feat(tools): 2.0 phase 5 wave 4 - office, ebooks, data, archives (14 tools) (#224)

tests/unit/doc-engine/libreoffice.test.ts

@@ -0,0 +1,27 @@
+import { parseConvertTarget } from "@snapotter/doc-engine";
+import { describe, expect, it } from "vitest";
+
+describe("parseConvertTarget", () => {
+  it("returns a bare extension unchanged", () => {
+    const result = parseConvertTarget("pdf");
+    expect(result).toEqual({ ext: "pdf", convertTo: "pdf" });
+  });
+
+  it("splits on the first colon for qualified filter targets", () => {
+    const result = parseConvertTarget("docx:MS Word 2007 XML");
+    expect(result).toEqual({ ext: "docx", convertTo: "docx:MS Word 2007 XML" });
+  });
+
+  it("handles filter names with colons after the first", () => {
+    const result = parseConvertTarget("csv:Text - txt - csv (StarCalc):44,34,76,1");
+    expect(result).toEqual({
+      ext: "csv",
+      convertTo: "csv:Text - txt - csv (StarCalc):44,34,76,1",
+    });
+  });
+
+  it("handles single-character extensions", () => {
+    const result = parseConvertTarget("a:SomeFilter");
+    expect(result).toEqual({ ext: "a", convertTo: "a:SomeFilter" });
+  });
+});

tests/unit/doc-engine/pandoc.test.ts

@@ -0,0 +1,34 @@
+import { readFileSync } from "node:fs";
+import { mkdtemp, rm, writeFile } from "node:fs/promises";
+import { tmpdir } from "node:os";
+import { join } from "node:path";
+import { pandocAvailable, runPandoc } from "@snapotter/doc-engine";
+import { afterAll, beforeAll, describe, expect, it } from "vitest";
+
+describe("pandocAvailable", () => {
+  it("returns a boolean without throwing", () => {
+    expect(typeof pandocAvailable()).toBe("boolean");
+  });
+});
+
+describe.skipIf(!pandocAvailable())("runPandoc (requires pandoc)", () => {
+  let tmpDir: string;
+
+  beforeAll(async () => {
+    tmpDir = await mkdtemp(join(tmpdir(), "pandoc-test-"));
+  });
+
+  afterAll(async () => {
+    await rm(tmpDir, { recursive: true, force: true }).catch(() => {});
+  });
+
+  it("converts a trivial markdown file to html", async () => {
+    const mdPath = join(tmpDir, "test.md");
+    const htmlPath = join(tmpDir, "test.html");
+    await writeFile(mdPath, "# Hello\n\nWorld\n");
+    await runPandoc(mdPath, htmlPath, { extraArgs: ["--standalone"] });
+    const html = readFileSync(htmlPath, "utf8");
+    expect(html).toContain("Hello");
+    expect(html).toContain("<");
+  }, 30_000);
+});