Upload install.sh with huggingface_hub

install.sh

@@ -0,0 +1,177 @@
+#!/usr/bin/env bash
+# Install the Lain parser service: one frozen artifact, a private venv and a
+# systemd user unit. Everything lands under --prefix (default
+# ~/.local/share/lain-parser) and ~/.config/systemd/user; no sudo anywhere.
+#
+#   bash install.sh                     # install, enable and start the unit
+#   bash install.sh --no-start          # install files only
+#   bash install.sh --verify            # re-check the installed hashes
+#
+# Source of truth for the artifact and the vendored app: this repository.
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+REPO="$(cd "${SCRIPT_DIR}/../.." && pwd)"
+ARTIFACT_NAME="e5-s2"
+ARTIFACT_SRC="${REPO}/artifacts/e5-s2"
+ARTIFACT_SHA="d928668e5ed3c8ea4ac69e42f8d11c90f82120e3307dffaf3a02eb801c077e54"
+SNAPSHOT_SRC="${ANIFILEBERT_SNAPSHOT:-${REPO}/artifacts/anifilebert-dapt512-long}"
+SNAPSHOT_FILES=(anifilebert config.json model.safetensors tokenizer_config.json \
+                vocab.json vocab.char.json label_schema.json dapt.json)
+PREFIX="${LAIN_PARSER_PREFIX:-${HOME}/.local/share/lain-parser}"
+UNIT_DIR="${HOME}/.config/systemd/user"
+UNIT="lain-parser.service"
+PORT="${LAIN_PARSER_PORT:-9377}"
+HOSTS="${LAIN_PARSER_HOST:-127.0.0.1,172.17.0.1}"
+# The socket lives inside Lain's data dir on purpose: a container that already
+# bind-mounts that directory sees it at /data/parser.sock with no extra mount
+# and no firewall rule (host TCP is firewalled from the bridge on many hosts).
+SOCKET="${LAIN_PARSER_SOCKET:-${HOME}/.local/share/lain/parser.sock}"
+START=1
+VERIFY=0
+FORCE=0
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --prefix) PREFIX="$2"; shift 2 ;;
+    --port) PORT="$2"; shift 2 ;;
+    --host) HOSTS="$2"; shift 2 ;;
+    --socket) SOCKET="$2"; shift 2 ;;
+    --no-start) START=0; shift ;;
+    --force) FORCE=1; shift ;;
+    --verify) VERIFY=1; shift ;;
+    -h|--help) sed -n '2,10p' "$0" | sed 's/^# \{0,1\}//'; exit 0 ;;
+    *) echo "unknown argument: $1" >&2; exit 2 ;;
+  esac
+done
+
+step() { printf '\n== %s\n' "$*"; }
+die() { echo "error: $*" >&2; exit 1; }
+
+sha_of() { sha256sum "$1" | cut -d' ' -f1; }
+
+verify_installed() {
+  step "verifying ${PREFIX}"
+  local venv="${PREFIX}/venv/bin/python"
+  [[ -x "$venv" ]] || die "venv missing under ${PREFIX}"
+  local sum installed
+  sum="$(sha_of "${PREFIX}/artifact/${ARTIFACT_NAME}/model.int8.pt")"
+  [[ "$sum" == "$ARTIFACT_SHA" ]] || die "artifact hash mismatch: ${sum}"
+  echo "artifact ${ARTIFACT_NAME} hash ok"
+  "$venv" "${PREFIX}/app/serve.py" --check "Example.Show.S01E02.1080p-GROUP.mkv" \
+    --artifact "${PREFIX}/artifact/${ARTIFACT_NAME}" --snapshot "${PREFIX}/anifilebert" \
+    >/dev/null
+  echo "model loads and parses"
+}
+
+if [[ "$VERIFY" == 1 ]]; then
+  verify_installed
+  exit 0
+fi
+
+[[ -f "${ARTIFACT_SRC}/model.int8.pt" ]] || die "artifact not found at ${ARTIFACT_SRC}"
+[[ -f "${ARTIFACT_SRC}/config.json" ]] || die "artifact config not found at ${ARTIFACT_SRC}"
+sum="$(sha_of "${ARTIFACT_SRC}/model.int8.pt")"
+[[ "$sum" == "$ARTIFACT_SHA" ]] || die "source artifact hash mismatch: ${sum}"
+[[ -d "$SNAPSHOT_SRC" ]] || die "snapshot not found at ${SNAPSHOT_SRC}"
+
+step "installing into ${PREFIX}"
+mkdir -p "${PREFIX}/app/fp" "${PREFIX}/artifact/${ARTIFACT_NAME}" "${PREFIX}/anifilebert"
+cp -f "${SCRIPT_DIR}/app/serve.py" "${PREFIX}/app/serve.py"
+cp -f "${SCRIPT_DIR}"/app/fp/*.py "${PREFIX}/app/fp/"
+cp -f "${SCRIPT_DIR}/requirements.txt" "${PREFIX}/requirements.txt"
+cp -f "${ARTIFACT_SRC}/model.int8.pt" "${ARTIFACT_SRC}/config.json" \
+      "${PREFIX}/artifact/${ARTIFACT_NAME}/"
+for entry in "${SNAPSHOT_FILES[@]}"; do
+  cp -RL "${SNAPSHOT_SRC}/${entry}" "${PREFIX}/anifilebert/"
+done
+
+step "recording hashes"
+{
+  echo "# installed $(date -u +%Y-%m-%dT%H:%M:%SZ) by $(id -un)@$(hostname)"
+  (cd "$PREFIX" && find app artifact anifilebert requirements.txt -type f \
+     -exec sha256sum {} + | sort -k2)
+} > "${PREFIX}/SHA256SUMS"
+echo "wrote ${PREFIX}/SHA256SUMS"
+
+step "creating the venv (CPU wheels only)"
+if [[ ! -x "${PREFIX}/venv/bin/python" ]]; then
+  if command -v uv >/dev/null 2>&1; then
+    uv venv --python "${LAIN_PARSER_PYTHON:-3.12}" "${PREFIX}/venv"
+    uv pip install --python "${PREFIX}/venv/bin/python" \
+      --index-url https://download.pytorch.org/whl/cpu "torch==2.14.0"
+    uv pip install --python "${PREFIX}/venv/bin/python" -r "${PREFIX}/requirements.txt"
+  else
+    python3 -m venv "${PREFIX}/venv"
+    "${PREFIX}/venv/bin/pip" install --upgrade pip >/dev/null
+    "${PREFIX}/venv/bin/pip" install --index-url https://download.pytorch.org/whl/cpu \
+      "torch==2.14.0"
+    "${PREFIX}/venv/bin/pip" install -r "${PREFIX}/requirements.txt"
+  fi
+else
+  echo "venv already present, keeping it"
+fi
+"${PREFIX}/venv/bin/python" -c "import torch, transformers; print('torch', torch.__version__, '| transformers', transformers.__version__)"
+
+step "systemd user unit"
+mkdir -p "${UNIT_DIR}"
+unit_tmp="$(mktemp)"
+cat > "$unit_tmp" <<UNIT
+# Generated by filename-parser packaging/lain-parser/install.sh -- yours to edit.
+[Unit]
+Description=Lain local filename parser (${ARTIFACT_NAME})
+After=network.target docker.service
+
+[Service]
+WorkingDirectory=${PREFIX}
+ExecStart=${PREFIX}/venv/bin/python ${PREFIX}/app/serve.py --artifact ${PREFIX}/artifact/${ARTIFACT_NAME} --snapshot ${PREFIX}/anifilebert --host ${HOSTS} --port ${PORT} --unix-socket ${SOCKET}
+Environment=PYTHONUNBUFFERED=1
+Restart=on-failure
+RestartSec=2
+
+[Install]
+WantedBy=default.target
+UNIT
+if [[ ! -f "${UNIT_DIR}/${UNIT}" ]]; then
+  install -m 0644 "$unit_tmp" "${UNIT_DIR}/${UNIT}"
+  echo "wrote ${UNIT_DIR}/${UNIT}"
+elif diff -q "$unit_tmp" "${UNIT_DIR}/${UNIT}" >/dev/null; then
+  echo "unit already up to date"
+elif [[ "$FORCE" == 1 ]]; then
+  install -m 0644 "$unit_tmp" "${UNIT_DIR}/${UNIT}"
+  echo "unit replaced (--force)"
+else
+  install -m 0644 "$unit_tmp" "${UNIT_DIR}/${UNIT}.new"
+  echo "unit differs: wrote ${UNIT_DIR}/${UNIT}.new for review (or pass --force)"
+fi
+rm -f "$unit_tmp"
+
+if [[ "$START" == 1 ]] && command -v systemctl >/dev/null 2>&1; then
+  systemctl --user daemon-reload
+  systemctl --user enable --now "$UNIT" || true
+  if command -v curl >/dev/null 2>&1; then
+    ok=""
+    for _ in $(seq 1 20); do
+      sleep 1
+      for addr in ${HOSTS//,/ }; do
+        if curl -fsS --max-time 2 "http://${addr}:${PORT}/health" >/dev/null 2>&1; then ok="$addr"; break 2; fi
+      done
+    done
+    if [[ -n "$ok" ]]; then
+      curl -fsS --max-time 3 "http://${ok}:${PORT}/health"; echo
+    else
+      echo "warning: health check did not answer in 20s (see: journalctl --user -u ${UNIT})"
+    fi
+  fi
+fi
+
+step "done"
+cat <<SUMMARY
+prefix      ${PREFIX}
+artifact    ${ARTIFACT_NAME} (${ARTIFACT_SHA:0:16}…)
+unit        ${UNIT_DIR}/${UNIT}
+native URL  http://127.0.0.1:${PORT}
+docker URL  http://172.17.0.1:${PORT}   (or host.docker.internal with host-gateway)
+docker UDS  unix:///data/parser.sock    (socket at ${SOCKET})
+verify      bash install.sh --verify
+SUMMARY