Create document drafts from new contract PDFs in Drive

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

FOLDER, TEMPLATE, FIELDS = "contracts-folder-id", "template-doc-id", ["parties", "start_date", "end_date", "amount"]

@on_trigger
def run(event):
    creds = Credentials(credentials.connection("google", "work"))
    drive = build("drive", "v3", credentials=creds, cache_discovery=False).files()
    seen = state.get("seen") or []
    pdfs = [f for f in drive.list(q=f"'{FOLDER}' in parents and mimeType='application/pdf' and trashed=false",
                                  fields="files(id,name)").execute()["files"] if f["id"] not in seen]
    report_fetched(len(pdfs))
    if not pdfs:
        return {"summary": "No new PDFs"}
    docs = build("docs", "v1", credentials=creds, cache_discovery=False).documents()
    for f in pdfs:
        text = drive.copy(fileId=f["id"], body={"mimeType": "application/vnd.google-apps.document"}).execute()
        body = http.get(f"https://www.googleapis.com/drive/v3/files/{text['id']}/export", params={"mimeType": "text/plain"},
                        headers={"Authorization": f"Bearer {creds.token}"}).text[:12000]
        drive.delete(fileId=text["id"]).execute()
        data = ai.generate(f"Extract {FIELDS} from this contract:\n{body}", max_output_tokens=200,
                           json_schema={"type": "object", "properties": {k: {"type": "string"} for k in FIELDS}}).json
        draft = drive.copy(fileId=TEMPLATE, body={"name": f"Summary - {f['name']}", "parents": [FOLDER]}).execute()
        docs.batchUpdate(documentId=draft["id"], body={"requests": [{"replaceAllText": {"containsText": {"text": "{{" + k + "}}"},
                         "replaceText": v or ""}} for k, v in data.items()]}).execute()
        seen.append(f["id"])
    state.set("seen", seen[-500:])
    return {"drafts": len(pdfs)}