feat: Parallelize chunking

[Pringled]

Chunking and BM25 tokenization now run in worker processes when an index build has 200+ files to (re)chunk. The CLI and MCP server use 4 workers by default. There's a SEMBLE_INDEX_WORKERS var to overwrite this, and setting it to 0 turns it off.

Cold build timing on a large repo of 27k files:

	main	new
27k files	~120s	~84s

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
src/semble/cli.py	100.00% <100.00%> (ø)
src/semble/index/create.py	100.00% <100.00%> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[greptile-apps]

Confidence Score: 5/5

[High impact] The PR appears safe to merge; no new blocking issue remains.

_{Reviews (2) · Last reviewed commit: "Drop the Windows stdin workaround now th..." · Reviewed by Greptile}