ci: sync gitleaks allowlist

.gitea/workflows/ci.yml

@@ -0,0 +1,68 @@
+---
+# Homelab CI — Docker/heavy lane (git-ci-02)
+name: CI
+
+on:
+  push:
+    branches: [master, main]
+  pull_request:
+    types: [opened, synchronize, reopened]
+
+jobs:
+  skip-ci-check:
+    runs-on: [homelab, self-hosted, linux]
+    container:
+      image: node:20-bookworm
+    outputs:
+      should-skip: ${{ steps.check.outputs.skip }}
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 1
+      - id: check
+        run: |
+          SKIP=0
+          BRANCH="${GITHUB_HEAD_REF:-${GITHUB_REF#refs/heads/}}"
+          MSG="${GITHUB_EVENT_HEAD_COMMIT_MESSAGE:-$(git log -1 --pretty=%B 2>/dev/null || true)}"
+          echo "$BRANCH" "$MSG" | grep -qi '@skipci' && SKIP=1
+          echo "skip=$SKIP" >> $GITHUB_OUTPUT
+
+  docker-ci:
+    needs: skip-ci-check
+    if: needs.skip-ci-check.outputs.should-skip != '1'
+    runs-on: [homelab, self-hosted, linux, heavy, docker]
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Hadolint
+        run: |
+          shopt -s globstar nullglob
+          found=0
+          for f in Dockerfile docker/**/Dockerfile */Dockerfile; do
+            [ -f "$f" ] || continue
+            found=1
+            # Warnings (unpinned apt/pip) are advisory; only errors fail the job
+            docker run --rm -i hadolint/hadolint hadolint --failure-threshold error - < "$f"
+          done
+          [ "$found" -eq 1 ] || echo "No Dockerfile — skip hadolint"
+
+      - name: Trivy config scan (advisory)
+        run: |
+          docker run --rm -v "$PWD:/repo" aquasec/trivy:latest config /repo || true
+
+  secret-scan:
+    needs: skip-ci-check
+    if: needs.skip-ci-check.outputs.should-skip != '1'
+    runs-on: [homelab, self-hosted, linux, heavy]
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+      - name: Gitleaks
+        run: |
+          extra=""
+          if [ -f .gitleaks.toml ]; then
+            extra="--config /repo/.gitleaks.toml"
+          fi
+          docker run --rm -v "$PWD:/repo" ghcr.io/gitleaks/gitleaks:latest \
+            detect --source /repo --no-banner --redact ${extra}

.gitleaks.toml

@@ -0,0 +1,28 @@
+# Homelab bootstrap — gitleaks allowlist (tests, examples, placeholders)
+#
+# IMPORTANT: `useDefault = true` is required — without it gitleaks loads ONLY
+# this file (title + allowlist) with ZERO detection rules, so it would never
+# flag a real secret. Fixed 2026-07 (security-hardening track); if you're
+# re-pushing this template to a repo that already had the old version, that
+# repo's secret scanning was a no-op until this lands.
+title = "homelab gitea bootstrap"
+
+[extend]
+useDefault = true
+
+[allowlist]
+description = "Test fixtures and example configs are not production secrets"
+paths = [
+  '''(?i).*\.test\.(ts|tsx|js|jsx|py)$''',
+  '''(?i).*\.spec\.(ts|tsx|js|jsx)$''',
+  '''(?i).*/tests/.*''',
+  '''(?i).*/__tests__/.*''',
+  '''(?i).*\.example\.(yml|yaml|env|json|toml)$''',
+  '''(?i).*vault\.example\.(yml|yaml)$''',
+  '''(?i).*\.env\.example$''',
+]
+regexes = [
+  '''(?i)(invalid|fake|dummy|placeholder|example|changeme|change_me|not-a-real)''',
+  '''(?i)sk-or-invalid''',
+  '''(?i)msk-or-invalid''',
+]

extractors/hiringcafe/src/main.ts

@@ -111,6 +111,12 @@ function parseWorkplaceTypes(
   }
 }
 
+function encodeSearchState(searchState: unknown): string {
+  const json = JSON.stringify(searchState);
+  const urlEncodedJson = encodeURIComponent(json);
+  return Buffer.from(urlEncodedJson, "utf-8").toString("base64");
+}
+
 function asRecord(value: unknown): Record<string, unknown> | null {
   if (!value || typeof value !== "object" || Array.isArray(value)) return null;
   return value as Record<string, unknown>;
@@ -224,18 +230,7 @@ function extractResultsBatch(payload: unknown): RawHiringCafeJob[] {
   }
 
   const payloadRecord = asRecord(payload);
-  if (!payloadRecord) return [];
-
-  const pageProps = asRecord(payloadRecord.pageProps);
-  const ssrHits = pageProps?.ssrHits ?? payloadRecord.ssrHits;
-  if (Array.isArray(ssrHits)) {
-    return ssrHits.filter(
-      (item): item is RawHiringCafeJob =>
-        Boolean(item) && typeof item === "object" && !Array.isArray(item),
-    );
-  }
-
-  const results = payloadRecord.results;
+  const results = payloadRecord?.results;
   if (!Array.isArray(results)) return [];
 
   return results.filter(
@@ -247,22 +242,7 @@ function extractResultsBatch(payload: unknown): RawHiringCafeJob[] {
 function parseTotalCount(payload: unknown): number | null {
   const payloadRecord = asRecord(payload);
   if (!payloadRecord) return null;
-
-  const pageProps = asRecord(payloadRecord.pageProps);
-  return (
-    toNumberOrNull(pageProps?.ssrTotalCount) ??
-    toNumberOrNull(payloadRecord.ssrTotalCount) ??
-    toNumberOrNull(payloadRecord.total)
-  );
-}
-
-function isLastSsrPage(payload: unknown): boolean {
-  const payloadRecord = asRecord(payload);
-  if (!payloadRecord) return true;
-
-  const pageProps = asRecord(payloadRecord.pageProps);
-  const isLastPage = pageProps?.ssrIsLastPage ?? payloadRecord.ssrIsLastPage;
-  return isLastPage === true;
+  return toNumberOrNull(payloadRecord.total);
 }
 
 function buildCityLocationId(input: string): string {
@@ -521,37 +501,23 @@ function createCitySearchState(args: {
   };
 }
 
-async function readBuildIdFromPage(page: Page): Promise<string> {
-  const buildId = await page.evaluate(() => {
-    const element = document.querySelector("#__NEXT_DATA__");
-    if (!element?.textContent) return null;
-    const data = JSON.parse(element.textContent) as { buildId?: string };
-    return data.buildId ?? null;
-  });
-  if (!buildId) {
-    throw new Error("Hiring Cafe page did not expose Next.js buildId");
-  }
-  return buildId;
-}
-
-async function fetchSsrSearchPage(
+async function callHiringCafeApi(
   page: Page,
-  buildId: string,
-  searchState: Record<string, unknown>,
-  pageNo: number,
+  endpoint: string,
+  params: Record<string, string>,
 ): Promise<unknown> {
   const response = await page.evaluate(
-    async ({ buildIdArg, searchStateArg, pageNoArg }) => {
-      const params = new URLSearchParams();
-      params.set("searchState", JSON.stringify(searchStateArg));
-      params.set("page", String(pageNoArg));
-      const url = `/_next/data/${buildIdArg}/index.json?${params.toString()}`;
+    async ({ endpointArg, paramsArg }) => {
+      const url = new URL(endpointArg, window.location.origin);
+      for (const [key, value] of Object.entries(paramsArg)) {
+        url.searchParams.set(key, value);
+      }
 
-      const res = await fetch(url, {
+      const res = await fetch(url.toString(), {
+        method: "GET",
         credentials: "include",
         headers: {
-          Accept: "application/json",
-          "x-nextjs-data": "1",
+          Accept: "application/json, text/plain, */*",
         },
       });
 
@@ -573,7 +539,7 @@ async function fetchSsrSearchPage(
 
       return output;
     },
-    { buildIdArg: buildId, searchStateArg: searchState, pageNoArg: pageNo },
+    { endpointArg: endpoint, paramsArg: params },
   );
 
   const result = response as BrowserApiResponse;
@@ -581,14 +547,14 @@ async function fetchSsrSearchPage(
   if (!result.ok) {
     const snippet = result.responseText.slice(0, 250);
     throw new Error(
-      `Hiring Cafe SSR search failed (${result.status} ${result.statusText}): ${snippet}`,
+      `Hiring Cafe API ${endpoint} failed (${result.status} ${result.statusText}): ${snippet}`,
     );
   }
 
   if (result.data === null) {
     const snippet = result.responseText.slice(0, 250);
     throw new Error(
-      `Hiring Cafe SSR search returned non-JSON response: ${snippet}`,
+      `Hiring Cafe API ${endpoint} returned non-JSON response: ${snippet}`,
     );
   }
 
@@ -660,8 +626,6 @@ async function run(): Promise<void> {
       await initializePage();
     }
 
-    const buildId = await readBuildIdFromPage(page);
-
     const countryLocation = resolveHiringCafeCountryLocation(country);
     const countryLong =
       countryLocation?.address_components[0]?.long_name ?? "United Kingdom";
@@ -700,23 +664,40 @@ async function run(): Promise<void> {
             dateFetchedPastNDays,
             workplaceTypes,
           });
-      let termTarget = maxJobsPerTerm;
+      const encodedSearchState = encodeSearchState(searchState);
+
+      let totalAvailable: number | null = null;
+      try {
+        const countPayload = await callHiringCafeApi(
+          page,
+          "/api/search-jobs/get-total-count",
+          {
+            s: encodedSearchState,
+          },
+        );
+        totalAvailable = parseTotalCount(countPayload);
+      } catch (error) {
+        const message = error instanceof Error ? error.message : String(error);
+        console.warn(
+          `Hiring Cafe count request failed for term '${searchTerm}': ${message}`,
+        );
+      }
+
+      const termTarget =
+        totalAvailable !== null
+          ? Math.min(maxJobsPerTerm, totalAvailable)
+          : maxJobsPerTerm;
 
       let pageNo = 0;
       let termCollected = 0;
 
       while (termCollected < termTarget && pageNo < PAGE_LIMIT) {
-        const jobsPayload = await fetchSsrSearchPage(
-          page,
-          buildId,
-          searchState,
-          pageNo,
-        );
-
-        const totalAvailable = parseTotalCount(jobsPayload);
-        if (totalAvailable !== null) {
-          termTarget = Math.min(maxJobsPerTerm, totalAvailable);
-        }
+        const size = Math.min(1000, termTarget - termCollected);
+        const jobsPayload = await callHiringCafeApi(page, "/api/search-jobs", {
+          size: String(size),
+          page: String(pageNo),
+          s: encodedSearchState,
+        });
 
         const batch = extractResultsBatch(jobsPayload);
         if (batch.length === 0) break;
@@ -746,7 +727,7 @@ async function run(): Promise<void> {
           totalCollected: termCollected,
         });
 
-        if (isLastSsrPage(jobsPayload)) break;
+        if (batch.length < size) break;
         pageNo += 1;
       }