fix(neo4j): write content_hash, so the incremental push actually diffs

[rahlk]

Closes #118. Replaces #119, which GitHub auto-closed when its base branch was deleted on merging #117 — same commit, cherry-picked onto main.

The bug

bolt.ts reads each module's stored content_hash to decide which modules changed:

const res = await s.run("MATCH (m:TSModule) RETURN m._module AS k, m.content_hash AS h");

Nothing wrote it. finalizeAnalysis stripped the field from the wire envelope, and the Neo4j
projection — built from that envelope — omitted it, with project.ts naming the strip as the
reason. The diff compared against NULL for every module and concluded everything had changed, so
every push was a full re-upsert.

Three parts of the system disagreed: schema.ts declared content_hash on :TSModule,
bolt.ts read it, project.ts never wrote it. The contract needed no change — only the
writer was missing, so schema.neo4j.json is untouched.

Parity

codeanalyzer-python is the reference and does this correctly: content_hash is wire payload
(schema/py_schema.py:450), projected (neo4j/project.py:689), read back (neo4j/bolt.py:105).

Verification

Live Neo4j, superset-frontend (1,841 modules):

	result
push 1	1841 modules (1841 changed) — 55s
push 2, unchanged source	1841 modules (0 changed) — 25s
content_hash stored	1841 / 1841 (was 0)

Also confirmed on sample-app that editing exactly one file yields (1 changed).

Suite: 240 pass, 0 fail.

About the test

The regression test asserts every projected :TSModule carries a non-null content_hash, and is
break-checked — reverting the projection line fails it.

That shape is deliberate. The existing bolt test hand-seeded content_hash:'stale' into a fixture
node, exercising the diff against data the real projection could never produce. A test that supplies
the value it is verifying cannot catch this class of bug.

Risk worth naming

This enables a code path that has never run in production — the permanent full re-upsert was
masking anything latent in the incremental route. The superset double-push is evidence it holds at
real scale, but it is a first execution.