CLDR-18745 cldr_rule_classifier.py

See unicode-org#4951

Author: preetsojitra2712

tools/scripts/llm/cldr_rule_classifier.py

@@ -0,0 +1,125 @@
+import json
+import os
+import re
+from jira import JIRA
+
+def classify_ticket_with_rules_advanced(title, description):
+    """
+    Classifies a Jira ticket using an advanced rule-set with
+    regex, a scoring system, and specific keyword mapping.
+    """
+    text = (title + " " + (description or "")).lower()
+
+    # --- Initialize default classification and scoring ---
+    classification = {
+        "is_cldr_related": True,
+        "ticket_type": "task",
+        "component": "other",
+        "needs_engineering_work": False,
+        "needs_language_specialist": False,
+        "is_potential_duplicate": False,
+        "potential_duplicate_of": None,
+        "summary": "Classification based on advanced rules."
+    }
+    scores = {"priority": 0}
+
+    # --- Rule 1: Use Regex to find duplicate ticket keys ---
+    duplicate_match = re.search(r"cldr-\d+", text)
+    if duplicate_match:
+        classification["is_potential_duplicate"] = True
+        classification["potential_duplicate_of"] = duplicate_match.group(0).upper()
+
+    # --- Rule 2: Scoring System for Priority ---
+    priority_keywords = {
+        "critical": ["critical", "crash", "urgent", "blocker", "unusable"],
+        "high": ["important", "major", "severe"],
+        "low": ["trivial", "minor", "cosmetic"]
+    }
+    for word in priority_keywords["critical"]:
+        if word in text: scores["priority"] += 20
+    for word in priority_keywords["high"]:
+        if word in text: scores["priority"] += 10
+    for word in priority_keywords["low"]:
+        if word in text: scores["priority"] -= 5
+
+    # --- Rule 3: Deduce Component from Keywords ---
+    component_keywords = {
+        "plural-rules": ["plural", "plurals", "cardinal", "ordinal"],
+        "date-time-formats": ["date", "time", "datetime", "timezone", "calendar", "format"],
+        "units": ["unit", "measurement", "gallon", "meter", "km", "celsius"],
+        "locale-data": ["locale", "country", "language", "territory", "subdivision"],
+        "charts": ["chart", "graph", "visualization"]
+    }
+    for component, keywords in component_keywords.items():
+        if any(word in text for word in keywords):
+            classification["component"] = component
+            break
+
+    # --- Rule 4: Determine Ticket Type ---
+    if any(word in text for word in ["error", "bug", "fails", "broken", "crash"]):
+        classification["ticket_type"] = "bug"
+    elif any(word in text for word in ["add", "create", "implement", "feature", "proposal"]):
+        classification["ticket_type"] = "feature"
+    elif any(word in text for word in ["improve", "enhance", "update", "refactor"]):
+        classification["ticket_type"] = "enhancement"
+
+    # --- Rule 5: Set Routing Group based on content ---
+    routing_keywords = {
+        "CLDR Design WG": ["structure", "api", "icu", "icu4x", "design", "proposal", "architecture"],
+        "PMs": ["data", "accurate", "incorrect", "validate", "country", "language", "locale"]
+    }
+    classification["routing_group"] = "CLDR Ops"
+    if any(word in text for word in routing_keywords["CLDR Design WG"]):
+        classification["routing_group"] = "CLDR Design WG"
+    elif any(word in text for word in routing_keywords["PMs"]):
+        classification["routing_group"] = "PMs"
+        
+    # --- Make final decisions based on scores ---
+    if scores["priority"] > 15:
+        classification["priority"] = "critical"
+    elif scores["priority"] > 5:
+        classification["priority"] = "high"
+    elif scores["priority"] < 0:
+        classification["priority"] = "low"
+    else:
+        classification["priority"] = "medium"
+
+    return {
+        "classification_status": "valid_ticket",
+        "details": classification
+    }
+
+# --- Main execution block ---
+if __name__ == "__main__":
+    # --- Configuration for Jira API Connection ---
+    # ⚠ Warning: Hardcoding credentials is a security risk.
+    JIRA_SERVER = "https://unicode-org.atlassian.net"
+    JIRA_USER_EMAIL = "EMAIL!!"  # <-- Replace with your email
+    JIRA_API_TOKEN = "API"    # <-- Replace with your API token
+
+    # --- Get Jira Ticket Key from User Input ---
+    TICKET_KEY_TO_FETCH = input("Please enter the Jira Ticket Key (e.g., CLDR-12345): ")
+    
+    print(f"\n--- Connecting to Jira server at: {JIRA_SERVER} ---")
+    try:
+        jira_connection = JIRA(server=JIRA_SERVER, basic_auth=(JIRA_USER_EMAIL, JIRA_API_TOKEN))
+        print(" Connection successful!")
+
+        print(f"\n--- Fetching ticket: {TICKET_KEY_TO_FETCH} ---")
+        issue = jira_connection.issue(TICKET_KEY_TO_FETCH)
+        ticket_summary = issue.fields.summary
+        ticket_description = issue.fields.description
+        print(f"Title: {ticket_summary}")
+        print(f"Description: {ticket_description if ticket_description else 'No description found.'}")
+
+        # --- Generate and Print the Dynamic Classification ---
+        print("\n" + "="*50)
+        print("--- Dynamic Advanced Rule-Based Classification ---")
+        print("="*50)
+        
+        advanced_classification = classify_ticket_with_rules_advanced(ticket_summary, ticket_description)
+        
+        print(json.dumps(advanced_classification, indent=2))
+
+    except Exception as e:
+        print(f" An error occurred: {e}")