Playwright¶

A transparent wrapper for Playwright that automatically collects comprehensive performance metrics during web automation testing.

Note

The integrations the team actively uses and supports are listed on the front page: Kafka, HTTP and InfluxDB. Playwright is not currently one of them, but it remains available, documented, and open to contributions — if you rely on it or want to improve it, it’s a good candidate.

Overview¶

mite_playwright provides seamless integration with Playwright, automatically capturing detailed performance metrics including DNS timing, TCP/TLS handshake times, paint events, and browser performance data without requiring any changes to your existing Playwright code.

Features¶

Zero Code Changes - Works with existing Playwright scripts
Comprehensive Metrics - Network timing, paint events, page load metrics
Transparent Proxy - Intercepts all Playwright operations automatically
Navigation Focused - Captures metrics for page navigation operations
Performance Optimized - Minimal overhead on test execution
Error Handling - Automatic error tracking and reporting

Quick start¶

Installation¶

pip install playwright
playwright install

Basic Usage¶

from mite_playwright import mite_playwright

@mite_playwright
async def simple_test(ctx, playwright):
    """Simple navigation test with automatic metrics"""
    browser = await playwright.chromium.launch()
    page = await browser.new_page()

    # Metrics automatically collected on navigation
    await page.goto("https://example.com")

    await browser.close()

Mock Context for Testing¶

class MockContext:
    def __init__(self):
        self.config = {}

    def send(self, metric_name, **kwargs):
        print(f"METRIC: {metric_name}")
        for key, value in kwargs.items():
            if isinstance(value, float):
                print(f"   {key}: {value:.4f}")
            else:
                print(f"   {key}: {value}")

# Run your test
ctx = MockContext()
asyncio.run(simple_test(ctx))

Collected Metrics¶

Network Timing Metrics¶

Metric	Description	Unit
`dns_lookup_time`	DNS resolution time	seconds
`tcp_time`	TCP connection establishment	seconds
`tls_time`	TLS handshake time (HTTPS only)	seconds
`time_to_first_byte`	Time until first response byte	seconds
`time_to_last_byte`	Time until response complete	seconds
`total_time`	Complete request duration	seconds

Response Metrics¶

Metric	Description
`url`	Requested URL
`status`	HTTP status code
`response_size`	Content-Length in bytes
`request_start_time`	Request start timestamp
`response_start_time`	Response start timestamp
`response_end_time`	Response end timestamp

Paint & Performance Metrics¶

Metric	Description	Unit
`first_paint`	First paint event	seconds
`first_contentful_paint`	First contentful paint	seconds
`js_onload_time`	JavaScript onload complete	seconds
`render_time`	Page render time	seconds
`time_to_interactive`	Time to interactive	seconds
`page_weight`	Total page size including resources	bytes

Error Tracking¶

The wrapper automatically captures and reports errors for:

Navigation failures
Timeout errors
Element not found errors
Any Playwright exceptions

These are sent as playwright_error, with the method name and error details.

File Structure¶

mite_playwright/
├── __init__.py                 # Main wrapper implementation
├── mite_playwright_example.py  # Usage examples
├── stats.py                    # Statistics utilities
└── README.md                   # Short pointer to this page

Troubleshooting¶

Common Issues¶

No metrics collected:

Ensure you’re using navigation methods (goto, reload, etc.)
Check that responses have valid timing data
Verify your context implements the send() method

Import errors:

Install Playwright: pip install playwright
Install browsers: playwright install