Update app.py

app.py

@@ -7,6 +7,7 @@ import networkx as nx
 import logging
 from sklearn.feature_extraction.text import TfidfVectorizer
 import re
+import os
 
 # Set up logging configuration
 logging.basicConfig(level=logging.INFO)
@@ -23,114 +24,94 @@ install_spacy_model()
 nlp = spacy.load("en_core_web_sm")
 
 # Neo4j credentials
-uri = "neo4j+s://ff701b1c.databases.neo4j.io"
-username = "neo4j"
-password = "BfZM7YRKpFz1b_V7acAmOtaSQHPU9xK03rJlfPep88g"
-
-driver = None
-try:
-    driver = GraphDatabase.driver(uri, auth=(username, password))
-    logging.info("✅ Connected to Neo4j!")
-
-    def close_driver():
-        if driver:
-            driver.close()
-            logging.info("🔒 Neo4j driver closed.")
-
-    def create_entity(tx, name: str):
-        tx.run("MERGE (e:Entity {name: $name})", name=name)
-
-    def create_relationship(tx, subj: str, pred: str, obj: str):
-        tx.run("""
-            MERGE (a:Entity {name: $subj})
-            MERGE (b:Entity {name: $obj})
-            MERGE (a)-[:RELATION {name: $pred}]->(b)
-        """, subj=subj, pred=pred, obj=obj)
-
-    # Text Processing
-    def load_and_clean_text(file_path: str) -> str:
-        with open(file_path, 'r', encoding='utf-8') as file:
-            text = file.read()
-        text = re.sub(r'\n+', ' ', text)
-        return re.sub(r'\s+', ' ', text).strip().lower()
-
-    # TF-IDF Filtering
-    def compute_tfidf_keywords(text: str, top_n=60):
-        vectorizer = TfidfVectorizer(stop_words='english')
-        X = vectorizer.fit_transform([text])
-        scores = zip(vectorizer.get_feature_names_out(), X.toarray()[0])
-        sorted_scores = sorted(scores, key=lambda x: x[1], reverse=True)
-        return {word for word, _ in sorted_scores[:top_n]}
-
-    # Triple Extraction
-    def get_full_phrase(token) -> str:
-        return ' '.join(tok.text for tok in token.subtree if tok.dep_ != 'punct').strip()
-
-    def extract_rich_triples(doc, tfidf_keywords) -> list:
-        triples = []
-        for sent in doc.sents:
-            subjects = [tok for tok in sent if "subj" in tok.dep_]
-            objects = [tok for tok in sent if "obj" in tok.dep_]
-            verbs = [tok for tok in sent if tok.pos_ == "VERB"]
-            for subj in subjects:
-                for obj in objects:
-                    for verb in verbs:
-                        s = get_full_phrase(subj)
-                        o = get_full_phrase(obj)
-                        if s.lower() in tfidf_keywords or o.lower() in tfidf_keywords:
-                            triples.append((s, verb.lemma_, o))
-        return triples
-
-    # Graph Visualization
-    def visualize_knowledge_graph(triples: list, title: str = "Knowledge Graph"):
-        G = nx.DiGraph()
-        for subj, pred, obj in triples:
-            G.add_node(subj, label='Subject')
-            G.add_node(obj, label='Object')
-            G.add_edge(subj, obj, label=pred)
-
-        pos = nx.spring_layout(G, k=1.2, seed=42)
-        node_colors = ['skyblue' if G.nodes[n]['label'] == 'Subject' else 'lightgreen' for n in G.nodes]
-
-        plt.figure(figsize=(16, 16))
-        nx.draw(G, pos, with_labels=True, node_size=1200, node_color=node_colors,
-                font_size=10, font_weight='bold', edge_color='gray', alpha=0.8)
-        nx.draw_networkx_edge_labels(G, pos, edge_labels={(u, v): d['label'] for u, v, d in G.edges(data=True)},
-                                     font_size=8, font_color='red')
-        plt.title(title, fontsize=20)
-        plt.show()
-
-    # Streamlit UI
-    st.title("🧠 Knowledge Graph Generator")
-
-    text_input = st.text_area("Paste your text here", height=200)
-
-    if st.button("Generate Graph"):
-        if text_input:
+uri = os.environ.get("NEO4J_URI")
+username = os.environ.get("NEO4J_USERNAME")
+password = os.environ.get("NEO4J_PASSWORD")
+
+# Connect to Neo4j to check the connection and then close it
+def connect_to_neo4j():
+    try:
+        driver = GraphDatabase.driver(uri, auth=(username, password))
+        logging.info("✅ Connected to Neo4j!")
+        driver.close()
+        logging.info("🔒 Neo4j driver closed.")
+    except Exception as e:
+        logging.error(f"❌ An error occurred while connecting to Neo4j: {e}", exc_info=True)
+        raise
+
+connect_to_neo4j()
+
+# Text Processing
+def load_and_clean_text(file_path: str) -> str:
+    with open(file_path, 'r', encoding='utf-8') as file:
+        text = file.read()
+    text = re.sub(r'\n+', ' ', text)
+    return re.sub(r'\s+', ' ', text).strip().lower()
+
+# TF-IDF Filtering
+def compute_tfidf_keywords(text: str, top_n=60):
+    vectorizer = TfidfVectorizer(stop_words='english')
+    X = vectorizer.fit_transform([text])
+    scores = zip(vectorizer.get_feature_names_out(), X.toarray()[0])
+    sorted_scores = sorted(scores, key=lambda x: x[1], reverse=True)
+    return {word for word, _ in sorted_scores[:top_n]}
+
+# Triple Extraction
+def get_full_phrase(token) -> str:
+    return ' '.join(tok.text for tok in token.subtree if tok.dep_ != 'punct').strip()
+
+def extract_rich_triples(doc, tfidf_keywords) -> list:
+    triples = []
+    for sent in doc.sents:
+        subjects = [tok for tok in sent if "subj" in tok.dep_]
+        objects = [tok for tok in sent if "obj" in tok.dep_]
+        verbs = [tok for tok in sent if tok.pos_ == "VERB"]
+        for subj in subjects:
+            for obj in objects:
+                for verb in verbs:
+                    s = get_full_phrase(subj)
+                    o = get_full_phrase(obj)
+                    if s.lower() in tfidf_keywords or o.lower() in tfidf_keywords:
+                        triples.append((s, verb.lemma_, o))
+    return triples
+
+# Graph Visualization
+def visualize_knowledge_graph(triples: list, title: str = "Knowledge Graph"):
+    G = nx.DiGraph()
+    for subj, pred, obj in triples:
+        G.add_node(subj, label='Subject')
+        G.add_node(obj, label='Object')
+        G.add_edge(subj, obj, label=pred)
+
+    pos = nx.spring_layout(G, k=1.2, seed=42)
+    node_colors = ['skyblue' if G.nodes[n]['label'] == 'Subject' else 'lightgreen' for n in G.nodes]
+
+    plt.figure(figsize=(16, 16))
+    nx.draw(G, pos, with_labels=True, node_size=1200, node_color=node_colors,
+            font_size=10, font_weight='bold', edge_color='gray', alpha=0.8)
+    nx.draw_networkx_edge_labels(G, pos, edge_labels={(u, v): d['label'] for u, v, d in G.edges(data=True)},
+                                 font_size=8, font_color='red')
+    plt.title(title, fontsize=20)
+    plt.show()
+
+# Streamlit UI
+st.set_page_config(page_title="Knowledge Graph Generator", layout="wide")
+st.title("🧠 Knowledge Graph Generator")
+
+text_input = st.text_area("Paste your text here", height=200)
+
+if st.button("Generate Graph"):
+    if text_input:
+        try:
             triples = extract_rich_triples(nlp(text_input), compute_tfidf_keywords(text_input))
             logging.info(f"🧠 Extracted {len(triples)} filtered triples.")
             for t in triples[:10]:
                 st.write("🔗", t)
 
-            # Push to Neo4j
-            with driver.session() as session:
-                for subj, pred, obj in triples:
-                    session.execute_write(create_entity, subj)
-                    session.execute_write(create_entity, obj)
-                    session.execute_write(create_relationship, subj, pred, obj)
-
-            logging.info("📡 Triples successfully stored in Neo4j.")
-            st.success("Triples successfully stored in Neo4j.")
-
             # Final Visualization
             visualize_knowledge_graph(triples, title="Filtered Knowledge Graph (TF-IDF)")
-        else:
-            st.warning("Please enter some text.")
-
-except Exception as e:
-    logging.error(f"❌ An error occurred: {e}", exc_info=True)
-    st.error("An error occurred. Please check the logs.")
-
-finally:
-    if driver:
-        close_driver()
+        except Exception as e:
+            logging.error(f"❌ An error occurred: {e}", exc_info=True)
+            st.error("An error occurred. Please check the logs.")
+    else:
+        st.warning("Please enter some text.")