Replace scan extraction with @dittowords/text-extract

jest.config.ts

@@ -2,9 +2,10 @@ import type { Config } from "jest";
 
 const config: Config = {
   transformIgnorePatterns: [],
+  // globby v16 is ESM-only. Jest resolves its `unicorn-magic/node` subpath
+  // import to the ESM entry, which babel-jest cannot load.
   moduleNameMapper: {
-    "^unicorn-magic/node$":
-      "<rootDir>/node_modules/unicorn-magic/node.js",
+    "^unicorn-magic/node$": "<rootDir>/node_modules/unicorn-magic/node.js",
   },
   maxWorkers: 1,
   verbose: true,

lib/ditto.ts

@@ -25,6 +25,6 @@ const main = async () => {
   }
 };
 
-export type * from "./src/scan/types";
+export type * from "@dittowords/text-extract";
 
 main();

lib/src/commands/scan.ts

@@ -5,8 +5,11 @@ import { prompt } from "enquirer";
 import open from "open";
 
 import logger from "../utils/logger";
-import { DittoScanExtractSummary, runExtract } from "../scan/extract";
-import { DittoScanCandidate } from "../scan/types";
+import {
+  DittoScanCandidate,
+  DittoScanExtractSummary,
+  runExtract,
+} from "@dittowords/text-extract";
 import { quit } from "../utils/quit";
 import initAPIToken from "../services/apiToken/initAPIToken";
 import appContext from "../utils/appContext";
@@ -91,6 +94,13 @@ function logExtractSummary(
       `  files.skipped_minified: ${summary.filesSkippedMinified}\n`
     );
   }
+  // A failed file is dropped from the results, so its strings are missing.
+  if (summary.filesFailed > 0) {
+    process.stderr.write(`  files.failed: ${summary.filesFailed}\n`);
+    for (const f of summary.failures) {
+      process.stderr.write(`    ${f.file} (${f.language}): ${f.message}\n`);
+    }
+  }
   process.stderr.write(
     `[ditto-cli scan][extract] emitted ${
       summary.candidatesEmitted

lib/src/http/scan.ts

@@ -1,7 +1,7 @@
 import axios, { AxiosError } from "axios";
 import getHttpClient from "./client";
 import { IInitiateScanResponse, ZInitiateScanResponse } from "./types";
-import { DittoScanCandidate } from "../scan/types";
+import { DittoScanCandidate } from "@dittowords/text-extract";
 import DittoError, { ErrorType } from "../utils/DittoError";
 import { Blob } from "buffer";
 

lib/src/scan/extract.test.ts

@@ -0,0 +1,28 @@
+import path from "path";
+
+import { runExtract } from "@dittowords/text-extract";
+
+// The extraction itself is the library's to test. This only checks that the
+// CLI resolves the package and gets usable candidates back from a real
+// directory, so a bad dependency or a changed entry point fails here.
+describe("runExtract", () => {
+  const testfiles = path.resolve(__dirname, "../../../testfiles");
+
+  it("extracts candidates from the test fixtures", async () => {
+    const { candidates, summary } = await runExtract({ inputPath: testfiles });
+
+    expect(summary.filesFailed).toBe(0);
+    expect(summary.filesScanned).toBeGreaterThan(0);
+    expect(candidates.length).toBe(summary.candidatesEmitted);
+
+    const values = candidates.map((c) => c.value_raw);
+    expect(values).toContain("Hello World!");
+
+    // Every candidate carries the fields the CLI reads downstream.
+    for (const c of candidates) {
+      expect(typeof c.id).toBe("string");
+      expect(c.location.file).not.toBe("");
+      expect(path.isAbsolute(c.location.file)).toBe(false);
+    }
+  });
+});