Merge branch 'load_modules' into timestamping_rewrite

This commit is contained in:
Patrick Robertson
2025-02-11 15:21:31 +00:00
122 changed files with 3281 additions and 1011 deletions

View File

@@ -0,0 +1,2 @@
https://example.com/1/,data 1
https://example.com/2/,data 2
1 https://example.com/1/ data 1
2 https://example.com/2/ data 2

View File

@@ -0,0 +1,3 @@
webpages,other data
https://example.com/1/,data 1
https://example.com/2/,data 2
1 webpages other data
2 https://example.com/1/ data 1
3 https://example.com/2/ data 2

View File

@@ -0,0 +1 @@
from .example_module import ExampleModule

View File

@@ -0,0 +1,11 @@
{
"name": "Example Module",
"type": ["extractor", "feeder", "formatter", "storage", "enricher", "database"],
"requires_setup": False,
"dependencies": {"python": ["loguru"]
},
"configs": {
"csv_file": {"default": "db.csv", "help": "CSV file name"},
"required_field": {"required": True, "help": "required field in the CSV file"},
},
}

View File

@@ -0,0 +1,28 @@
from auto_archiver.core import Extractor, Enricher, Feeder, Database, Storage, Formatter, Metadata
class ExampleModule(Extractor, Enricher, Feeder, Database, Storage, Formatter):
def download(self, item):
print("download")
def __iter__(self):
yield Metadata().set_url("https://example.com")
def done(self, result):
print("done")
def enrich(self, to_enrich):
print("enrich")
def get_cdn_url(self, media):
return "nice_url"
def save(self, item):
print("save")
def uploadf(self, file, key, **kwargs):
print("uploadf")
def format(self, item):
print("format")

View File

@@ -0,0 +1,16 @@
steps:
feeders:
- example_module
extractors:
- example_module
formatters:
- example_module
storages:
- example_module
databases:
- example_module
enrichers:
- example_module
# Global configuration