Architecture

The project is split into a reusable library layer and a thin CLI.

Module map

• allure_report_exporter/cli.py
• Argument parsing and orchestration.
• allure_report_exporter/allure_cli.py
• Allure binary detection, version check, and allure generate execution.
• allure_report_exporter/parse_widgets.py
• Parsing widgets/*.json and data/test-cases/*.json.
• allure_report_exporter/model.py
• Dataclasses for internal report model.
• allure_report_exporter/render_html.py
• Jinja2 render of print-friendly HTML report.
• allure_report_exporter/pdf.py
• Playwright PDF rendering from local print HTML.
• allure_report_exporter/docx.py
• DOCX generation via python-docx.
• allure_report_exporter/assets/
• Default CSS and HTML template.

Data flow

1. Generate official Allure HTML output.
2. Parse report widgets and test-case JSON.
3. Build normalized in-memory model.
4. Render print HTML for review/debugging and PDF generation.
5. Render DOCX from the same model.

Design constraints

• Avoid parsing raw Allure result internals directly.
• Tolerate missing widget files and missing attachments.
• Keep behavior deterministic for CI usage.
• Keep CLI thin to support future integrations (for example, editor extension UI).