feat: add enricher helper files (types, fetchLightRepo, updateEnrichedRepos)

Signed-off-by: Mouad BANI <mouad-mb@outlook.com>

Author: mbani01

services/apps/packages_worker/src/enricher/fetchLightRepo.ts

@@ -0,0 +1,96 @@
+import { FetchError, LightRepoResult } from './types'
+
+const GRAPHQL_URL = 'https://api.github.com/graphql'
+
+const REPO_QUERY = `
+  query($owner: String!, $name: String!) {
+    repository(owner: $owner, name: $name) {
+      description
+      primaryLanguage { name }
+      repositoryTopics(first: 25) { nodes { topic { name } } }
+      stargazerCount
+      forkCount
+      watchers { totalCount }
+      issues(states: OPEN) { totalCount }
+      pushedAt
+      isArchived
+      isDisabled
+      isFork
+      createdAt
+    }
+  }
+`
+
+export function parseGithubUrl(url: string): { owner: string; name: string } {
+  const match = url.match(/https?:\/\/github\.com\/([^/]+)\/([^/]+?)(?:\.git)?\/?$/)
+  if (!match) throw new FetchError('MALFORMED', `Cannot parse GitHub URL: ${url}`)
+  return { owner: match[1], name: match[2] }
+}
+
+export async function fetchLightRepo(url: string, token: string): Promise<LightRepoResult> {
+  const { owner, name } = parseGithubUrl(url)
+
+  let response: Response
+  try {
+    response = await fetch(GRAPHQL_URL, {
+      method: 'POST',
+      headers: {
+        Authorization: `bearer ${token}`,
+        'Content-Type': 'application/json',
+      },
+      body: JSON.stringify({ query: REPO_QUERY, variables: { owner, name } }),
+    })
+  } catch (err) {
+    throw new FetchError('TRANSIENT', `Network error for ${url}: ${(err as Error).message}`)
+  }
+
+  const resetSec = parseInt(response.headers.get('x-ratelimit-reset') ?? '0', 10)
+  const resetMs = resetSec ? resetSec * 1000 + 5_000 : Date.now() + 65_000
+
+  if (response.status === 401) {
+    throw new FetchError('AUTH', `401 Unauthorized for ${url}`)
+  }
+
+  if (response.status === 403) {
+    const body = await response.text()
+    if (body.toLowerCase().includes('rate limit')) {
+      throw new FetchError('RATE_LIMIT', `Rate limited on ${url}`, resetMs)
+    }
+    throw new FetchError('AUTH', `403 Forbidden for ${url}`)
+  }
+
+  if (response.status === 404) throw new FetchError('NOT_FOUND', `404 for ${url}`)
+  if (response.status >= 500) throw new FetchError('TRANSIENT', `${response.status} for ${url}`)
+
+  // eslint-disable-next-line @typescript-eslint/no-explicit-any
+  const json = (await response.json()) as any
+
+  if (json.errors?.length) {
+    const err = json.errors[0]
+    if (err.type === 'RATE_LIMITED') throw new FetchError('RATE_LIMIT', `RATE_LIMITED for ${url}`, resetMs)
+    if (err.type === 'NOT_FOUND') throw new FetchError('NOT_FOUND', `NOT_FOUND for ${url}`)
+    throw new FetchError('TRANSIENT', `GraphQL error for ${url}: ${err.message ?? err.type}`)
+  }
+
+  const repo = json.data?.repository
+  if (!repo) throw new FetchError('NOT_FOUND', `No repository data for ${url}`)
+
+  return {
+    url,
+    host: 'github',
+    owner,
+    name,
+    description: repo.description ?? null,
+    primaryLanguage: repo.primaryLanguage?.name ?? null,
+    topics: (repo.repositoryTopics?.nodes ?? []).map((n: { topic: { name: string } }) => n.topic.name),
+    stars: repo.stargazerCount ?? 0,
+    forks: repo.forkCount ?? 0,
+    watchers: repo.watchers?.totalCount ?? 0,
+    openIssues: repo.issues?.totalCount ?? 0,
+    lastCommitAt: repo.pushedAt ?? null,
+    archived: repo.isArchived ?? false,
+    disabled: repo.isDisabled ?? false,
+    isFork: repo.isFork ?? false,
+    createdAt: repo.createdAt ?? null,
+  }
+}

services/apps/packages_worker/src/enricher/types.ts

@@ -0,0 +1,31 @@
+export interface LightRepoResult {
+  url: string
+  host: 'github'
+  owner: string
+  name: string
+  description: string | null
+  primaryLanguage: string | null
+  topics: string[]
+  stars: number
+  forks: number
+  watchers: number
+  openIssues: number
+  lastCommitAt: string | null
+  archived: boolean
+  disabled: boolean
+  isFork: boolean
+  createdAt: string | null
+}
+
+export type FetchErrorKind = 'RATE_LIMIT' | 'TRANSIENT' | 'NOT_FOUND' | 'AUTH' | 'MALFORMED'
+
+export class FetchError extends Error {
+  constructor(
+    public readonly kind: FetchErrorKind,
+    message: string,
+    public readonly resetAt?: number,
+  ) {
+    super(message)
+    this.name = 'FetchError'
+  }
+}

services/apps/packages_worker/src/enricher/updateEnrichedRepos.ts

@@ -0,0 +1,36 @@
+import { getServiceChildLogger } from '@crowd/logging'
+import { QueryExecutor } from '@crowd/data-access-layer/src/queryExecutor'
+
+import { LightRepoResult } from './types'
+
+const log = getServiceChildLogger('github-repos-enricher:update')
+
+export async function updateEnrichedRepos(qx: QueryExecutor, rows: LightRepoResult[]): Promise<void> {
+  if (rows.length === 0) return
+
+  for (const r of rows) {
+    await qx.result(
+      `UPDATE repos SET
+        host             = COALESCE(host, $(host)),
+        owner            = COALESCE(owner, $(owner)),
+        name             = COALESCE(name, $(name)),
+        description      = $(description),
+        primary_language = $(primaryLanguage),
+        topics           = $(topics)::text[],
+        stars            = $(stars),
+        forks            = $(forks),
+        watchers         = $(watchers),
+        open_issues      = $(openIssues),
+        last_commit_at   = $(lastCommitAt)::timestamptz,
+        archived         = $(archived),
+        disabled         = $(disabled),
+        is_fork          = $(isFork),
+        created_at       = COALESCE(created_at, $(createdAt)::timestamptz),
+        last_synced_at   = NOW()
+       WHERE url = $(url)`,
+      r,
+    )
+  }
+
+  log.debug({ count: rows.length }, 'Updated enriched repos')
+}