feat(cli): real progress bars for indexing

[r0h1tb]

Closes #5

Uses the existing rich dependency rather than adding tqdm.

Parsing

Was console.status() — an indeterminate spinner. Now a bar with count, elapsed and ETA, plus the current file name, which is what the issue asks for.

Embeddings — the phase that actually needed this

build_embeddings() ran behind a single "Building embeddings..." spinner and emitted nothing until it finished. On a first run, which also downloads the model, that's many minutes of apparent hang. When I hit this I ended up querying Qdrant directly to work out whether the process was alive or wedged.

It now accepts an optional progress_callback(done, total), invoked after each batch, and the CLI renders it as a bar:

def build_embeddings(
    self,
    nodes: list[ASTNode],
    batch_size: int = 64,
    progress_callback: Optional[Callable[[int, int], None]] = None,
) -> int:

Default None, so existing callers are unaffected — this is additive.

One deliberate detail: the callback also fires once with (0, 0) when nothing is embeddable, so a caller driving a UI always gets a terminal update rather than a bar that never resolves.

Bars are transient=True, so they clear on completion and the existing summary lines remain the only residue.

Tests

4 new tests in tests/test_indexing_progress.py:

• callback completion is monotonic, ends at (total, total), and total doesn't shift mid-run
• omitting the callback preserves the original return value
• the empty-input terminal update
• _index_progress() actually carries a count column and an ETA (the issue explicitly asks for a file count)

The rendering itself isn't asserted — rich suppresses live output when not attached to a TTY, so a test would be asserting nothing. Instead the tests pin the mechanism the bar is driven by. I verified the visible behaviour by re-indexing a 105-file Spring Boot project end to end.

Verification

ruff check ast_rag/ tests/     All checks passed!
ruff format --check            76 files already formatted
pytest tests/                  3 failed, 178 passed, 2 xfailed

Baseline on main is 3 failed, 174 passed — same three pre-existing failures (fixed in #51), plus the 4 new tests.

Independent of #50, #51, #52 and #53; branches from main. CI here will be red at the Lint with ruff step until #51 lands — that's the pre-existing breakage, not this change.

[lexasub]

Good pr