#!/usr/bin/env python3 """Validate heroku-to-aws migration-report.html (thin stakeholder report). Required sections: decision-summary, exec-costs, cost-optimization, next-steps. Conditional: what-if-scenarios when scenarios/index.json has ≥2 entries. Footer must contain "draft for review". Also enforces the report's own decision-UX and a11y contract (ported subset of the gcp-to-aws report validator, #223 — gated to what the Heroku one-pager actually emits): - decision-summary must NOT contain any badge-verdict-* pill (the skeleton mandates a typography-first `verdict-headline`; color-only verdict carriers are banned). - when estimation-infra.json declares recommendation.outcome, decision-summary must contain a `verdict-headline` element. - must declare a lang attribute; any must declare scope=col|row; any

must carry aria-label +

. Deliberately NOT enforced (would false-fail the intentionally-thin Heroku report, whose skeleton puts

before the verdict and emits no

/ table ): decision-before-TOC ordering, single-

, per-table , glossary/appendix set. Exit 0 on PASS, 1 on FAIL. Usage: python3 validate-heroku-migration-report.py /path/to/migration-report.html \\ --migration-dir "$MIGRATION_DIR" """ from future import annotations import argparse import json import re import sys from html.parser import HTMLParser from pathlib import Path # Sections required in BOTH report modes. cost-optimization is required in full # mode only (see _required_sections) — the decision pack is pre-execution and does # not carry the optimization table. COMMON_REQUIRED_SECTION_IDS = [ "decision-summary", "exec-costs", ] # The structural difference between modes: full (migration-report.html) ends on a # next-steps list AND carries cost-optimization; decision (decision-report.html) ends # on a decision-cta pointing at Generate (next-steps assumes MIGRATION_GUIDE.md / # terraform/ already exist — they don't yet in decision mode). MODE_REQUIRED_SECTION_ID = { "full": "next-steps", "decision": "decision-cta", } def _required_sections(mode: str) -> list[str]: required = [*COMMON_REQUIRED_SECTION_IDS, MODE_REQUIRED_SECTION_ID[mode]] if mode == "full": required.append("cost-optimization") return required # Section identity/count/fragment come from the stdlib HTML parser, never a # raw-source regex, so that: # - a
that exists ONLY inside an HTML comment (e.g. the # skeleton's commented `` placeholder) # is never counted — HTMLParser routes comment text to handle_comment and # never re-tokenizes it as a tag; # - a section wrapped in an inert subtree (