feat: add generic saveHTMLAssets util

src/gcc/assets-list.ts

@@ -1,4 +1,4 @@
-import { saveGitHubAssetList } from "../utils/github/fetch-assets.ts"
+import { saveGitHubAssetList } from "../utils/asset/fetch-github-assets.ts"
 
 /**
  * Generate the list of all releases of a GitHub repository and save it to a json file

src/gcc/gcc.ts

@@ -13,10 +13,10 @@ import { addUpdateAlternativesToRc, installAptPack } from "setup-apt"
 import { installBrewPack } from "setup-brew"
 import { rcOptions } from "../cli-options.js"
 import { setupMacOSSDK } from "../macos-sdk/macos-sdk.js"
+import { loadGitHubAssetList, matchAsset } from "../utils/asset/load-assets.js"
 import { hasDnf } from "../utils/env/hasDnf.js"
 import { isArch } from "../utils/env/isArch.js"
 import { isUbuntu } from "../utils/env/isUbuntu.js"
-import { loadGitHubAssetList, matchAsset } from "../utils/github/load-assets.js"
 import { extract7Zip } from "../utils/setup/extract.js"
 import { type InstallationInfo, type PackageInfo, setupBin } from "../utils/setup/setupBin.js"
 import { setupChocoPack } from "../utils/setup/setupChocoPack.js"

src/llvm/assets-list.ts

@@ -1,9 +1,5 @@
-import { mkdir, readFile, readdir, writeFile } from "fs/promises"
-import he from "he"
-import { DownloaderHelper } from "node-downloader-helper"
-import JsonStringify from "safe-stable-stringify"
-import { saveGitHubAssetList } from "../utils/github/fetch-assets.ts"
-import { compareVersion } from "../utils/setup/version.ts"
+import { saveGitHubAssetList } from "../utils/asset/fetch-github-assets.ts"
+import { saveHTMLAssets } from "../utils/asset/fetch-html-assets.ts"
 
 /**
  * Generate the list of all releases of a GitHub repository and save it to a json file
@@ -18,14 +14,23 @@ async function main() {
   )
 
   // go through https://releases.llvm.org/x.y.z and get all the assets
-  await fetchLLVMOrgReleases()
+  await saveHTMLAssets({
+    htmlDownloadDir: "./src/llvm/assets/",
+    path: "./src/llvm/llvm_org_releases.json",
+    *getAssetVersionAndURL() {
+      for (let major = 1; major <= 9; major++) {
+        for (let minor = 0; minor <= 9; minor++) {
+          for (let patch = 0; patch <= 9; patch++) {
+            const version = `${major}.${minor}.${patch}`
+            yield [version, `https://releases.llvm.org/${version}`] as [string, string]
+          }
+        }
+      }
+    },
+    filterAssets: isAssetArchive,
+  })
 }
 
-main().catch((err) => {
-  console.error(err)
-  process.exit(1)
-})
-
 function isAssetArchive(asset: string): boolean {
   return asset.endsWith("tar.xz")
     || asset.endsWith("zip")
@@ -33,96 +38,7 @@ function isAssetArchive(asset: string): boolean {
     || asset.endsWith("tar.gz")
 }
 
-async function fetchLLVMOrgReleases() {
-  const assetHTMLDir = "./src/llvm/assets/"
-
-  await fetchIndexFiles(assetHTMLDir)
-
-  const assets: Record<string, string[]> = await extractAssetsFromHTML(assetHTMLDir)
-
-  // sort the assets by version
-  const jsonStringify = JsonStringify.configure({
-    deterministic: compareVersion,
-  })
-
-  // write the assets to a json file
-  const data = jsonStringify(assets, null, 2)
-  await writeFile("./src/llvm/llvm_org_releases.json", data)
-}
-
-async function fetchIndexFiles(assetHTMLDir: string) {
-  const promises: Promise<void>[] = []
-
-  await mkdir(assetHTMLDir, { recursive: true })
-
-  for (let major = 1; major <= 9; major++) {
-    for (let minor = 0; minor <= 9; minor++) {
-      for (let patch = 0; patch <= 9; patch++) {
-        promises.push(fetchIndexFile(major, minor, patch, assetHTMLDir))
-      }
-    }
-  }
-  await Promise.all(promises)
-  return assetHTMLDir
-}
-
-async function fetchIndexFile(major: number, minor: number, patch: number, assetHTMLDir: string) {
-  try {
-    const version = `${major}.${minor}.${patch}`
-
-    const dl = new DownloaderHelper(
-      `https://releases.llvm.org/${version}`,
-      assetHTMLDir,
-      {
-        fileName: `${version}.html`,
-        override: {
-          skip: true,
-        },
-      },
-    )
-    dl.on("start", () => {
-      console.log(`Downloading ${version}`)
-    })
-    dl.on("error", (err) => {
-      console.error(`Failed to download ${version}.html: ${err}`)
-    })
-    await dl.start()
-  } catch (err) {
-    console.error(err)
-  }
-}
-
-async function extractAssetsFromHTML(assetHTMLDir: string) {
-  const assets: Record<string, string[]> = {}
-
-  const linkRegex = /href="([^"]+)"/g // match all href in the html
-
-  const indexFiles = await readdir(assetHTMLDir)
-  await Promise.all(indexFiles.map(async (indexFile) => {
-    const version = indexFile.replace(".html", "")
-    if (!(version in assets)) {
-      assets[version] = []
-    }
-
-    // read the html file
-    const body = await readFile(`${assetHTMLDir}/${indexFile}`, "utf8")
-
-    // parse the html via regex
-    let match: RegExpExecArray | null
-    // biome-ignore lint/suspicious/noAssignInExpressions: ignore
-    while ((match = linkRegex.exec(body)) !== null) {
-      const asset = match[1]
-
-      if (isAssetArchive(asset)) {
-        assets[version].push(he.decode(asset))
-      }
-    }
-
-    if (assets[version].length === 0) {
-      // eslint-disable-next-line @typescript-eslint/no-dynamic-delete
-      delete assets[version]
-    }
-  }))
-
-  return assets
-}
+main().catch((err) => {
+  console.error(err)
+  process.exit(1)
+})

src/utils/asset/fetch-html-assets.ts

@@ -0,0 +1,119 @@
+import { mkdir, readFile, readdir, writeFile } from "fs/promises"
+import he from "he"
+import { DownloaderHelper } from "node-downloader-helper"
+import JsonStringify from "safe-stable-stringify"
+import { compareVersion } from "../setup/version.ts"
+
+type Options = {
+  /**
+   * The directory to download the HTML files
+   */
+  htmlDownloadDir: string
+  /**
+   * The path to write the output json file
+   */
+  path: string
+  /**
+   * A generator that returns the version and the URL of the asset to download
+   *
+   * The generator should return a tuple of the version and the URL
+   */
+  getAssetVersionAndURL: () => Generator<[string, string], void, unknown>
+  /**
+   * Filter the assets
+   */
+  filterAssets?: (asset: string) => boolean
+}
+
+/**
+ * Save the assets of the HTML files to a json file
+ *
+ * The assets are extracted from the href of the html files
+ */
+export async function saveHTMLAssets(opts: Options) {
+  await fetchIndexFiles(opts)
+
+  const assets: Record<string, string[]> = await extractAssetsFromHTML(opts)
+
+  // sort the assets by version
+  const jsonStringify = JsonStringify.configure({
+    deterministic: compareVersion,
+  })
+
+  // write the assets to a json file
+  const data = jsonStringify(assets, null, 2)
+  await writeFile(opts.path, data)
+}
+
+async function fetchIndexFiles(opts: Options) {
+  const promises: Promise<void>[] = []
+
+  await mkdir(opts.htmlDownloadDir, { recursive: true })
+
+  for (const [version, url] of opts.getAssetVersionAndURL()) {
+    promises.push(fetchIndexFile(version, url, opts.htmlDownloadDir))
+  }
+
+  await Promise.all(promises)
+}
+
+async function fetchIndexFile(version: string, url: string, htmlDownloadDir: string) {
+  try {
+    const dl = new DownloaderHelper(
+      url,
+      htmlDownloadDir,
+      {
+        fileName: `${version}.html`,
+        override: {
+          skip: true,
+        },
+      },
+    )
+    dl.on("start", () => {
+      console.log(`Downloading ${version}`)
+    })
+    dl.on("error", (err) => {
+      console.error(`Failed to download ${version}.html: ${err}`)
+    })
+    await dl.start()
+  } catch (err) {
+    console.error(err)
+  }
+}
+
+async function extractAssetsFromHTML(opts: Options) {
+  const assets: Record<string, string[]> = {}
+
+  const linkRegex = /href="([^"]+)"/g // match all href in the html
+
+  const indexFiles = await readdir(opts.htmlDownloadDir)
+  await Promise.all(indexFiles.map(async (indexFile) => {
+    const version = indexFile.replace(".html", "")
+    if (!(version in assets)) {
+      assets[version] = []
+    }
+
+    // read the html file
+    const body = await readFile(`${opts.htmlDownloadDir}/${indexFile}`, "utf8")
+
+    // parse the html via regex
+    let match: RegExpExecArray | null
+    // biome-ignore lint/suspicious/noAssignInExpressions: ignore
+    while ((match = linkRegex.exec(body)) !== null) {
+      const asset = match[1]
+
+      if (opts.filterAssets !== undefined && !opts.filterAssets(asset)) {
+        continue
+      }
+
+      assets[version].push(he.decode(asset))
+    }
+
+    if (assets[version].length === 0) {
+      // eslint-disable-next-line @typescript-eslint/no-dynamic-delete
+      delete assets[version]
+    }
+  }))
+
+  return assets
+}