Merge branch 'release/v0.21' into pulsar-api-support

This commit is contained in:
Tyler O 2025-02-10 17:29:00 +00:00
commit 5bdb9c1919
149 changed files with 3916 additions and 1823 deletions

View file

@ -12,14 +12,13 @@ import os
import time
import uuid
from trustgraph.schema import TextDocument, text_ingest_queue
from trustgraph.schema import Metadata, Triple, Value
from trustgraph.log_level import LogLevel
from trustgraph.knowledge import hash, to_uri, Literal, Uri
from trustgraph.api import Api
from trustgraph.knowledge import hash, to_uri
from trustgraph.knowledge import PREF_PUBEV, PREF_DOC, PREF_ORG
from trustgraph.knowledge import Organization, PublicationEvent
from trustgraph.knowledge import DigitalDocument
default_url = os.getenv("TRUSTGRAPH_URL", 'http://localhost:8088/')
default_user = 'trustgraph'
default_collection = 'default'
@ -27,32 +26,14 @@ class Loader:
def __init__(
self,
pulsar_host,
output_queue,
url,
user,
collection,
log_level,
metadata,
pulsar_api_key=None,
):
if pulsar_api_key:
auth = pulsar.AuthenticationToken(pulsar_api_key)
self.client = pulsar.Client(
pulsar_host,
authentication=auth,
logger=pulsar.ConsoleLogger(log_level.to_pulsar())
)
else:
self.client = pulsar.Client(
pulsar_host,
logger=pulsar.ConsoleLogger(log_level.to_pulsar())
)
self.producer = self.client.create_producer(
topic=output_queue,
schema=JsonSchema(TextDocument),
chunking_enabled=True,
)
self.api = Api(url)
self.user = user
self.collection = collection
@ -75,49 +56,18 @@ class Loader:
id = to_uri(PREF_DOC, id)
triples = []
def emit(t):
triples.append(t)
self.metadata.id = id
self.metadata.emit(emit)
r = TextDocument(
metadata=Metadata(
id=id,
metadata=[
Triple(
s=Value(
value=t["s"],
is_uri=isinstance(t["s"], Uri)
),
p=Value(
value=t["p"],
is_uri=isinstance(t["p"], Uri)
),
o=Value(
value=t["o"],
is_uri=isinstance(t["o"], Uri)
),
)
for t in triples
],
user=self.user,
collection=self.collection,
),
text=data,
self.api.load_text(
text=data, id=id, metadata=self.metadata,
# user=self.user,
# collection=self.collection,
)
self.producer.send(r)
print(f"{file}: Loaded successfully.")
except Exception as e:
print(f"{file}: Failed: {str(e)}", flush=True)
def __del__(self):
self.client.close()
def main():
@ -126,37 +76,20 @@ def main():
description=__doc__,
)
default_pulsar_host = os.getenv("PULSAR_HOST", 'pulsar://localhost:6650')
default_pulsar_api_key = os.getenv("PULSAR_API_KEY", None)
default_output_queue = text_ingest_queue
parser.add_argument(
'-p', '--pulsar-host',
default=default_pulsar_host,
help=f'Pulsar host (default: {default_pulsar_host})',
)
parser.add_argument(
'--pulsar-api-key',
default=default_pulsar_api_key,
help=f'Pulsar API key',
'-u', '--url',
default=default_url,
help=f'API URL (default: {default_url})',
)
parser.add_argument(
'-o', '--output-queue',
default=default_output_queue,
help=f'Output queue (default: {default_output_queue})'
)
parser.add_argument(
'-u', '--user',
'-U', '--user',
default=default_user,
help=f'User ID (default: {default_user})'
)
parser.add_argument(
'-c', '--collection',
'-C', '--collection',
default=default_collection,
help=f'Collection ID (default: {default_collection})'
)
@ -198,7 +131,7 @@ def main():
)
parser.add_argument(
'--url', help=f'Document URL'
'--document-url', help=f'Document URL'
)
parser.add_argument(
@ -209,14 +142,6 @@ def main():
'--identifier', '--id', help=f'Document ID'
)
parser.add_argument(
'-l', '--log-level',
type=LogLevel,
default=LogLevel.ERROR,
choices=list(LogLevel),
help=f'Output queue (default: info)'
)
parser.add_argument(
'files', nargs='+',
help=f'File to load'
@ -236,7 +161,7 @@ def main():
copyright_holder=args.copyright_holder,
copyright_year=args.copyright_year,
license=args.license,
url=args.url,
url=args.document_url,
keywords=args.keyword,
)
@ -254,12 +179,9 @@ def main():
)
p = Loader(
pulsar_host=args.pulsar_host,
pulsar_api_key=args.pulsar_api_key,
output_queue=args.output_queue,
url=args.url,
user=args.user,
collection=args.collection,
log_level=args.log_level,
metadata=document,
)