API request via includes/init.php; the SPA stores credentials in localStorage and attaches them to each /api call. Public Slack endpoints opt out via IS_PUBLIC_API.
• APIs: All /api/*.php endpoints boot through init.php, which wires configuration, database access, and feature toggles (for example the ONLINE constant that can gate requests).
• Database: includes/model.php boots a MySQL schema at runtime using the DB_SCHEMA SQL string so developers do not run migrations manually.
• External services: Firecrawl supplies article scraping, OpenAI + Anthropic handle keyword extraction and checks, and Slack-compatible webhooks announce newly created sources or fact-check results.
• WordPress bridge: WordPressConfig in includes/config-schema.php knows how to backfill posts from the remote WP site while tagging authors, categories, tags, and derived keywords.

Directory Overview

| Path | Description | | — | — | | index.html | Serves the SPA shell (styles.css, main.js). | | main.js | Implements the entire frontend (routing, UI state, API calls, WordPress helpers). | | styles.css | Default styling for the dashboard (Flexbox layout, panels, chips, etc.). | | includes/ | Configuration schema + defaults, bootstrap (init.php), ORM-like Model, WordPress helpers, and cron utilities. | | api/ | Thin endpoint scripts that validate inputs and dispatch to Model methods. | | docs/ | Documentation set (this file plus backend, frontend, and operations guides). |

High-Level Flow

1. User signs in via the SPA login form and loads index.html.
2. Frontend initializes (main.js), pulling /api/users.php to hydrate appState.users, reading the cached author id and Basic Auth credentials from localStorage, and mounting hash-based routes.
3. Editors capture sources through the “新規ソース” workflow:
   • Paste a URL; Firecrawl scrapes metadata and Markdown via /api/crawl.php.
   • The UI triggers /api/detect-keywords.php, stores suggested keywords/title, and forces /api/sources.php duplicate checks.
   • If the reviewer approves, /api/sources.php creates the record, kicks the Slack webhook, and enforces rate limits (WORKING_SOURCES_LIMIT).
4. Source triage + keyword views read from /api/sources.php, /api/search-sources.php, /api/keywords.php, /api/keyword-counts.php, etc.
5. WordPress ingestion flows through cron endpoints like /api/cron.php or CLI backfill scripts that call WordPressConfig::back_fill(), persist posts, and detect keywords per post. New WordPress posts can also trigger background fact/editorial checks.
6. WordPress browsing inside the SPA uses /api/wp-post.php+/api/search-keywords.php and displays titles via the decoding helper introduced in main.js (resolveWordPressTitle).

Additional Documents

• docs/backend.md — Database, server bootstrap, Model, and endpoint specifics.
• docs/frontend.md — SPA architecture, routes, and UI behaviors.
• docs/operations.md — Setup, configuration, cron/backfill jobs, and external service notes.