Update app_logic.py

app_logic.py

@@ -1,78 +1,85 @@
 # app_logic.py
-import requests
-import json
-from py2neo import Graph, Node, Relationship
+from neo4j import GraphDatabase
+from huggingface_hub import InferenceClient
+import time
 
-def run_my_application():
-    """
-    Fetch and process a remote JSON file.
-    Replace this with your real application logic.
-    """
-    # json_file_path = ["https://arinsight.co/skills/Customer_Service_skills.json", "https://arinsight.co/skills/Computer_Programming_skills.json",
-                      # "https://arinsight.co/skills/Data_Science_skills.json", "https://arinsight.co/skills/Communications_skills.json", 
-                      # "https://arinsight.co/skills/Cleaning_and_Facilities_skills.json",
-                      # "https://arinsight.co/skills/Caregiving_skills.json","https://arinsight.co/skills/Business_Development_and_Sales_skills.json", 
-                      # "https://arinsight.co/skills/Arts_and_Design_skills.json", "https://arinsight.co/skills/Architecture_and_Construction_skills.json",
-                      # "https://arinsight.co/skills/Civil_Engineering_skills.json"]
+# Setup Hugging Face inference client
+client = InferenceClient(api_key=HF_API_KEY)
 
-    json_file_path = ["https://arinsight.co/skills/Accounting_and_Finance_skills.json"]
+# Setup Neo4j database connection
+uri = "bolt://20.64.235.115:7687"
+auth = ("neo4j", "123456789")
+graph = GraphDatabase.driver(uri, auth=auth)
 
-    # Connection to Neo4j
-    graph = Graph("bolt://20.64.235.115:7687", auth=("neo4j", "123456789"))
-    
-    # Read the JSON file
-    # json_file_path = ["Engineering_skills.json", "Energy and Utilities_skills.json", "Electrical Engineering_skills.json"]
-    # graph.delete_all()
-    for file in json_file_path:
-        # with open(file, "r") as file:
-        #     data = json.load(file)
-        response = requests.get(file)
-            
-        #     # Raise an error if the request failed
-        response.raise_for_status()
-    
-        #     # Parse the JSON content
-        data = response.json()
-    
-        # Clear the graph database before starting
+# Function to generate description using Hugging Face
+def generate_description(tertiary_skill, secondary_skill, primary_skill):
+    messages = [
+        {
+            "role": "user",
+            "content": f"Give me 3 line description of {tertiary_skill} as tertiary skill, with background of Secondary Skill \"{secondary_skill}\" and degree or primary skill \"{primary_skill}\". Just 3 line description of {tertiary_skill} nothing else within scope of primary and secondary skill"
+        }
+    ]
+    try:
+        completion = client.chat.completions.create(
+            model="meta-llama/Llama-3.3-70B-Instruct",
+            messages=messages,
+            max_tokens=500
+        )
+        return completion.choices[0].message['content']
+    except Exception as e:
+        print(f"Error generating description for {tertiary_skill}: {e}")
+        return ""
+
+# Function to update tertiary skill description in the database
+def update_description(tx, tertiary_skill_name, secondary_skill_name, primary_skill_name, description):
+    query = (
+        "MATCH (p:PrimarySkill {name: $primary_skill_name})-[:HAS_SECONDARY]->(s:SecondarySkill {name: $secondary_skill_name})-[:HAS_TERTIARY]->(t:TertiarySkill {name: $tertiary_skill_name}) "
+        "SET t.description = $description"
+    )
+    tx.run(query, tertiary_skill_name=tertiary_skill_name, secondary_skill_name=secondary_skill_name, primary_skill_name=primary_skill_name, description=description)
+
+# Main processing function
+def run_my_application():
+    with graph.session() as session:
+        # Fetch all primary skills
+        primary_skills = session.run("MATCH (p:PrimarySkill) RETURN p.name AS primary_name")
         
-    
-        # Create graph nodes and relationships
-        for entry in data:
-            primary_skill = entry["Primary Skill"]
-            secondary_skill = entry["Secondary Skill"]
-            primary_description = entry.get("Description", "")
-            educational_alignment = entry.get("Educational Alignment", "")
-            curriculum_areas = "; ".join(entry.get("Curriculum Areas", []))
-            tertiary_skills = entry["Tertiary Skills"]
-    
-            # Create or get the Primary Skill node
-            primary_node = Node("PrimarySkill", name=primary_skill, description=primary_description)
-            graph.merge(primary_node, "PrimarySkill", "name")
-    
-            # Create the Secondary Skill node
-            secondary_node = Node(
-                "SecondarySkill",
-                name=secondary_skill,
-                description=primary_description,
-                educational_alignment=educational_alignment,
-                curriculum_areas=curriculum_areas,
+        for primary in primary_skills:
+            primary_name = primary["primary_name"]
+            print(f"Processing Primary Skill: {primary_name}")
+
+            # Fetch secondary skills for each primary skill
+            secondary_skills = session.run(
+                "MATCH (p:PrimarySkill {name: $primary_name})-[:HAS_SECONDARY]->(s:SecondarySkill) RETURN s.name AS secondary_name, p.name AS primary_name",
+                primary_name=primary_name
             )
-            graph.merge(secondary_node, "SecondarySkill", "name")
-    
-            # Relationship: PrimarySkill -> has_secondary -> SecondarySkill
-            relationship = Relationship(primary_node, "HAS_SECONDARY", secondary_node)
-            graph.merge(relationship)
-    
-            # Create Tertiary Skills nodes and relationships
-            for tertiary_skill in tertiary_skills:
-                tertiary_node = Node("TertiarySkill", name=tertiary_skill, description="")
-                graph.merge(tertiary_node, "TertiarySkill", "name")
-    
-                # Relationship: SecondarySkill -> has_tertiary -> TertiarySkill
-                tertiary_relationship = Relationship(secondary_node, "HAS_TERTIARY", tertiary_node)
-                graph.merge(tertiary_relationship)
-    
+
+            for secondary in secondary_skills:
+                secondary_name = secondary["secondary_name"]
+                print(f"  Processing Secondary Skill: {secondary_name}")
+
+                # Fetch tertiary skills for each secondary skill
+                tertiary_skills = session.run(
+                    "MATCH (p:PrimarySkill {name: $primary_name})-[:HAS_SECONDARY]->(s:SecondarySkill {name: $secondary_name})-[:HAS_TERTIARY]->(t:TertiarySkill) RETURN t.name AS tertiary_name, s.name AS secondary_name, p.name AS primary_name",
+                    primary_name=primary_name, secondary_name=secondary_name
+                )
+
+                for tertiary in tertiary_skills:
+                    tertiary_name = tertiary["tertiary_name"]
+                    # print(f"    Processing Tertiary Skill: {tertiary_name}")
+
+                    # Generate description for the tertiary skill
+                    description = generate_description(tertiary_name, secondary_name, primary_name)
+
+                    if description:
+                        # Update the tertiary skill's description in the database
+                        session.write_transaction(update_description, tertiary_name, secondary_name, primary_name, description)
+                        # print(f"      Updated description for {tertiary_name}")
+
+                    # To prevent rate-limiting or overload, introduce a small delay
+                    time.sleep(0.5)
+
+
     return "Graph database has been populated."