[Ingest Manager] Remove dupe code between two unpack*ToCache fns (#82801) (#82952)

## Summary

Update `unpackRegistryPackageToCache` to call `unpackArchiveToCache` instead of duplicating much of it.

Now an archive is iterated & put in cache via the same function regardless of its initial source.

package.json

@@ -234,6 +234,7 @@
     "markdown-it": "^10.0.0",
     "md5": "^2.1.0",
     "mime": "^2.4.4",
+    "mime-types": "^2.1.27",
     "mini-css-extract-plugin": "0.8.0",
     "minimatch": "^3.0.4",
     "moment": "^2.24.0",
@@ -849,4 +850,4 @@
     "yargs": "^15.4.1",
     "zlib": "^1.0.5"
   }
-}
+}

x-pack/plugins/ingest_manager/server/services/epm/archive/index.ts

@@ -43,7 +43,7 @@ export async function unpackArchiveToCache(
     await bufferExtractor(archiveBuffer, filter, (entry: ArchiveEntry) => {
       const { path, buffer } = entry;
       // skip directories
-      if (path.slice(-1) === '/') return;
+      if (path.endsWith('/')) return;
       if (buffer) {
         cacheSet(path, buffer);
         paths.push(path);
@@ -51,7 +51,7 @@ export async function unpackArchiveToCache(
     });
   } catch (error) {
     throw new PackageInvalidArchiveError(
-      `Error during extraction of uploaded package: ${error}. Assumed content type was ${contentType}, check if this matches the archive type.`
+      `Error during extraction of package: ${error}. Assumed content type was ${contentType}, check if this matches the archive type.`
     );
   }
 
@@ -59,7 +59,7 @@ export async function unpackArchiveToCache(
   // unpacking a zip file with untarBuffer() just results in nothing.
   if (paths.length === 0) {
     throw new PackageInvalidArchiveError(
-      `Uploaded archive seems empty. Assumed content type was ${contentType}, check if this matches the archive type.`
+      `Archive seems empty. Assumed content type was ${contentType}, check if this matches the archive type.`
     );
   }
   return paths;

x-pack/plugins/ingest_manager/server/services/epm/registry/index.ts

@@ -3,7 +3,7 @@
  * or more contributor license agreements. Licensed under the Elastic License;
  * you may not use this file except in compliance with the Elastic License.
  */
-
+import mime from 'mime-types';
 import semver from 'semver';
 import { Response } from 'node-fetch';
 import { URL } from 'url';
@@ -18,15 +18,15 @@ import {
   RegistrySearchResults,
   RegistrySearchResult,
 } from '../../../types';
+import { unpackArchiveToCache } from '../archive';
 import {
   cacheGet,
-  cacheSet,
   cacheDelete,
   getArchiveFilelist,
   setArchiveFilelist,
   deleteArchiveFilelist,
 } from './cache';
-import { ArchiveEntry, getBufferExtractor } from './extract';
+import { ArchiveEntry } from './extract';
 import { fetchUrl, getResponse, getResponseStream } from './requests';
 import { streamToBuffer } from './streams';
 import { getRegistryUrl } from './registry_url';
@@ -137,22 +137,12 @@ export async function unpackRegistryPackageToCache(
   pkgVersion: string,
   filter = (entry: ArchiveEntry): boolean => true
 ): Promise<string[]> {
-  const paths: string[] = [];
   const { archiveBuffer, archivePath } = await fetchArchiveBuffer(pkgName, pkgVersion);
-  const bufferExtractor = getBufferExtractor({ archivePath });
-  if (!bufferExtractor) {
-    throw new Error('Unknown compression format. Please use .zip or .gz');
+  const contentType = mime.lookup(archivePath);
+  if (!contentType) {
+    throw new Error(`Unknown compression format for '${archivePath}'. Please use .zip or .gz`);
   }
-  await bufferExtractor(archiveBuffer, filter, (entry: ArchiveEntry) => {
-    const { path, buffer } = entry;
-    const { file } = pathParts(path);
-    if (!file) return;
-    if (buffer) {
-      cacheSet(path, buffer);
-      paths.push(path);
-    }
-  });
-
+  const paths: string[] = await unpackArchiveToCache(archiveBuffer, contentType);
   return paths;
 }
 

x-pack/test/ingest_manager_api_integration/apis/epm/install_by_upload.ts

@@ -92,7 +92,7 @@ export default function (providerContext: FtrProviderContext) {
         .send(buf)
         .expect(400);
       expect(res.error.text).to.equal(
-        '{"statusCode":400,"error":"Bad Request","message":"Uploaded archive seems empty. Assumed content type was application/gzip, check if this matches the archive type."}'
+        '{"statusCode":400,"error":"Bad Request","message":"Archive seems empty. Assumed content type was application/gzip, check if this matches the archive type."}'
       );
     });
 
@@ -105,7 +105,7 @@ export default function (providerContext: FtrProviderContext) {
         .send(buf)
         .expect(400);
       expect(res.error.text).to.equal(
-        '{"statusCode":400,"error":"Bad Request","message":"Error during extraction of uploaded package: Error: end of central directory record signature not found. Assumed content type was application/zip, check if this matches the archive type."}'
+        '{"statusCode":400,"error":"Bad Request","message":"Error during extraction of package: Error: end of central directory record signature not found. Assumed content type was application/zip, check if this matches the archive type."}'
       );
     });
 

yarn.lock

@@ -20000,7 +20000,7 @@ mime-db@1.44.0, mime-db@1.x.x, "mime-db@>= 1.40.0 < 2":
   resolved "https://registry.yarnpkg.com/mime-db/-/mime-db-1.44.0.tgz#fa11c5eb0aca1334b4233cb4d52f10c5a6272f92"
   integrity sha512-/NOTfLrsPBVeH7YtFPgsVWveuL+4SjjYxaQ1xtM1KMFj7HdxlBlxeyNLzhyJVx7r4rZGJAZ/6lkKCitSc/Nmpg==
 
-mime-types@^2.1.12, mime-types@^2.1.25, mime-types@^2.1.26, mime-types@~2.1.17, mime-types@~2.1.19, mime-types@~2.1.24:
+mime-types@^2.1.12, mime-types@^2.1.25, mime-types@^2.1.26, mime-types@^2.1.27, mime-types@~2.1.17, mime-types@~2.1.19, mime-types@~2.1.24:
   version "2.1.27"
   resolved "https://registry.yarnpkg.com/mime-types/-/mime-types-2.1.27.tgz#47949f98e279ea53119f5722e0f34e529bec009f"
   integrity sha512-JIhqnCasI9yD+SsmkquHBxTSEuZdQX5BuQnS2Vc7puQQQ+8yiP5AY5uWhpdv4YL4VM5c6iliiYWPgJ/nJQLp7w==