#!/bin/bash
set -u
TMP=/tmp/pdf-seed-fix2
API=http://localhost/pdfs/api/upload
mkdir -p "$TMP"

PDFS=(
"https://archive.org/download/fm21-150-combatives/fm21-150-combatives.pdf|FM 21-150 Combatives (1992)|defense"
"https://archive.org/download/WhereThereIsNoDoctor-English-DavidWerner/WhereThereIsNoDoctor-English-DavidWerner.pdf|Where There Is No Doctor (Hesperian)|medical"
"https://www.globalsecurity.org/military/library/policy/army/fm/3-22-9/fm3-22-9_c1_2011.pdf|FM 3-22.9 Rifle Marksmanship M16/M4 (2011)|defense"
)

ok=0; fail=0; total=${#PDFS[@]}; i=0
for entry in "${PDFS[@]}"; do
  i=$((i+1))
  IFS='|' read -r url title tag <<< "$entry"
  fname=$(echo "$title" | tr -c 'A-Za-z0-9._-' '_' | head -c 100).pdf
  out="$TMP/$fname"
  echo ""
  echo "==> [$i/$total] [$title]"
  if wget -q --show-progress -U "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36" --tries=2 --timeout=120 -O "$out" "$url"; then
    sz_bytes=$(stat -c%s "$out" 2>/dev/null || echo 0)
    if [ "$sz_bytes" -gt 100000 ] && file "$out" | grep -qi "PDF"; then
      sz=$(du -h "$out" | cut -f1)
      curl -s -X POST "$API" -F "file=@$out;type=application/pdf" -F "title=$title" -F "tag=$tag" > /dev/null
      echo "    ✓ uploaded ($sz)"
      ok=$((ok+1))
    else
      echo "    ✗ skipped (not a valid PDF, $sz_bytes bytes)"
      fail=$((fail+1))
    fi
    rm -f "$out"
  else
    echo "    ✗ download failed"
    fail=$((fail+1))
  fi
done

echo ""
echo "================================"
echo "OK: $ok   Failed: $fail   Total: $total"
echo "================================"
rm -rf "$TMP"