Fix StreamBuf parser hang on non-xdr:wsDr drawing root (#56)

[senoff]

Summary

Fixes #56.

Root cause: DrawingXform.parseClose() compared the closing tag name against the hardcoded string "xdr:wsDr" and fell to default: return true for everything else. BaseXform.parse() treats false as the loop-exit signal for its for await loop. When a workbook's drawing file uses a non-standard namespace root (e.g. a proprietary or future application that doesn't use the xdr: spreadsheet drawing namespace), parseClose never returns false — so the loop exhausts all SAX events and then blocks waiting for stream data that StreamBuf never delivers, hanging wb.xlsx.load() indefinitely.

Fix (approach b — least invasive): In parseOpen(), when the opening element is not the expected xdr:wsDr, capture the actual tag name into this._rootTag instead of trying to look it up as a child anchor type. In parseClose(), the existing switch already exits on this.tag (xdr:wsDr); the default branch now also returns false when name === this._rootTag, providing the loop-exit signal regardless of what namespace the drawing root element uses.

Regression test: spec/integration/issues/issue-56-streambuf-non-xdr-root.spec.js takes the test-issue-1575.xlsx fixture, replaces xl/drawings/drawing1.xml's <xdr:wsDr .../> root with <some:other xmlns:some="http://example.com/unknown-drawing-ns"/> via JSZip, then asserts that wb.xlsx.load(buffer) completes within 5 seconds. Without the fix the test times out.

Also includes a one-line .prettierrc fix (trailingComma: "all" → "es5") to resolve the prettier↔eslint comma-dangle conflict on function arguments that was preventing the pre-commit hook from passing.

Test results

884 unit passing
201 integration passing  (+1 new regression test)
1 e2e passing

[senoff]

Local test run on this branch — npm install --legacy-peer-deps && npm run test:unit && npm run test:integration:

Tier	This PR	Baseline (master @ 4c9e73c)	Δ
test:unit	884 passing, 1 pending, 0 failing	884 passing, 1 pending, 0 failing	+1 int (this PR adds a regression test)
test:integration	201 passing, 0 failing	200 passing, 0 failing	+1 int (this PR adds a regression test)

No regressions vs baseline. Skipped test:end-to-end + test:jasmine (browser/grunt tiers) — same regression-catching surface for unit + integration on a feature branch like this one.

Posted because the tests.yml workflow at this repo hasn't auto-fired on this PR (GitHub's first-time-contributor approval gate). Just want to make the merge feel less risky if/when you get to it. Happy to wait for the official CI run if you'd prefer to flip the approval gate; this is purely supplementary.

[senoff]

Regenerated per AGENTS.md guidance — see new PR #70. Closing this one in favor of the regenerated version.