ureport-pytest-reporter 1.0.1

Pytest plugin for UReport — ships test results automatically

ureport-pytest-reporter

A pytest plugin that automatically ships test results to UReport after every test run. Zero code changes required in your tests — configure once and it hooks into pytest's lifecycle automatically.

Installation

pip install ureport-pytest-reporter

Quick Start

Add the following to your pytest.ini or pyproject.toml:

# pytest.ini
[pytest]
ureport_server_url = http://your-ureport-server:4100
ureport_api_token  = your-api-token
ureport_product    = MyProduct
ureport_type       = E2E

# pyproject.toml
[tool.pytest.ini_options]
ureport_server_url = "http://your-ureport-server:4100"
ureport_api_token  = "your-api-token"
ureport_product    = "MyProduct"
ureport_type       = "E2E"

That's it. Run pytest as normal — results are submitted to UReport at the end of the session.

---

Configuration Reference

Required

Option	Env Variable	Description
ureport_server_url	UREPORT_SERVER_URL	UReport server base URL
ureport_api_token	UREPORT_API_TOKEN	API token from UReport → User Settings → API Token
ureport_product	UREPORT_PRODUCT	Product name in UReport (groups tests across builds)
ureport_type	UREPORT_TYPE	Build type: E2E, API, UI, or any custom string

Build Metadata (Optional)

Option	Env Variable	Default	Description
ureport_build_number	UREPORT_BUILD_NUMBER	current timestamp	CI build / run number. Use your CI's build ID (e.g. $GITHUB_RUN_NUMBER) to correlate runs
ureport_team	UREPORT_TEAM	—	Team name for organisational tracking
ureport_browser	UREPORT_BROWSER	—	Browser name (e.g. CHROME, FIREFOX)
ureport_device	UREPORT_DEVICE	—	Device name (e.g. DESKTOP-WINDOWS, MOBILE-PIXEL 5)
ureport_platform	UREPORT_PLATFORM	auto-detected	OS platform (darwin, linux, win32)
ureport_platform_version	UREPORT_PLATFORM_VERSION	auto-detected	OS release version
ureport_stage	UREPORT_STAGE	—	Deployment stage (e.g. staging, prod)
ureport_version	UREPORT_VERSION	—	Application version under test

Behaviour (Optional)

Option	Default	Description
ureport_batch_size	50	Number of tests per POST request to UReport
ureport_include_steps	true	Send step-level detail (steps added via step())
ureport_include_screenshots	true	Embed screenshots as base64 in step payloads
ureport_save_relations	true	Save test metadata (tags, components, teams) for cross-run tracking
ureport_auto_detect_platform	true	Auto-detect platform and platform_version from the OS
ureport_auto_token	true	Auto-derive a failure token (path/to/file.py:line) from the stack trace
ureport_output_file	—	Write the full submitted payload to this JSON file (useful for debugging)
ureport_quick_info_markers	—	Space-separated marker names stored as quick-info (execution-specific, never saved to relations)

Environment Variables

All required and build metadata options can be set via environment variables. Environment variables take priority over ini options. This is the recommended approach for CI/CD pipelines:

export UREPORT_SERVER_URL=http://your-ureport-server:4100
export UREPORT_API_TOKEN=your-api-token
export UREPORT_PRODUCT=MyProduct
export UREPORT_TYPE=E2E
export UREPORT_BUILD_NUMBER=$GITHUB_RUN_NUMBER
export UREPORT_STAGE=staging

---

Adding Steps

Use the step() context manager to add named steps to your tests. Steps appear in UReport's step comparison view and can be nested.

from ureport_reporter import step

def test_user_checkout():
    with step("Login"):
        page.goto("/login")
        page.fill("#email", "[e-mail]")
        page.fill("#password", "secret")
        page.click("button[type=submit]")

    with step("Add item to cart"):
        page.click(".product-card:first-child .add-to-cart")
        with step("Verify cart badge"):          # nested step
            assert page.locator(".cart-count").inner_text() == "1"

    with step("Checkout"):
        page.click(".checkout-btn")
        page.fill("#card-number", "4242424242424242")
        page.click("#pay-now")
        assert page.locator(".order-confirmation").is_visible()

Steps are automatically routed to the correct section based on the pytest phase they run in:

• Steps inside fixtures with autouse/setup scope → setup section
• Steps inside the test body → body section
• Steps inside teardown fixtures → teardown section

---

Attaching Files

Use attach() to attach screenshots, API responses, or any content to the current step (or directly to the test if no step is active).

from ureport_reporter import step, attach

def test_api_response():
    with step("Call /api/users"):
        response = requests.get("/api/users")
        attach("response_body", body=response.content, content_type="application/json")
        attach("request_curl",  body=b"curl -X GET /api/users", content_type="text/x-curl")
        assert response.status_code == 200

def test_ui_screenshot(page):
    page.goto("/dashboard")
    screenshot = page.screenshot()
    attach("dashboard", body=screenshot, content_type="image/png")

Supported content types:

Content-Type	Rendered in UReport as
image/png, image/jpeg	Screenshot thumbnail
application/json	Syntax-highlighted JSON
application/xml, text/xml	Syntax-highlighted XML
text/x-curl	Curl command
text/plain	Plain text

You can also attach from a file path instead of bytes:

attach("report", path="/tmp/report.xml", content_type="application/xml")

---

Markers

Add metadata to tests using pytest markers. This metadata is saved to UReport and used for filtering, grouping, and cross-run tracking.

import pytest

@pytest.mark.ureport_uid("checkout-happy-path")
@pytest.mark.ureport_components("Cart", "Checkout", "Payment")
@pytest.mark.ureport_teams("frontend")
@pytest.mark.ureport_tags("@smoke", "@p1")
def test_checkout():
    ...

Marker	Description
@pytest.mark.ureport_uid("my-uid")	Override the test's UID in UReport. Useful for stable identifiers across file renames
@pytest.mark.ureport_components("A", "B")	Component names this test covers
@pytest.mark.ureport_teams("team-name")	Team responsible for this test
@pytest.mark.ureport_tags("@smoke", "@p1")	Additional tags (native pytest marks like @pytest.mark.smoke are also captured automatically)

Tags from Native pytest Marks

Any native pytest mark without arguments is automatically captured as a tag:

@pytest.mark.smoke     # captured as @smoke
@pytest.mark.regression
def test_login():
    ...

Quick Info Markers

Quick info markers store execution-specific values (e.g. environment URL, run ID) that appear as copy-to-clipboard fields in UReport but are never saved to test relations.

Configure which markers are treated as quick info:

# pytest.ini
ureport_quick_info_markers = env build_url run_id

Then use them in tests:

@pytest.mark.env("staging")
@pytest.mark.build_url("https://ci.example.com/builds/123")
def test_something():
    ...

---

Using the ureport Fixture

Alternatively, use the built-in ureport fixture instead of importing step and attach directly:

def test_something(ureport):
    with ureport.step("Do something"):
        result = do_something()
        ureport.attach("result", body=result.encode(), content_type="application/json")
    assert result == "ok"

---

Retry Support

Works with pytest-rerunfailures out of the box. The final attempt's result is what gets submitted to UReport:

• Final attempt passes after retries → RERUN_PASS
• Final attempt fails after retries → FAIL with is_rerun: true

# pytest.ini
addopts = --reruns 2 --reruns-delay 1

---

CI/CD Example

GitHub Actions

- name: Run tests
  env:
    UREPORT_SERVER_URL: ${{ secrets.UREPORT_SERVER_URL }}
    UREPORT_API_TOKEN:  ${{ secrets.UREPORT_API_TOKEN }}
    UREPORT_PRODUCT:    MyProduct
    UREPORT_TYPE:       E2E
    UREPORT_BUILD_NUMBER: ${{ github.run_number }}
    UREPORT_STAGE:      ${{ github.ref == 'refs/heads/main' && 'prod' || 'staging' }}
  run: pytest

GitLab CI

test:
  variables:
    UREPORT_SERVER_URL: "http://your-ureport-server:4100"
    UREPORT_API_TOKEN: "$UREPORT_TOKEN"
    UREPORT_PRODUCT: "MyProduct"
    UREPORT_TYPE: "E2E"
    UREPORT_BUILD_NUMBER: "$CI_PIPELINE_IID"
  script:
    - pytest

---

Debugging

Inspect the payload

Use ureport_output_file to dump the full payload to a JSON file after the run:

ureport_output_file = ureport-payload.json

Disable the reporter temporarily

Remove or comment out ureport_server_url (or unset UREPORT_SERVER_URL). If none of the four required fields are present, the plugin stays inactive and adds zero overhead.

---

Full Configuration Example

# pytest.ini
[pytest]
ureport_server_url          = http://localhost:4100
ureport_api_token           = abc123
ureport_product             = MyProduct
ureport_type                = E2E
ureport_build_number        = 42
ureport_team                = platform
ureport_browser             = CHROME
ureport_stage               = staging
ureport_version             = 2.1.0
ureport_batch_size          = 100
ureport_include_steps       = true
ureport_include_screenshots = true
ureport_save_relations      = true
ureport_auto_token          = true
ureport_output_file         = ureport-payload.json
ureport_quick_info_markers  = env build_url