trustgraph-cli 1.2.10__tar.gz → 1.2.12__tar.gz

{trustgraph_cli-1.2.10 → trustgraph_cli-1.2.12}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: trustgraph-cli
-Version: 1.2.10
+Version: 1.2.12
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Author-email: "trustgraph.ai" <security@trustgraph.ai>
 Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph

{trustgraph_cli-1.2.10 → trustgraph_cli-1.2.12}/pyproject.toml

@@ -50,6 +50,7 @@ tg-load-pdf = "trustgraph.cli.load_pdf:main"
 tg-load-sample-documents = "trustgraph.cli.load_sample_documents:main"
 tg-load-text = "trustgraph.cli.load_text:main"
 tg-load-turtle = "trustgraph.cli.load_turtle:main"
+tg-load-knowledge = "trustgraph.cli.load_knowledge:main"
 tg-put-flow-class = "trustgraph.cli.put_flow_class:main"
 tg-put-kg-core = "trustgraph.cli.put_kg_core:main"
 tg-remove-library-document = "trustgraph.cli.remove_library_document:main"

trustgraph_cli-1.2.12/trustgraph/cli/load_knowledge.py

@@ -0,0 +1,202 @@
+"""
+Loads triples and entity contexts into the knowledge graph.
+"""
+
+import asyncio
+import argparse
+import os
+import time
+import rdflib
+import json
+from websockets.asyncio.client import connect
+from typing import List, Dict, Any
+
+from trustgraph.log_level import LogLevel
+
+default_url = os.getenv("TRUSTGRAPH_URL", 'ws://localhost:8088/')
+default_user = 'trustgraph'
+default_collection = 'default'
+
+class KnowledgeLoader:
+
+    def __init__(
+            self,
+            files,
+            flow,
+            user,
+            collection,
+            document_id,
+            url = default_url,
+    ):
+
+        if not url.endswith("/"):
+            url += "/"
+
+        self.triples_url = url + f"api/v1/flow/{flow}/import/triples"
+        self.entity_contexts_url = url + f"api/v1/flow/{flow}/import/entity-contexts"
+
+        self.files = files
+        self.user = user
+        self.collection = collection
+        self.document_id = document_id
+
+    async def run(self):
+
+        try:
+            # Load triples first
+            async with connect(self.triples_url) as ws:
+                for file in self.files:
+                    await self.load_triples(file, ws)
+
+            # Then load entity contexts
+            async with connect(self.entity_contexts_url) as ws:
+                for file in self.files:
+                    await self.load_entity_contexts(file, ws)
+
+        except Exception as e:
+            print(e, flush=True)
+
+    async def load_triples(self, file, ws):
+
+        g = rdflib.Graph()
+        g.parse(file, format="turtle")
+
+        def Value(value, is_uri):
+            return { "v": value, "e": is_uri }
+
+        for e in g:
+            s = Value(value=str(e[0]), is_uri=True)
+            p = Value(value=str(e[1]), is_uri=True)
+            if type(e[2]) == rdflib.term.URIRef:
+                o = Value(value=str(e[2]), is_uri=True)
+            else:
+                o = Value(value=str(e[2]), is_uri=False)
+
+            req = {
+                "metadata": {
+                    "id": self.document_id,
+                    "metadata": [],
+                    "user": self.user,
+                    "collection": self.collection
+                },
+                "triples": [
+                    {
+                        "s": s,
+                        "p": p,
+                        "o": o,
+                    }
+                ]
+            }
+
+            await ws.send(json.dumps(req))
+
+    async def load_entity_contexts(self, file, ws):
+        """
+        Load entity contexts by extracting entities from the RDF graph
+        and generating contextual descriptions based on their relationships.
+        """
+
+        g = rdflib.Graph()
+        g.parse(file, format="turtle")
+
+        for s, p, o in g:
+            # If object is a URI, do nothing
+            if isinstance(o, rdflib.term.URIRef):
+                continue
+                
+            # If object is a literal, create entity context for subject with literal as context
+            s_str = str(s)
+            o_str = str(o)
+            
+            req = {
+                "metadata": {
+                    "id": self.document_id,
+                    "metadata": [],
+                    "user": self.user,
+                    "collection": self.collection
+                },
+                "entities": [
+                    {
+                        "entity": {
+                            "v": s_str,
+                            "e": True
+                        },
+                        "context": o_str
+                    }
+                ]
+            }
+
+            await ws.send(json.dumps(req))
+
+
+def main():
+
+    parser = argparse.ArgumentParser(
+        prog='tg-load-knowledge',
+        description=__doc__,
+    )
+
+    parser.add_argument(
+        '-u', '--api-url',
+        default=default_url,
+        help=f'API URL (default: {default_url})',
+    )
+
+    parser.add_argument(
+        '-i', '--document-id',
+        required=True,
+        help=f'Document ID)',
+    )
+
+    parser.add_argument(
+        '-f', '--flow-id',
+        default="default",
+        help=f'Flow ID (default: default)'
+    )
+
+    parser.add_argument(
+        '-U', '--user',
+        default=default_user,
+        help=f'User ID (default: {default_user})'
+    )
+
+    parser.add_argument(
+        '-C', '--collection',
+        default=default_collection,
+        help=f'Collection ID (default: {default_collection})'
+    )
+
+
+    parser.add_argument(
+        'files', nargs='+',
+        help=f'Turtle files to load'
+    )
+
+    args = parser.parse_args()
+
+    while True:
+
+        try:
+            loader = KnowledgeLoader(
+                document_id = args.document_id,
+                url = args.api_url,
+                flow = args.flow_id,
+                files = args.files,
+                user = args.user,
+                collection = args.collection,
+            )
+
+            asyncio.run(loader.run())
+
+            print("Triples and entity contexts loaded.")
+            break
+
+        except Exception as e:
+
+            print("Exception:", e, flush=True)
+            print("Will retry...", flush=True)
+
+        time.sleep(10)
+
+if __name__ == "__main__":
+    main()

trustgraph_cli-1.2.12/trustgraph/cli_version.py

@@ -0,0 +1 @@
+__version__ = "1.2.12"

{trustgraph_cli-1.2.10 → trustgraph_cli-1.2.12}/trustgraph_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: trustgraph-cli
-Version: 1.2.10
+Version: 1.2.12
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Author-email: "trustgraph.ai" <security@trustgraph.ai>
 Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph

{trustgraph_cli-1.2.10 → trustgraph_cli-1.2.12}/trustgraph_cli.egg-info/SOURCES.txt

@@ -21,6 +21,7 @@ trustgraph/cli/invoke_mcp_tool.py
 trustgraph/cli/invoke_prompt.py
 trustgraph/cli/load_doc_embeds.py
 trustgraph/cli/load_kg_core.py
+trustgraph/cli/load_knowledge.py
 trustgraph/cli/load_pdf.py
 trustgraph/cli/load_sample_documents.py
 trustgraph/cli/load_text.py

{trustgraph_cli-1.2.10 → trustgraph_cli-1.2.12}/trustgraph_cli.egg-info/entry_points.txt

@@ -17,6 +17,7 @@ tg-invoke-mcp-tool = trustgraph.cli.invoke_mcp_tool:main
 tg-invoke-prompt = trustgraph.cli.invoke_prompt:main
 tg-load-doc-embeds = trustgraph.cli.load_doc_embeds:main
 tg-load-kg-core = trustgraph.cli.load_kg_core:main
+tg-load-knowledge = trustgraph.cli.load_knowledge:main
 tg-load-pdf = trustgraph.cli.load_pdf:main
 tg-load-sample-documents = trustgraph.cli.load_sample_documents:main
 tg-load-text = trustgraph.cli.load_text:main

trustgraph_cli-1.2.10/trustgraph/cli_version.py

@@ -1 +0,0 @@
-__version__ = "1.2.10"