Skip to main content

Allure uploader

Tool that parses allure report and uploads results to TestY.

User flow

  1. Go to "host"/plugins/allure-megauploader/
  2. Fill out config creation form
    1. Verbose name: name that will be displayed in config choice
    2. Project: project in which testplan is stored where your results will be uploaded
    3. Plan: Already existing plan where results will be uploaded to (if child plan not specified)
    4. Child plan: Name of test plan that will be found/created in Plan from third step
    5. Parameters: Parameters of child plan
    6. Auto suite name: If this field is not empty all created cases will be put in "Actual suite" -> "auto suite name"( example: suite name is Audit and auto suite name is auto cases will be created in suite auto with parent suite Audit
    7. Jira projects: list of prefixes of jira projects separated by comma (example: TST-, STOR-, TMS-) These links are parsed from allure report
    8. Custom attributes: A list of custom attributes parsed from Allure labels, for example: @allure.label('Place', 'Route 66') will produce Test Case with attribute Place and value Route 66.
    9. Envs to parse: list of names of variables separated by end-line. If your report contains env variables you can specify which ones you want, if none specified all variables will be parsed. Variables will be added as attributes to your test result. You can provide replacement name for variables like this IMAGE_VERSION:version. Example input for env vars:
      IMAGE_VERSION:version
      PORT
      SOME_VAR:var
    10. Additional parameters: json field that will be interpreted as key = Attribute name, value = Attribute value and added to every result. (example: {"version":"1.1.0"})
    11. Allowed users: list of selectable users that can view and edit this config
    12. Labels: list of comma separated label names that will be added to found/created cases. This list will replace all labels on cases!
    13. Service: service is select that considers parsing logic for allure report. It is an extension point for teams to implement their own logic in allure uploader. Service records can be added in admin panel only by staff user.
    14. Hit save and you all done with config!
  3. In Navigation bar at the top of page hit parse report.
  4. Choose your config
  5. Choose allure url that will allow to download zip file or upload archive.
  6. If both fields are specified url will be prioritized and archive from local will not be processed.
  7. You will be redirected to page with progress. As soon as it's done hit Upload confirmation
  8. You will see Processing status it can be Allure content was not received yet or timestamp when it was parsed
  9. On confirmation page you will see "Suites to be created" if suite not exist it will be displayed here and all cases will be displayed under suite. "Cases to be created within existing suites" will show which cases will be created in format "suite_name | case_name". If auto suite name specified cases will appear here because suites exists but cases within auto don't
  10. !!!!!You will have 30 minutes to check what elements will be created. In case you ok with changes hit submit and another upload with progress bar will start. In case of decline is pressed temporary parsed report will be erased, and you will have to parse it once again. In case you didn't make it on time parsed report will be erased, and you will have to parse it once again.
  11. In task list you can watch for statuses of tasks launched by user, like parsing task and upload task.
  12. In service list you can check out available parsing services. Feel free to go to dev server and test your reports there.
  13. By default we provide 5 parsing strategies:
    1. Default: Default service for parsing reports, does not take into account suites hierarchy (service_code: 1)
    2. Hierarchy service: Creates hierarchy based on allure suites (service_code: 2)
    3. Unique cases: As default but takes into account historyId from allure, no discoverage by values of cases (service_code: 3)
    4. Hierarchy unique cases: Combination of services 2 and 3 (service_code: 4)
    5. Performance service: Works like strategy 4, but for large datasets like 50k tests in one report, preview for upload is not available (service_code: 5)

Metadata

Release files for allure-megauploader 0.0.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for allure-megauploader 0.0.2
File Size Uploaded Provenance Actions
allure_megauploader-0.0.2.tar.gz 141.0 kB Without provenance Details for allure_megauploader-0.0.2.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for allure-megauploader 0.0.2
File Interpreter ABI Platform Actions
allure_megauploader-0.0.2-py3-none-any.whl Python 3 none any Details for allure_megauploader-0.0.2-py3-none-any.whl

Total release size: 297.4 kB

Release files / allure_megauploader-0.0.2.tar.gz

Download URL allure_megauploader-0.0.2.tar.gz
Size 141.0 kB
Tags Source
SHA-256 checksum
How to use checksums
8395fcf844a50e6fc96634d2cca80153de70139bd9fc31b43b27d1adfa0b9477
BLAKE2b-256 checksum
How to use checksums
25eb1226180148e0f89897e93bc83970aa93dc4b16560c2404d8d185358632c0
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.11.9

Release files / allure_megauploader-0.0.2-py3-none-any.whl

Download URL allure_megauploader-0.0.2-py3-none-any.whl
Size 156.3 kB
Tags Python 3
SHA-256 checksum
How to use checksums
43f888a17bb256d80566614cc26169d6151aef37950b9b2456ce96b44b26dc01
BLAKE2b-256 checksum
How to use checksums
22aafad166902155dacd27b4e4fc4c6c6abb2fe218f517840d926a97e19866dc
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.11.9

Security for allure-megauploader 0.0.2

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.0.2 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page