From 5244e866154f054346a673fbb8e2f6cb4dfe8159 Mon Sep 17 00:00:00 2001
From: Cyber MacGeddon <cybermaggedon@gmail.com>
Date: Thu, 4 Sep 2025 15:37:12 +0100
Subject: [PATCH] Added CLI

---
 trustgraph-cli/pyproject.toml                 |   2 +
 .../trustgraph/cli/invoke_nlp_query.py        | 111 ++++++++++++
 .../trustgraph/cli/invoke_structured_query.py | 159 ++++++++++++++++++
 3 files changed, 272 insertions(+)
 create mode 100644 trustgraph-cli/trustgraph/cli/invoke_nlp_query.py
 create mode 100644 trustgraph-cli/trustgraph/cli/invoke_structured_query.py

diff --git a/trustgraph-cli/pyproject.toml b/trustgraph-cli/pyproject.toml
index feb9a306..aa2be8cb 100644
--- a/trustgraph-cli/pyproject.toml
+++ b/trustgraph-cli/pyproject.toml
@@ -43,8 +43,10 @@ tg-invoke-document-rag = "trustgraph.cli.invoke_document_rag:main"
 tg-invoke-graph-rag = "trustgraph.cli.invoke_graph_rag:main"
 tg-invoke-llm = "trustgraph.cli.invoke_llm:main"
 tg-invoke-mcp-tool = "trustgraph.cli.invoke_mcp_tool:main"
+tg-invoke-nlp-query = "trustgraph.cli.invoke_nlp_query:main"
 tg-invoke-objects-query = "trustgraph.cli.invoke_objects_query:main"
 tg-invoke-prompt = "trustgraph.cli.invoke_prompt:main"
+tg-invoke-structured-query = "trustgraph.cli.invoke_structured_query:main"
 tg-load-doc-embeds = "trustgraph.cli.load_doc_embeds:main"
 tg-load-kg-core = "trustgraph.cli.load_kg_core:main"
 tg-load-pdf = "trustgraph.cli.load_pdf:main"
diff --git a/trustgraph-cli/trustgraph/cli/invoke_nlp_query.py b/trustgraph-cli/trustgraph/cli/invoke_nlp_query.py
new file mode 100644
index 00000000..8b01187c
--- /dev/null
+++ b/trustgraph-cli/trustgraph/cli/invoke_nlp_query.py
@@ -0,0 +1,111 @@
+"""
+Uses the NLP Query service to convert natural language questions to GraphQL queries
+"""
+
+import argparse
+import os
+import json
+import sys
+from trustgraph.api import Api
+
+default_url = os.getenv("TRUSTGRAPH_URL", 'http://localhost:8088/')
+
+def nlp_query(url, flow_id, question, max_results, output_format='json'):
+
+    api = Api(url).flow().id(flow_id)
+
+    resp = api.nlp_query(
+        question=question,
+        max_results=max_results
+    )
+
+    # Check for errors
+    if "error" in resp and resp["error"]:
+        print("Error:", resp["error"].get("message", "Unknown error"), file=sys.stderr)
+        sys.exit(1)
+
+    # Format output based on requested format
+    if output_format == 'json':
+        print(json.dumps(resp, indent=2))
+    elif output_format == 'graphql':
+        # Just print the GraphQL query
+        if "graphql_query" in resp:
+            print(resp["graphql_query"])
+        else:
+            print("No GraphQL query generated", file=sys.stderr)
+            sys.exit(1)
+    elif output_format == 'summary':
+        # Print a human-readable summary
+        if "graphql_query" in resp:
+            print(f"Generated GraphQL Query:")
+            print("-" * 40)
+            print(resp["graphql_query"])
+            print("-" * 40)
+            if "detected_schemas" in resp and resp["detected_schemas"]:
+                print(f"Detected Schemas: {', '.join(resp['detected_schemas'])}")
+            if "confidence" in resp:
+                print(f"Confidence: {resp['confidence']:.2%}")
+            if "variables" in resp and resp["variables"]:
+                print(f"Variables: {json.dumps(resp['variables'], indent=2)}")
+        else:
+            print("No GraphQL query generated", file=sys.stderr)
+            sys.exit(1)
+
+def main():
+
+    parser = argparse.ArgumentParser(
+        prog='tg-invoke-nlp-query',
+        description=__doc__,
+    )
+
+    parser.add_argument(
+        '-u', '--url',
+        default=default_url,
+        help=f'API URL (default: {default_url})',
+    )
+
+    parser.add_argument(
+        '-f', '--flow-id',
+        default="default",
+        help=f'Flow ID (default: default)'
+    )
+
+    parser.add_argument(
+        '-q', '--question',
+        required=True,
+        help='Natural language question to convert to GraphQL',
+    )
+
+    parser.add_argument(
+        '-m', '--max-results',
+        type=int,
+        default=100,
+        help='Maximum number of results (default: 100)'
+    )
+
+    parser.add_argument(
+        '--format', 
+        choices=['json', 'graphql', 'summary'],
+        default='summary',
+        help='Output format (default: summary)'
+    )
+
+    args = parser.parse_args()
+
+    try:
+
+        nlp_query(
+            url=args.url,
+            flow_id=args.flow_id,
+            question=args.question,
+            max_results=args.max_results,
+            output_format=args.format,
+        )
+
+    except Exception as e:
+
+        print("Exception:", e, flush=True, file=sys.stderr)
+        sys.exit(1)
+
+if __name__ == "__main__":
+    main()
\ No newline at end of file
diff --git a/trustgraph-cli/trustgraph/cli/invoke_structured_query.py b/trustgraph-cli/trustgraph/cli/invoke_structured_query.py
new file mode 100644
index 00000000..8f34e747
--- /dev/null
+++ b/trustgraph-cli/trustgraph/cli/invoke_structured_query.py
@@ -0,0 +1,159 @@
+"""
+Uses the Structured Query service to execute natural language questions against structured data
+"""
+
+import argparse
+import os
+import json
+import sys
+import csv
+import io
+from trustgraph.api import Api
+from tabulate import tabulate
+
+default_url = os.getenv("TRUSTGRAPH_URL", 'http://localhost:8088/')
+
+def format_output(data, output_format):
+    """Format structured query response data in the specified format"""
+    if not data:
+        return "No data returned"
+    
+    # Handle case where data contains multiple query results
+    if isinstance(data, dict) and len(data) == 1:
+        # Single query result - extract the list
+        query_name, result_list = next(iter(data.items()))
+        if isinstance(result_list, list):
+            return format_table_data(result_list, query_name, output_format)
+    
+    # Multiple queries or non-list data - use JSON format
+    if output_format == 'json':
+        return json.dumps(data, indent=2)
+    else:
+        return json.dumps(data, indent=2)  # Fallback to JSON
+
+def format_table_data(rows, table_name, output_format):
+    """Format a list of rows in the specified format"""
+    if not rows:
+        return f"No {table_name} found"
+    
+    if output_format == 'json':
+        return json.dumps({table_name: rows}, indent=2)
+    
+    elif output_format == 'csv':
+        # Get field names in order from first row, then add any missing ones
+        fieldnames = list(rows[0].keys()) if rows else []
+        # Add any additional fields from other rows that might be missing
+        all_fields = set(fieldnames)
+        for row in rows:
+            for field in row.keys():
+                if field not in all_fields:
+                    fieldnames.append(field)
+                    all_fields.add(field)
+        
+        # Create CSV string
+        output = io.StringIO()
+        writer = csv.DictWriter(output, fieldnames=fieldnames)
+        writer.writeheader()
+        writer.writerows(rows)
+        return output.getvalue().rstrip()
+    
+    elif output_format == 'table':
+        # Get field names in order from first row, then add any missing ones
+        fieldnames = list(rows[0].keys()) if rows else []
+        # Add any additional fields from other rows that might be missing
+        all_fields = set(fieldnames)
+        for row in rows:
+            for field in row.keys():
+                if field not in all_fields:
+                    fieldnames.append(field)
+                    all_fields.add(field)
+        
+        # Create table data
+        table_data = []
+        for row in rows:
+            table_row = [row.get(field, '') for field in fieldnames]
+            table_data.append(table_row)
+        
+        return tabulate(table_data, headers=fieldnames, tablefmt='pretty')
+    
+    else:
+        return json.dumps({table_name: rows}, indent=2)
+
+def structured_query(url, flow_id, question, output_format='table'):
+
+    api = Api(url).flow().id(flow_id)
+
+    resp = api.structured_query(question=question)
+
+    # Check for errors
+    if "error" in resp and resp["error"]:
+        print("Error:", resp["error"].get("message", "Unknown error"), file=sys.stderr)
+        sys.exit(1)
+
+    # Check for query errors
+    if "errors" in resp and resp["errors"]:
+        print("Query Errors:", file=sys.stderr)
+        for error in resp["errors"]:
+            print(f"  - {error}", file=sys.stderr)
+        # Still print data if available
+        if "data" in resp and resp["data"]:
+            print(format_output(resp["data"], output_format))
+        sys.exit(1)
+
+    # Print the data
+    if "data" in resp:
+        print(format_output(resp["data"], output_format))
+    else:
+        print("No data returned", file=sys.stderr)
+        sys.exit(1)
+
+def main():
+
+    parser = argparse.ArgumentParser(
+        prog='tg-invoke-structured-query',
+        description=__doc__,
+    )
+
+    parser.add_argument(
+        '-u', '--url',
+        default=default_url,
+        help=f'API URL (default: {default_url})',
+    )
+
+    parser.add_argument(
+        '-f', '--flow-id',
+        default="default",
+        help=f'Flow ID (default: default)'
+    )
+
+    parser.add_argument(
+        '-q', '--question',
+        required=True,
+        help='Natural language question to execute',
+    )
+
+    parser.add_argument(
+        '--format', 
+        choices=['table', 'json', 'csv'],
+        default='table',
+        help='Output format (default: table)'
+    )
+
+    args = parser.parse_args()
+
+    try:
+
+        structured_query(
+            url=args.url,
+            flow_id=args.flow_id,
+            question=args.question,
+            output_format=args.format,
+        )
+
+    except Exception as e:
+
+        print("Exception:", e, flush=True, file=sys.stderr)
+        sys.exit(1)
+
+if __name__ == "__main__":
+    main()
\ No newline at end of file