mirror of
https://github.com/bellingcat/auto-archiver.git
synced 2026-06-11 12:48:28 +03:00
51 lines
1.2 KiB
YAML
51 lines
1.2 KiB
YAML
steps:
|
|
# only 1 feeder allowed
|
|
# a feeder could be in an "infinite loop" for example: gsheets_infinite feeder which holds-> this could be an easy logic addiction by modifying for each to while not feeder.done() if it becomes necessary
|
|
feeder: gsheets_feeder # default -> only expects URL from CLI
|
|
archivers: # order matters
|
|
- tiktok
|
|
- telethon
|
|
- twitter
|
|
- instagram
|
|
- webarchive # this way it runs as a failsafe only
|
|
enrichers:
|
|
- screenshot
|
|
- wacz
|
|
- webarchive # this way it runs for every case, webarchive extends archiver and enrichment
|
|
- thumbnails
|
|
formatters:
|
|
- HTMLFormater
|
|
- PdfFormater
|
|
storages:
|
|
- local_storage
|
|
- s3
|
|
databases:
|
|
- gsheets_db
|
|
- mongo_db
|
|
|
|
|
|
|
|
configurations:
|
|
global:
|
|
- save_logs: False
|
|
gsheets_feeder:
|
|
- sheet: "Auto archiver"
|
|
- header: "" # defaults to 1 in GSheetsFeeder
|
|
- service_account: "secrets/service_account.json"
|
|
tiktok:
|
|
username: "abc"
|
|
password: "123"
|
|
token: "here"
|
|
screenshot:
|
|
width: 1280
|
|
height: 720000
|
|
wacz:
|
|
profile: secrets/profile.tar.gz
|
|
webarchive:
|
|
api_key: "12345"
|
|
s3:
|
|
- bucket: 123
|
|
- region: "nyc3"
|
|
- cdn: "{region}{bucket}"
|
|
|