Refactor: Break generate_report_task into a staged pipeline

[techmore]

Type

refactor

Severity

medium

Area

nmapui/workflows.py — generate_report_task()

Description

generate_report_task is approximately 350 lines handling: input validation, customer resolution, scan rules, folder creation, chunked nmap scanning with recursive subnet splitting, XML merging, HTML conversion (two variants), PDF generation, desktop file copy, metadata persistence, statistics extraction, Google Drive upload orchestration, and result emission.

This makes it extremely difficult to:

• Test individual stages in isolation
• Retry a specific failed stage (e.g., just re-upload to Google Drive)
• Add new output formats or destinations
• Reason about error recovery

Additionally, _scan_subnets_with_fallback can recurse to depth 4, creating up to 4096 sequential nmap invocations for a large subnet with persistent timeouts, with no overall scan count budget.

Proposed Fix

Extract into a pipeline of smaller functions:

def generate_report_task(context, sid, data):
    params = _resolve_report_params(context, sid, data)  # validation, customer, rules
    scan_result = _execute_scan_phase(context, sid, params)  # nmap + merge
    report_result = _generate_report_artifacts(context, sid, scan_result)  # HTML + PDF
    _persist_and_distribute(context, sid, report_result)  # metadata, Drive, desktop copy

Add a global scan count budget to _scan_subnets_with_fallback (e.g., max 100 total invocations).

Related Issues

#145 (Break reporting.py into rendering, artifact, naming, and diff services)
#166 (Reporting module modularization)