feat: add Bitbucket workspace support for PR complexity analysis

[OhadPerryBoomi]

---

This PR was opened using this rule: https://github.com/RiveryIO/rivery-cursor-rules/blob/main/administrative/github-pr.md

---

Description

Adds full Bitbucket workspace support so the complexity analyzer can scan and tag PRs from both GitHub and Bitbucket in a single pipeline.

What changed

• New cli/bitbucket.py — Bitbucket API client: fetch PR diffs, metadata, discover project repos, search merged PRs, post complexity comments (with <!-- [complexity-analyzer] --> marker for idempotency).
• cli/utils.py — parse_pr_url() now accepts both GitHub and Bitbucket URLs (same 3-tuple return — backward-compatible). Added detect_pr_provider().
• cli/csv_handler.py — Added source column (auto-detected from URL, appended at end so existing data is unaffected). pandas.read_csv() in reports handles missing column gracefully.
• cli/main.py — analyze_pr_to_dict() and dry-run path route to GH or BB fetch based on URL. New --bb-project option on batch-analyze.
• cli/batch.py — generate_pr_list_from_bb_project() discovers repos by project UUID, collects merged PRs. Label check/apply routes to BB comments for Bitbucket URLs. Source-filtered incremental fetch (get_max_merged_for_source).
• cli/analyze.py — Both analyze_single_pr() and handle_dry_run() route by provider.
• cli/config.py — get_bitbucket_credentials() reads BITBUCKET_EMAIL + BITBUCKET_API_TOKEN.
• cli/migrate.py — Backfills source column from URL; skips GitHub API enrichment for BB rows.
• scripts/sync-new-prs.sh — Two-pass daily sync: Pass 1 (GitHub via --repos-file), Pass 2 (Bitbucket via --bb-project). Supports SKIP_GH=1 / SKIP_BB=1 env vars.
• complexity-report.csv — Rewritten with source column, sorted by date. Includes 12 initial Bitbucket PR scores.
• Tests — New tests for BB URL parsing, detect_pr_provider(), BB source auto-detection in CSV writer, updated verify test fixture.

Backward compatibility

• parse_pr_url() return type unchanged (3-tuple) — zero caller changes needed.
• source column added at end of CSV schema — existing pd.read_csv() consumers unaffected.
• Running migrate-csv on old CSVs backfills source as "github" automatically.

Testing

• All 38 unit tests pass (test_utils, test_csv_handler, test_verify)
• Live smoke test: fetched BB PR #41 diff + metadata successfully
• Live batch: discovered 9 repos in BB project, analyzed and tagged 12 merged PRs
• BB comments posted with complexity scores and verified via API
• CSV rewritten with source column, reports generate without errors

Usage

# Single BB PR analysis
complexity-cli analyze-pr "https://bitbucket.org/workspace/repo/pull-requests/41"

# Batch: scan entire BB project
complexity-cli batch-analyze --bb-project "workspace/{project-uuid}" --days 14 --output complexity-report.csv --label

# Daily sync (both sources)
./scripts/sync-new-prs.sh --days 14