feat(ops): automate backups and private CV benchmarks

scripts/run-cv-benchmark.sh

@@ -4,16 +4,32 @@ set -euo pipefail
 ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
 OUTPUT_DIR="${CV_BENCHMARK_OUTPUT_DIR:-$ROOT_DIR/tmp/cv-benchmarks/latest}"
 APPROVED_DIR="${CV_BENCHMARK_APPROVED_DIR:-$ROOT_DIR/tmp/cv-benchmarks/approved-fixtures}"
+CORPUS_DIR="${CV_BENCHMARK_CORPUS_DIR:-/home/pi/cvs}"
+
+if [[ ! -d "$CORPUS_DIR" ]]; then
+  echo "CV corpus directory not found: $CORPUS_DIR" >&2
+  exit 1
+fi
+
+if ! find "$CORPUS_DIR" -maxdepth 1 -type f \( -iname '*.pdf' -o -iname '*.docx' -o -iname '*.txt' -o -iname '*.md' \) -print -quit | grep -q .; then
+  echo "No supported CV files were found in: $CORPUS_DIR" >&2
+  exit 1
+fi
 
 mkdir -p "$OUTPUT_DIR" "$APPROVED_DIR"
 
+echo "CV benchmark corpus: $CORPUS_DIR"
 echo "CV benchmark output: $OUTPUT_DIR"
 echo "Approved fixtures: $APPROVED_DIR"
 
+CV_BENCHMARK_CORPUS_DIR="$CORPUS_DIR" \
 CV_BENCHMARK_OUTPUT_DIR="$OUTPUT_DIR" \
 CV_BENCHMARK_APPROVED_DIR="$APPROVED_DIR" \
   dotnet test "$ROOT_DIR/JobTrackerApi.Tests/JobTrackerApi.Tests.csproj" --filter CvCorpusHarnessTests
 
+test -s "$OUTPUT_DIR/index.json"
+test -s "$OUTPUT_DIR/report.md"
+
 if [[ -f "$OUTPUT_DIR/report.md" ]]; then
   echo
   echo "Benchmark report written to: $OUTPUT_DIR/report.md"