odoo-integration

Use the xmlrpc npm package to create clients against the /xmlrpc/2/common and /xmlrpc/2/object endpoints. Authenticate with database, login, and password to get a uid, then call execute_kw for methods like search_read, read, create, and write.

Query stock.picking with searchRead using a state domain filter, then append rows via the Google Sheets API with a service account. Track synced records with a deduplication key column and a last-date value so each run only processes new records.

The XML-RPC protocol is identical across versions, but some field names changed. For example, move_lines became move_ids in v13 and min_date became scheduled_date. Use fallback patterns like picking.move_ids || picking.move_lines when the version is unknown.

The xmlrpc library returns false instead of throwing when credentials or the database name are wrong. Always check if the returned uid is falsy after authenticate and reject explicitly, otherwise later calls fail with confusing errors.

Add a Procfile containing 'web: node index.js', set engines to Node 18 or higher in package.json, and configure all credentials as Railway environment variables. Paste the Google service account JSON as a single minified line in GOOGLE_CREDENTIALS.

Odoo XML-RPC never returns all fields by default; you must explicitly list the fields you need. Pass a fields array such as ['id', 'name', 'state', 'date', 'origin'] in every searchRead call, and raise the limit for production volumes.