"""
Integration test for Jira tracker synchronization with Preloop.
This test verifies the complete end-to-end flow including:
- Tracker registration
- Initial issue indexing via polling
- Webhook registration and propagation
- Bi-directional sync (Jira -> Preloop, Preloop -> Jira)
- Comment synchronization
- MCP tools integration
- Proper cleanup
Environment Variables Required:
- PRELOOP_TEST_URL: Preloop instance URL
- PRELOOP_TEST_API_KEY: API key for Preloop authentication
- JIRA_URL: Jira instance URL (e.g., https://example.atlassian.net)
- JIRA_API_KEY: Jira API token
- JIRA_USERNAME: Jira username/email
- JIRA_ISSUE_KEY: Test issue in format "PROJECT-123"
- JIRA_ORG_ID: Organization identifier (for scope filtering)
- JIRA_PROJECT_ID: Project key like "PROJ" (for scope filtering)
Usage:
pytest tests/integration/test_tracker_sync_jira.py -v -s
"""
import base64
import json
import os
import re
import time
from urllib.parse import quote
import httpx
import pytest
from tests.integration.test_tracker_sync_common import (
INDEX_TIMEOUT,
PRELOOP_API_KEY,
PRELOOP_URL,
TEST_RUN_ID,
WEBHOOK_PROPAGATION_TIMEOUT,
wait_for_issue,
wait_for_issue_update,
)
# Jira config
JIRA_URL = os.getenv("JIRA_URL", "").rstrip("/")
JIRA_API_KEY = os.getenv("JIRA_API_KEY", "")
JIRA_USERNAME = os.getenv("JIRA_USERNAME", "")
JIRA_ISSUE_KEY = os.getenv("JIRA_ISSUE_KEY", "") # e.g., "PROJECT-123"
JIRA_ORG_ID = os.getenv("JIRA_ORG_ID", "") # Usually the Jira URL domain
JIRA_PROJECT_ID = os.getenv("JIRA_PROJECT_ID", "") # Project key (e.g., "PROJ")
@pytest.fixture(scope="module")
def jira_client():
"""Create Jira API client."""
if not JIRA_API_KEY or not JIRA_USERNAME:
pytest.skip("JIRA_API_KEY and JIRA_USERNAME required")
auth_str = f"{JIRA_USERNAME}:{JIRA_API_KEY}"
encoded_auth = base64.b64encode(auth_str.encode()).decode()
headers = {
"Authorization": f"Basic {encoded_auth}",
"Content-Type": "application/json",
}
with httpx.Client(base_url=JIRA_URL, headers=headers) as client:
yield client
@pytest.mark.integration
@pytest.mark.jira
@pytest.mark.skipif(
not all([JIRA_API_KEY, JIRA_USERNAME, JIRA_ISSUE_KEY, JIRA_URL]),
reason="JIRA_API_KEY, JIRA_USERNAME, JIRA_ISSUE_KEY, and JIRA_URL required",
)
def test_jira_tracker_sync(preloop_client, jira_client):
"""
Complete integration test for Jira tracker synchronization.
This test verifies:
- Tracker registration
- Initial issue indexing via polling
- Webhook registration and propagation
- Bi-directional sync (Jira -> Preloop, Preloop -> Jira)
- Comment synchronization
- MCP tools integration
- Proper cleanup
"""
print("\n" + "=" * 80)
print("JIRA TRACKER SYNC TEST")
print("=" * 80)
# Build scope rules to only sync specific org and project
scope_rules = []
if JIRA_ORG_ID and JIRA_PROJECT_ID:
scope_rules = [
{
"scope_type": "ORGANIZATION",
"rule_type": "INCLUDE",
"identifier": JIRA_ORG_ID,
},
{
"scope_type": "PROJECT",
"rule_type": "INCLUDE",
"identifier": JIRA_PROJECT_ID,
},
]
# Variables for cleanup
tracker_id = None
original_title = None
original_description = None
created_comment_ids = []
try:
# Step 2: Register tracker
print("\n" + "=" * 80)
print("STEP 2: Tracker Registration (Jira)")
print("=" * 80)
request_body = {
"name": f"Jira Test Tracker {TEST_RUN_ID}",
"type": "jira",
"url": JIRA_URL, # URL at top level for Jira
"api_key": JIRA_API_KEY,
"config": {
"username": JIRA_USERNAME,
},
"scope_rules": scope_rules,
}
print("Request body (api_key redacted):")
redacted_body = request_body.copy()
redacted_body["api_key"] = "***REDACTED***"
print(json.dumps(redacted_body, indent=2))
register_response = preloop_client.post(
"/api/v1/trackers",
json=request_body,
)
assert register_response.status_code == 201, (
f"Failed to register Jira tracker (status {register_response.status_code}): {register_response.text}"
)
tracker_data = register_response.json()
tracker_id = tracker_data["id"]
print(f"✓ Registered Jira tracker: {tracker_id}")
print("Response data:")
print(json.dumps(tracker_data, indent=2))
# Step 3: Verify tracker is listed
print("\n" + "=" * 80)
print("STEP 3: Tracker Verification")
print("=" * 80)
list_response = preloop_client.get("/api/v1/trackers")
assert list_response.status_code == 200
trackers = list_response.json()
assert any(t["id"] == tracker_id for t in trackers), "Jira tracker not in list"
print("✓ Jira tracker appears in tracker list")
# Step 5: Wait for initial indexing
print("\n" + "=" * 80)
print("STEP 5: Initial Indexing (Polling)")
print("=" * 80)
issue_data = wait_for_issue(preloop_client, JIRA_ISSUE_KEY, INDEX_TIMEOUT)
original_title = issue_data["title"]
original_description = issue_data.get("description", "")
print(f"✓ Issue indexed: {JIRA_ISSUE_KEY}")
print(f" Original title: {original_title}")
# Step 7: Update issue via Jira API
print("\n" + "=" * 80)
print("STEP 7: External Update (Jira API)")
print("=" * 80)
new_title = f"{original_title} {TEST_RUN_ID}"
new_description = f"{original_description} {TEST_RUN_ID}"
update_response = jira_client.put(
f"/rest/api/3/issue/{JIRA_ISSUE_KEY}",
json={
"fields": {
"summary": new_title,
"description": {
"type": "doc",
"version": 1,
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": new_description}],
}
],
},
}
},
)
update_response.raise_for_status()
print("✓ Updated issue via Jira API")
print(f" New title: {new_title}")
# Step 8: Wait for webhook propagation
print("\n" + "=" * 80)
print("STEP 8: Webhook Propagation Test")
print("=" * 80)
updated_issue = wait_for_issue_update(
preloop_client, JIRA_ISSUE_KEY, new_title, WEBHOOK_PROPAGATION_TIMEOUT
)
assert updated_issue["title"] == new_title
assert TEST_RUN_ID in updated_issue.get("description", "")
print("✓ Webhook propagation successful")
# Step 9: Create comment via Jira API
print("\n" + "=" * 80)
print("STEP 9: Comment Sync Test")
print("=" * 80)
comment_text = f"Test comment {TEST_RUN_ID}"
comment_response = jira_client.post(
f"/rest/api/3/issue/{JIRA_ISSUE_KEY}/comment",
json={
"body": {
"type": "doc",
"version": 1,
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": comment_text}],
}
],
}
},
)
comment_response.raise_for_status()
comment_id = str(comment_response.json()["id"])
created_comment_ids.append(comment_id)
print(f"✓ Created comment via Jira API: {comment_id}")
# Wait for comment to appear in Preloop
print(" Waiting for comment to sync...")
time.sleep(10) # Give webhook time to propagate
issue_with_comments = wait_for_issue(
preloop_client, JIRA_ISSUE_KEY, WEBHOOK_PROPAGATION_TIMEOUT
)
assert any(
comment_text in c.get("body", "")
for c in issue_with_comments.get("comments", [])
), "Comment not synced to Preloop"
print("✓ Comment synced to Preloop")
# Step 10: Update issue via Preloop API (remove test suffix)
print("\n" + "=" * 80)
print("STEP 10: Preloop Update")
print("=" * 80)
# URL-encode the issue key for the PUT request
encoded_issue_key = quote(JIRA_ISSUE_KEY, safe="")
update_response = preloop_client.put(
f"/api/v1/issues/{encoded_issue_key}",
json={
"title": original_title,
"description": original_description,
},
)
assert update_response.status_code == 200, (
f"Failed to update issue via Preloop: {update_response.text}"
)
print("✓ Updated issue via Preloop API")
# Step 11: Verify update propagated to Jira
print("\n" + "=" * 80)
print("STEP 11: Tracker Verification (Jira)")
print("=" * 80)
time.sleep(5) # Give sync time to complete
verify_response = jira_client.get(f"/rest/api/3/issue/{JIRA_ISSUE_KEY}")
verify_response.raise_for_status()
jira_issue = verify_response.json()
assert jira_issue["fields"]["summary"] == original_title, (
f"Title not synced to Jira: got '{jira_issue['fields']['summary']}', expected '{original_title}'"
)
# Note: Jira description is complex format, so we check for the text content
jira_description = ""
if jira_issue["fields"].get("description"):
desc_content = jira_issue["fields"]["description"].get("content", [])
for block in desc_content:
for content in block.get("content", []):
if content.get("type") == "text":
jira_description += content.get("text", "")
assert original_description in jira_description, (
"Description not synced to Jira"
)
print("✓ Update propagated from Preloop to Jira")
# Step 12: Test MCP Tools
print("\n" + "=" * 80)
print("STEP 12: MCP Tools Integration Test")
print("=" * 80)
from tests.integration.helpers import run_mcp_test
async def test_mcp_operations(mcp_client):
"""Test MCP operations using direct client."""
# Test create_issue via MCP
create_title = f"MCP Test Issue {TEST_RUN_ID}"
create_description = f"Issue created via MCP for testing - {TEST_RUN_ID}"
print(f"📝 Creating issue via MCP: {create_title}")
create_result = await mcp_client.create_issue(
project=JIRA_PROJECT_ID,
title=create_title,
description=create_description,
)
# Extract issue key from response
created_issue_key = None
# Try structuredContent first (most reliable) - FastMCP puts Pydantic model fields here
if (
hasattr(create_result, "structuredContent")
and create_result.structuredContent
):
url = create_result.structuredContent.get("url")
if url:
# Parse Jira URL: https://example.atlassian.net/browse/PROJECT-123
url_match = re.search(r"/browse/([A-Z]+-\d+)", url)
if url_match:
created_issue_key = url_match.group(1)
# Fallback to parsing content text - FastMCP puts JSON serialized model here
if not created_issue_key and hasattr(create_result, "content"):
for content_item in create_result.content:
if hasattr(content_item, "text"):
text = content_item.text
# Try to parse JSON and extract URL
try:
import json
data = json.loads(text)
url = data.get("url")
if url:
url_match = re.search(r"/browse/([A-Z]+-\d+)", url)
if url_match:
created_issue_key = url_match.group(1)
break
except json.JSONDecodeError:
# Try direct pattern match in text as final fallback
text_match = re.search(r"([A-Z]+-\d+)", text)
if text_match:
created_issue_key = text_match.group(1)
break
assert created_issue_key, (
f"Failed to extract issue key from MCP response. "
f"Debug info - has structuredContent: {hasattr(create_result, 'structuredContent')}, "
f"has content: {hasattr(create_result, 'content')}"
)
print(f"✓ Created issue via MCP: {created_issue_key}")
# Wait for issue to be indexed
time.sleep(5)
# Test search via MCP
print(f"🔍 Searching via MCP: {create_title}")
search_result = await mcp_client.search(
query=create_title, project=JIRA_PROJECT_ID, limit=10
)
print("✓ Found created issue via MCP search")
# Test update_issue via MCP to close the issue
print(f"✏️ Updating issue via MCP: {created_issue_key}")
await mcp_client.update_issue(created_issue_key, status="Done")
print("✓ Closed issue via MCP update_issue")
# Test get_issue via MCP to verify it's closed
time.sleep(3)
print(f"📄 Getting issue via MCP: {created_issue_key}")
get_result = await mcp_client.get_issue(created_issue_key)
# Verify status is closed
status_found = False
if hasattr(get_result, "content"):
for content_item in get_result.content:
if hasattr(content_item, "text"):
text = content_item.text.lower()
if "closed" in text or "done" in text:
status_found = True
break
assert status_found, "Issue does not appear to be closed in MCP response"
print("✓ Verified issue is closed via MCP get_issue")
return created_issue_key
# Run MCP tests
created_issue_key = run_mcp_test(
PRELOOP_URL, PRELOOP_API_KEY, test_mcp_operations
)
print("✓ MCP integration tests complete")
print("\n✅ Jira tracker sync test PASSED (including MCP)")
finally:
# Cleanup
print("\n" + "=" * 80)
print("CLEANUP")
print("=" * 80)
# Delete created comments
for comment_id in created_comment_ids:
try:
jira_client.delete(
f"/rest/api/3/issue/{JIRA_ISSUE_KEY}/comment/{comment_id}"
)
print(f"✓ Deleted comment: {comment_id}")
except Exception as e:
print(f"✗ Failed to delete comment {comment_id}: {e}")
# Note: We intentionally don't restore the issue title/description
# since the Preloop update in Step 10 already restored them
