feat(fallbacks): add priority font acquisition sources

packages/fallbacks/README.md

@@ -111,7 +111,7 @@ The full structured rows are exported as `SUBSTITUTION_EVIDENCE` for richer repo
 
 These maintainer tools use ignored `.cache` files and are not shipped in the package.
 
-`bun run acquire` downloads open-font candidates into `.cache/sources`. Sources come in two shapes: zip archives and pinned source trees. Set `DOCFONTS_SOURCE_CACHE` to use another cache directory, or pass `--source google-fonts` to acquire one source.
+`bun run acquire` downloads open-font candidates into `.cache/sources`. Sources come in two shapes: release archives (zip or tar.gz) and pinned source trees. Set `DOCFONTS_SOURCE_CACHE` to use another cache directory, or pass `--source google-fonts` to acquire one source.
 
 `bun run compare` checks a private reference font against acquired OTF/TTF candidates and prints a ranked Latin advance-width table. It writes no fonts, paths, or results to the tree.
 

packages/fallbacks/acquire.test.ts

@@ -22,6 +22,8 @@ describe("source acquisition catalog", () => {
       "AGPL-3.0-FE",
       "Apache-2.0",
       "UFL-1.0",
+      "GPL-2.0-FE",
+      "Bitstream-Vera-DejaVu",
     ]);
     for (const source of SOURCE_RELEASES) {
       expect(source.sourceId).toMatch(/^[a-z0-9]+(-[a-z0-9]+)*$/);
@@ -38,6 +40,8 @@ describe("source acquisition catalog", () => {
         expect(allowedLicenses.has(source.licenseFamily)).toBe(true);
         expect(source.downloadUrl.startsWith("https://")).toBe(true);
         expect(source.expectedFiles.length).toBeGreaterThan(0);
+        if (source.archiveFormat !== undefined)
+          expect(["zip", "tar.gz"]).toContain(source.archiveFormat);
       }
     }
   });
@@ -111,6 +115,69 @@ describe("source acquisition catalog", () => {
     ]);
   });
 
+  test("carries the priority Liberation, Selawik, and DejaVu sources", () => {
+    const byId = new Map(
+      SOURCE_RELEASES.map((source) => [source.sourceId, source]),
+    );
+    for (const id of [
+      "liberation-fonts",
+      "liberation-sans-narrow",
+      "selawik",
+      "dejavu",
+    ])
+      expect(byId.has(id)).toBe(true);
+  });
+
+  test("declares tar.gz sources and defaults the rest to zip", () => {
+    const byId = new Map(
+      SOURCE_RELEASES.map((source) => [source.sourceId, source]),
+    );
+    const archiveFormatOf = (id: string) => {
+      const source = byId.get(id);
+      if (!source || source.kind === "github-tree") return undefined;
+      return source.archiveFormat ?? "zip";
+    };
+
+    expect(archiveFormatOf("liberation-fonts")).toBe("tar.gz");
+    expect(archiveFormatOf("liberation-sans-narrow")).toBe("tar.gz");
+    expect(archiveFormatOf("selawik")).toBe("zip");
+    expect(archiveFormatOf("dejavu")).toBe("zip");
+    // Pre-existing sources never set the field, so they keep extracting as zip.
+    expect(archiveFormatOf("urw-base35")).toBe("zip");
+    expect(byId.get("urw-base35")).not.toHaveProperty("archiveFormat");
+  });
+
+  test("anchors every Selawik 1.01 TTF, WOFF, and WOFF2 member", () => {
+    const source = SOURCE_RELEASES.find(
+      (candidate) => candidate.sourceId === "selawik",
+    );
+    expect(source?.kind !== "github-tree").toBe(true);
+    if (!source || source.kind === "github-tree") return;
+
+    // Selawik 1.01 ships five weights, each as .ttf, .woff, and .woff2.
+    const weights = ["selawk", "selawkb", "selawkl", "selawksb", "selawksl"];
+    const expected = weights.flatMap((stem) => [
+      `${stem}.ttf`,
+      `${stem}.woff`,
+      `${stem}.woff2`,
+    ]);
+    expect([...source.expectedFiles].sort()).toEqual([...expected].sort());
+  });
+
+  test("uses the GPL font exception and DejaVu license families", () => {
+    const byId = new Map(
+      SOURCE_RELEASES.map((source) => [source.sourceId, source]),
+    );
+    const narrow = byId.get("liberation-sans-narrow");
+    const dejavu = byId.get("dejavu");
+    expect(narrow?.kind !== "github-tree" && narrow?.licenseFamily).toBe(
+      "GPL-2.0-FE",
+    );
+    expect(dejavu?.kind !== "github-tree" && dejavu?.licenseFamily).toBe(
+      "Bitstream-Vera-DejaVu",
+    );
+  });
+
   test("spans more than one project and license family", () => {
     expect(
       new Set(SOURCE_RELEASES.map((source) => source.project)).size,

packages/fallbacks/compare.test.ts

@@ -1,4 +1,5 @@
 import { describe, expect, test } from "bun:test";
+import { execFileSync } from "node:child_process";
 import { mkdirSync, mkdtempSync, rmSync, writeFileSync } from "node:fs";
 import { tmpdir } from "node:os";
 import { join } from "node:path";
@@ -465,3 +466,54 @@ describe("collectCandidates (GitHub tree sources)", () => {
     }
   });
 });
+
+describe("collectCandidates (archive sources)", () => {
+  test("reads tar.gz archive sources from the acquire snapshot format", () => {
+    const cacheDir = mkdtempSync(join(tmpdir(), "docfonts-tar-source-"));
+    try {
+      const sourceId = "tar-example";
+      const root = join(cacheDir, "archive-root");
+      mkdirSync(join(root, "fonts"), { recursive: true });
+      writeFileSync(
+        join(root, "fonts", "Example-Regular.ttf"),
+        syntheticFont(),
+      );
+      writeFileSync(
+        join(root, "fonts", "Example.woff"),
+        new Uint8Array([1, 2]),
+      );
+      execFileSync(
+        "tar",
+        [
+          "-czf",
+          join(cacheDir, `${sourceId}.tar.gz`),
+          "-C",
+          root,
+          "fonts/Example-Regular.ttf",
+          "fonts/Example.woff",
+        ],
+        { stdio: "ignore" },
+      );
+
+      const source: SnapshotSource = {
+        sourceId,
+        family: "Example",
+        targetFamilies: ["Some Proprietary"],
+        kind: "archive",
+        archiveFormat: "tar.gz",
+      };
+
+      const candidates = collectCandidates(source, cacheDir);
+      expect(candidates.map((candidate) => candidate.file)).toEqual([
+        "Example-Regular.ttf",
+      ]);
+      const score = scoreAdvances(
+        sampleMetrics(parseFont(syntheticFont())),
+        sampleMetrics(parseFont(candidates[0].bytes)),
+      );
+      expect(score.tier).toBe("metric_safe");
+    } finally {
+      rmSync(cacheDir, { recursive: true, force: true });
+    }
+  });
+});