feat: display readme from more sources

instead of just github. unfortunately most servers have
CORS enabled so a proxy is needed

src/lib/components/repo/utils.spec.ts

@@ -1,5 +1,5 @@
 import { describe, expect, test } from 'vitest'
-import { extractGithubDetails, selectRepoFromCollection } from './utils'
+import { cloneArrayToReadMeUrls, selectRepoFromCollection } from './utils'
 import {
   collection_defaults,
   event_defaults,
@@ -160,18 +160,92 @@ describe('getSelectedRepo', () => {
   })
 })
 
-describe('extractGithubDetails', () => {
-  ;[
-    'https://github.com/orgname/reponame.git',
-    'https://github.com/orgname/reponame',
-    'git@github.com:orgname/reponame',
-  ].forEach((clone) => {
-    describe(clone, () => {
-      test('returns correct org and repo', () => {
-        const res = extractGithubDetails(clone)
-        expect(res?.org).toEqual('orgname')
-        expect(res?.repo_name).toEqual('reponame')
-      })
+describe('cloneArrayToReadMeUrls', () => {
+  test('for each clone url returns url to /raw/HEAD/README.md and /raw/HEAD/readme.md', () => {
+    expect(
+      cloneArrayToReadMeUrls([
+        'https://gitea.com/orgname/reponame',
+        'https://gitlab.com/orgname/reponame',
+      ])
+    ).toEqual([
+      'https://gitea.com/orgname/reponame/raw/HEAD/README.md',
+      'https://gitea.com/orgname/reponame/raw/HEAD/readme.md',
+      'https://gitlab.com/orgname/reponame/raw/HEAD/README.md',
+      'https://gitlab.com/orgname/reponame/raw/HEAD/readme.md',
+    ])
+  })
+  test('for github link use raw.githubusercontent.com/HEAD', () => {
+    expect(
+      cloneArrayToReadMeUrls(['https://github.com/orgname/reponame'])
+    ).toEqual([
+      'https://raw.githubusercontent.com/HEAD/README.md',
+      'https://raw.githubusercontent.com/HEAD/readme.md',
+    ])
+  })
+  test('for sr.hr link to /blob/HEAD', () => {
+    expect(cloneArrayToReadMeUrls(['https://sr.ht/~orgname/reponame'])).toEqual(
+      [
+        'https://sr.ht/~orgname/reponame/blob/HEAD/README.md',
+        'https://sr.ht/~orgname/reponame/blob/HEAD/readme.md',
+      ]
+    )
+  })
+  test('for git.launchpad.net link to /plain', () => {
+    expect(
+      cloneArrayToReadMeUrls(['https://git.launchpad.net/orgname/reponame'])
+    ).toEqual([
+      'https://git.launchpad.net/orgname/reponame/plain/README.md',
+      'https://git.launchpad.net/orgname/reponame/plain/readme.md',
+    ])
+  })
+  test('for git.savannah.gnu.org link to /plain', () => {
+    expect(
+      cloneArrayToReadMeUrls(['https://git.savannah.gnu.org/orgname/reponame'])
+    ).toEqual([
+      'https://git.savannah.gnu.org/orgname/reponame/plain/README.md',
+      'https://git.savannah.gnu.org/orgname/reponame/plain/readme.md',
+    ])
+  })
+  describe('transform clone address to url', () => {
+    test('strips trailing / from address', () => {
+      expect(
+        cloneArrayToReadMeUrls(['https://codeberg.org/orgname/reponame/'])
+      ).toEqual([
+        'https://codeberg.org/orgname/reponame/raw/HEAD/README.md',
+        'https://codeberg.org/orgname/reponame/raw/HEAD/readme.md',
+      ])
+    })
+    test('strips .git from address', () => {
+      expect(
+        cloneArrayToReadMeUrls(['https://codeberg.org/orgname/reponame.git'])
+      ).toEqual([
+        'https://codeberg.org/orgname/reponame/raw/HEAD/README.md',
+        'https://codeberg.org/orgname/reponame/raw/HEAD/readme.md',
+      ])
+    })
+    test('git@codeberg.org:orgname/reponame.git to address', () => {
+      expect(
+        cloneArrayToReadMeUrls(['git@codeberg.org:orgname/reponame.git'])
+      ).toEqual([
+        'https://codeberg.org/orgname/reponame/raw/HEAD/README.md',
+        'https://codeberg.org/orgname/reponame/raw/HEAD/readme.md',
+      ])
+    })
+    test('ssh://codeberg.org/orgname/reponame to address', () => {
+      expect(
+        cloneArrayToReadMeUrls(['ssh://codeberg.org/orgname/reponame'])
+      ).toEqual([
+        'https://codeberg.org/orgname/reponame/raw/HEAD/README.md',
+        'https://codeberg.org/orgname/reponame/raw/HEAD/readme.md',
+      ])
+    })
+    test('https://custom.com/deep/deeper/deeper to address', () => {
+      expect(
+        cloneArrayToReadMeUrls(['https://custom.com/deep/deeper/deeper'])
+      ).toEqual([
+        'https://custom.com/deep/deeper/deeper/raw/HEAD/README.md',
+        'https://custom.com/deep/deeper/deeper/raw/HEAD/readme.md',
+      ])
     })
   })
 })

src/lib/components/repo/utils.ts

@@ -16,20 +16,56 @@ export const selectRepoFromCollection = (
   })[0]
 }
 
-export const extractGithubDetails = (
-  clone: string
-): { org: string; repo_name: string } | undefined => {
-  if (clone.indexOf('github.') > -1) {
-    const g_split = clone.split('github.')
-    if (g_split.length > 0) {
-      const final = g_split[1].replace(':', '/').split('/')
-      if (final.length > 2) {
-        return {
-          org: final[1],
-          repo_name: final[2].split('.')[0],
-        }
+/** most servers will produce a CORS error so a proxy should be used */
+export const cloneArrayToReadMeUrls = (clone: string[]): string[] => {
+  const addresses = clone.map(extractRepoAddress)
+  /**
+   * at the time of this commit these urls work for:
+   * self-hosted gitea (or forgejo), gitlab
+   * github.com
+   * bitbucket.org
+   * gitlab.org
+   * gitea.com
+   * codeberg.org (forgejo instance)
+   * sourcehut (git.sr.ht)
+   * launchpad.net
+   * It doesnt work for:
+   * self-hosted gogs (requires branch name repo/raw/master/README.md)
+   * sourceforge.net (https://sourceforge.net/p/mingw/catgets/ci/master/tree/README?format=raw)
+   * notabug.org (requires branch name notabug.org/org/repo/raw/master/README.md)
+   */
+  return [
+    ...addresses.flatMap((address) => {
+      let prefix = 'raw/HEAD'
+      if (address.includes('sr.ht')) prefix = 'blob/HEAD'
+      if (
+        address.includes('git.launchpad.net') ||
+        address.includes('git.savannah.gnu.org')
+      )
+        prefix = 'plain'
+      if (address.includes('github.com')) {
+        // raw.githubusercontent.com can be used without CORS error
+        address = address.replace('github.com', 'raw.githubusercontent.com')
+        prefix = 'HEAD'
       }
-    }
-  }
-  return undefined
+      return ['README.md', 'readme.md'].map(
+        (filename) => `https://${address}/${prefix}/${filename}`
+      )
+    }),
+  ]
+}
+
+const extractRepoAddress = (clone_string: string): string => {
+  let s = clone_string
+  // remove trailing slash
+  if (s.endsWith('/')) s = s.substring(0, s.length - 1)
+  // remove trailing .git
+  if (s.endsWith('.git')) s = s.substring(0, s.length - 4)
+  // remove :// and anything before
+  if (s.includes('://')) s = s.split('://')[1]
+  // remove @ and anything before
+  if (s.includes('@')) s = s.split('@')[1]
+  // replace : with /
+  s = s.replace(':', '/')
+  return s
 }

src/lib/stores/repo.ts

@@ -11,7 +11,7 @@ import {
 } from '$lib/components/repo/type'
 import { ensureRepoCollection } from './repos'
 import {
-  extractGithubDetails,
+  cloneArrayToReadMeUrls,
   selectRepoFromCollection,
 } from '$lib/components/repo/utils'
 import { get } from 'svelte/store'
@@ -104,33 +104,34 @@ const ensureRepoReadme = async (
       })
     }
   }
+  let text: string | undefined
   try {
-    const github_details = clone
-      .map(extractGithubDetails)
-      .find((details) => !!details)
-    let res: Response
-    if (github_details) {
+    let readme_urls = cloneArrayToReadMeUrls(clone)
+    // prioritise using github as it doesn't require a proxy
+    readme_urls = [
+      ...readme_urls.filter((url) => url.includes('raw.githubusercontent.com')),
+      ...readme_urls.filter(
+        (url) => !url.includes('raw.githubusercontent.com')
+      ),
+    ]
+    for (let i = 0; i < readme_urls.length; i++) {
       try {
-        res = await fetch(
-          `https://raw.githubusercontent.com/${github_details.org}/${github_details.repo_name}/HEAD/README.md`
+        const res = await fetch(
+          readme_urls[i].includes('raw.githubusercontent.com')
+            ? readme_urls[i]
+            : // use proxy as most servers produce a CORS error
+              `/git_proxy/readme/${encodeURIComponent(readme_urls[i])}`
         )
-        if (!res.ok) {
-          throw 'api request error'
+        if (res.ok) {
+          text = await res.text()
+          break
+        } else {
+          continue
         }
       } catch {
-        res = await fetch(
-          `https://raw.githubusercontent.com/${github_details.org}/${github_details.repo_name}/HEAD/readme.md`
-        )
+        continue
       }
-    } else
-      res = await fetch(`/git_proxy/readme/${encodeURIComponent(clone[0])}`)
-    if (!res.ok) {
-      throw 'api request error'
     }
-    let text = ''
-    text = await res.text()
-    update(text)
-  } catch (e) {
-    update()
-  }
+  } catch {}
+  update(text)
 }

src/routes/git_proxy/readme/[clone]/+server.ts

@@ -1,28 +0,0 @@
-import { extractGithubDetails } from '$lib/components/repo/utils'
-
-export const GET = async ({ params }: { params: { clone: string } }) => {
-  const github_details = extractGithubDetails(decodeURIComponent(params.clone))
-  if (github_details) {
-    const res = await fetch(
-      `https://raw.githubusercontent.com/${github_details.org}/${github_details.repo_name}/HEAD/README.md`
-    )
-    const text = await res.text()
-
-    return new Response(text)
-  } else {
-    // options:
-    //  * add support for different git server implementations that serve raw
-    //    files and cycle through the urls until we find the readme
-    //  * add a worker that can use 'git archive' to get specific files
-    //    * unfortunately the two options that can easily embeded within this
-    //      sveltekit backend (wasm-git and isomorphicgit) don't support the
-    //      'archive' command
-    //      https://github.com/petersalomonsen/wasm-git/
-    //      https://github.com/isomorphic-git
-    //    * 'git clone' is too expensive for retrieving single files. even when
-    //      done using treeless or blobless flags. see:
-    //      https://noise.getoto.net/2020/12/21/get-up-to-speed-with-partial-clone-and-shallow-clone/
-
-    return new Response(null)
-  }
-}

src/routes/git_proxy/readme/[readme_url]/+server.ts

@@ -0,0 +1,39 @@
+import { error } from '@sveltejs/kit'
+
+export const GET = async ({ params }: { params: { readme_url: string } }) => {
+  // prevent abuse of the proxy by ensuring the url contains 'readme.md'
+  if (
+    !(
+      params.readme_url.includes('readme.md') ||
+      params.readme_url.includes('README.md')
+    )
+  )
+    return new Response(null)
+
+  let text: string | undefined
+  try {
+    const res = await fetch(params.readme_url, {
+      signal: AbortSignal.timeout(5000),
+    })
+    if (res.ok) {
+      text = await res.text()
+    } else {
+      return error(res.status, res.statusText)
+    }
+  } catch {
+    return error(408, 'timeout')
+  }
+  return new Response(text || null)
+
+  // `https://raw.githubusercontent.com/${github_details.org}/${github_details.repo_name}/HEAD/README.md`
+  // alternative approaches:
+  //  * add a worker that can use 'git archive' to get specific files
+  //    * unfortunately the two options that can easily embeded within this
+  //      sveltekit backend (wasm-git and isomorphicgit) don't support the
+  //      'archive' command
+  //      https://github.com/petersalomonsen/wasm-git/
+  //      https://github.com/isomorphic-git
+  //    * 'git clone' is too expensive for retrieving single files. even when
+  //      done using treeless or blobless flags. see:
+  //      https://noise.getoto.net/2020/12/21/get-up-to-speed-with-partial-clone-and-shallow-clone/
+}