Pull invoice details out of PDFs in a folder

from google.oauth2.credentials import Credentials
from googleapiclient.discovery import build
import base64
from clickruntime import on_trigger, report_fetched, credentials, state, ai, notify

FOLDER = "your-invoices-folder-id"
SCHEMA = {"type": "object", "properties": {k: {"type": "string"} for k in ("supplier", "date", "total", "vat")},
          "required": ["supplier", "date", "total", "vat"]}

@on_trigger
def run(event):
    drive = build("drive", "v3", credentials=Credentials(credentials.connection("google", "work")), cache_discovery=False)
    since = state.get("since") or "2026-01-01T00:00:00Z"
    files = drive.files().list(q=f"'{FOLDER}' in parents and mimeType = 'application/pdf' and createdTime > '{since}'",
                               orderBy="createdTime", fields="files(id,name,createdTime)").execute()["files"]
    report_fetched(len(files))
    if not files:
        return {"summary": "Nothing new"}
    rows = ["file,supplier,date,total,vat"]
    for f in files:
        # Copy as a Google Doc so Drive runs OCR, then read it back as text.
        doc = drive.files().copy(fileId=f["id"], body={"mimeType": "application/vnd.google-apps.document"}).execute()
        text = drive.files().export(fileId=doc["id"], mimeType="text/plain").execute().decode()[:6000]
        drive.files().delete(fileId=doc["id"]).execute()
        d = ai.generate("Extract supplier, invoice date, total and VAT:\n" + text, json_schema=SCHEMA, max_output_tokens=120).json
        rows.append(",".join([f["name"], d["supplier"], d["date"], d["total"], d["vat"]]))
    notify.email_me(f"{len(files)} new invoices", "\n".join(rows))
    state.set("since", files[-1]["createdTime"])
    return {"invoices": len(files)}