TafsirQuran/k8s/gharaati-crawl-job.yaml

# One-shot crawl of Tafsir Noor into tafsirquran-audio PVC.
# Scale API to 0 first (RWO PVC), then create ConfigMap + Job:
#   kubectl -n tafsirquran create configmap tafsirquran-gharaati-scripts \
#     --from-file=crawl-gharaati.py=scripts/crawl-gharaati.py --dry-run=client -o yaml | kubectl apply -f -
#   kubectl -n tafsirquran apply -f k8s/gharaati-crawl-job.yaml
#   kubectl -n tafsirquran logs -f job/tafsirquran-gharaati-crawl
apiVersion: batch/v1
kind: Job
metadata:
  name: tafsirquran-gharaati-crawl
  namespace: tafsirquran
spec:
  backoffLimit: 2
  ttlSecondsAfterFinished: 86400
  template:
    spec:
      restartPolicy: Never
      containers:
        - name: crawl
          image: python:3.12-slim
          command: ["/bin/bash", "-lc"]
          args:
            - |
              set -euo pipefail
              apt-get update -qq
              DEBIAN_FRONTEND=noninteractive apt-get install -y -qq ca-certificates >/dev/null
              # Content only — PVC already has ~2997 Gooya MP3s; skip site audio re-download.
              python3 /scripts/crawl-gharaati.py /data/audio --skip-existing-json --delay 0.15
          volumeMounts:
            - name: audio
              mountPath: /data/audio
            - name: scripts
              mountPath: /scripts
          resources:
            requests:
              memory: 256Mi
              cpu: 250m
      volumes:
        - name: audio
          persistentVolumeClaim:
            claimName: tafsirquran-audio
        - name: scripts
          configMap:
            name: tafsirquran-gharaati-scripts
            defaultMode: 0755