makes orchestrator.run return the results to allow for code integration (#196)

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [project]
 name = "auto-archiver"
-version = "0.13.1"
+version = "0.13.2"
 description = "Automatically archive links to videos, images, and social media content from Google Sheets (and more)."
 
 requires-python = ">=3.10,<3.13"

src/auto_archiver/core/orchestrator.py

@@ -30,6 +30,7 @@ from loguru import logger
 
 DEFAULT_CONFIG_FILE = "orchestration.yaml"
 
+
 class JsonParseAction(argparse.Action):
     def __call__(self, parser, namespace, values, option_string=None):
         try:
@@ -60,6 +61,8 @@ class AuthenticationJsonParseAction(JsonParseAction):
             if not isinstance(site, str) or not isinstance(auth, dict):
                 raise argparse.ArgumentTypeError("Authentication must be a dictionary of site names and their authentication methods")
         setattr(namespace, self.dest, auth_dict)
+
+
 class UniqueAppendAction(argparse.Action):
     def __call__(self, parser, namespace, values, option_string=None):
         if not hasattr(namespace, self.dest):
@@ -68,6 +71,7 @@ class UniqueAppendAction(argparse.Action):
             if value not in getattr(namespace, self.dest):
                 getattr(namespace, self.dest).append(value)
 
+
 class ArchivingOrchestrator:
 
     feeders: List[Type[Feeder]]
@@ -166,7 +170,6 @@ class ArchivingOrchestrator:
         if not parser:
             parser = self.parser
 
-
         # allow passing URLs directly on the command line
         parser.add_argument('urls', nargs='*', default=[], help='URL(s) to archive, either a single URL or a list of urls, should not come from config.yaml')
 
@@ -180,7 +183,7 @@ class ArchivingOrchestrator:
         parser.add_argument('--authentication', dest='authentication', help='A dictionary of sites and their authentication methods \
                                                                             (token, username etc.) that extractors can use to log into \
                                                                             a website. If passing this on the command line, use a JSON string. \
-                                                                            You may also pass a path to a valid JSON/YAML file which will be parsed.',\
+                                                                            You may also pass a path to a valid JSON/YAML file which will be parsed.',
                                                                             default={},
                                                                             action=AuthenticationJsonParseAction)
         # logging arguments
@@ -188,7 +191,6 @@ class ArchivingOrchestrator:
         parser.add_argument('--logging.file', action='store', dest='logging.file', help='the logging file to write to', default=None)
         parser.add_argument('--logging.rotation', action='store', dest='logging.rotation', help='the logging rotation to use', default=None)
 
-
     def add_module_args(self, modules: list[LazyBaseModule] = None, parser: argparse.ArgumentParser = None) -> None:
 
         if not modules:
@@ -273,6 +275,7 @@ class ArchivingOrchestrator:
                         logger.error("No URLs provided. Please provide at least one URL via the command line, or set up an alternative feeder. Use --help for more information.")
                         exit()
                     # cli_feeder is a pseudo module, it just takes the command line args
+
                     def feed(self) -> Generator[Metadata]:
                         for url in urls:
                             logger.debug(f"Processing URL: '{url}'")
@@ -285,7 +288,6 @@ class ArchivingOrchestrator:
 
                     })()
 
-
                     pseudo_module.__iter__ = feed
                     step_items.append(pseudo_module)
                     continue
@@ -316,7 +318,7 @@ class ArchivingOrchestrator:
 
         return read_yaml(config_file)
 
-    def run(self, args: list) -> None:
+    def run(self, args: list) -> Generator[Metadata]:
 
         self.setup_basic_parser()
 
@@ -340,8 +342,8 @@ class ArchivingOrchestrator:
         for module_type in BaseModule.MODULE_TYPES:
             logger.info(f"{module_type.upper()}S: " + ", ".join(m.display_name for m in getattr(self, f"{module_type}s")))
 
-        for _ in self.feed():
-            pass
+        for result in self.feed():
+            yield result
 
     def cleanup(self) -> None:
         logger.info("Cleaning up")
@@ -393,7 +395,6 @@ class ArchivingOrchestrator:
                     m.tmp_dir = None
                 tmp_dir.cleanup()
 
-
     def archive(self, result: Metadata) -> Union[Metadata, None]:
         """
             Runs the archiving process for a single URL
@@ -489,7 +490,6 @@ class ArchivingOrchestrator:
             assert not ip.is_link_local, f"Invalid IP used"
             assert not ip.is_private, f"Invalid IP used"
 
-
     # Helper Properties
 
     @property