feat: add 7zz for CHM documentation extraction

ai/hermes-workspace/entrypoint-combined.sh

@@ -1,154 +0,0 @@
-#!/bin/bash
-# ── Hermes Workspace Combined Entrypoint ──
-# Waits for the Hermes gateway container (hermes:8642) to become healthy,
-# then starts the Hermes Workspace web UI in the foreground.
-# Supports graceful shutdown via SIGTERM/SIGINT.
-# ──────────────────────────────────────────
-
-set -euo pipefail
-
-# ── Configuration ──────────────────────────────────────────────
-GATEWAY_HOST="${GATEWAY_HOST:-hermes}"
-GATEWAY_PORT="${GATEWAY_PORT:-8642}"
-GATEWAY_URL="http://${GATEWAY_HOST}:${GATEWAY_PORT}"
-
-HEALTH_ENDPOINT="${HEALTH_ENDPOINT:-/health}"
-MAX_RETRIES="${HEALTH_MAX_RETRIES:-60}"
-RETRY_INTERVAL="${HEALTH_RETRY_INTERVAL:-2}"
-
-WORKSPACE_DIR="${WORKSPACE_DIR:-/workspace}"
-WORKSPACE_ENTRY="${WORKSPACE_ENTRY:-server-entry.js}"
-
-PID_FILE="${PID_FILE:-/tmp/workspace.pid}"
-
-# ── Logging ────────────────────────────────────────────────────
-log_info()  { echo "[$(date '+%Y-%m-%d %H:%M:%S')] [INFO]  $*"; }
-log_warn()  { echo "[$(date '+%Y-%m-%d %H:%M:%S')] [WARN]  $*"; }
-log_error() { echo "[$(date '+%Y-%m-%d %H:%M:%S')] [ERROR] $*"; }
-
-# ── Graceful Shutdown ──────────────────────────────────────────
-_workspace_pid=""
-_shutting_down=false
-
-cleanup() {
-    if [ "$_shutting_down" = true ]; then
-        return
-    fi
-    _shutting_down=true
-
-    log_info "Shutdown signal received, cleaning up..."
-
-    # Stop workspace process if running
-    if [ -n "$_workspace_pid" ] && kill -0 "$_workspace_pid" 2>/dev/null; then
-        log_info "Stopping workspace (PID: $_workspace_pid)..."
-        kill -TERM "$_workspace_pid" 2>/dev/null || true
-
-        # Give it time to shut down gracefully
-        local wait_sec=10
-        while kill -0 "$_workspace_pid" 2>/dev/null && [ "$wait_sec" -gt 0 ]; do
-            sleep 1
-            wait_sec=$((wait_sec - 1))
-        done
-
-        # Force kill if still running
-        if kill -0 "$_workspace_pid" 2>/dev/null; then
-            log_warn "Workspace did not shut down gracefully, force killing..."
-            kill -KILL "$_workspace_pid" 2>/dev/null || true
-        fi
-    fi
-
-    # Clean up PID file
-    [ -f "$PID_FILE" ] && rm -f "$PID_FILE"
-
-    log_info "Shutdown complete."
-    exit 0
-}
-
-# Trap termination signals for graceful shutdown
-trap cleanup SIGTERM SIGINT
-
-# ── Gateway Health Check ───────────────────────────────────────
-wait_for_gateway() {
-    local url="${GATEWAY_URL}${HEALTH_ENDPOINT}"
-    local retries="$MAX_RETRIES"
-    local interval="$RETRY_INTERVAL"
-    local attempt=0
-
-    log_info "Waiting for Hermes gateway at ${GATEWAY_URL}..."
-    log_info "Max retries: ${retries}, interval: ${interval}s"
-
-    while [ "$attempt" -lt "$retries" ]; do
-        attempt=$((attempt + 1))
-
-        if curl -fsS "${url}" >/dev/null 2>&1; then
-            log_info "Gateway is healthy after ${attempt} attempt(s) (${GATEWAY_URL})"
-            return 0
-        fi
-
-        if [ "$attempt" -lt "$retries" ]; then
-            log_info "Gateway not ready yet (attempt ${attempt}/${retries}), retrying in ${interval}s..."
-            sleep "$interval"
-        fi
-    done
-
-    log_error "Gateway did not become healthy after ${retries} attempts ($((retries * interval))s)"
-    return 1
-}
-
-# ── Workspace Startup ──────────────────────────────────────────
-start_workspace() {
-    local entry="${WORKSPACE_DIR}/${WORKSPACE_ENTRY}"
-
-    if [ ! -d "$WORKSPACE_DIR" ]; then
-        log_error "Workspace directory not found: ${WORKSPACE_DIR}"
-        return 1
-    fi
-
-    if [ ! -f "$entry" ]; then
-        log_error "Workspace entry point not found: ${entry}"
-        return 1
-    fi
-
-    log_info "Starting Hermes Workspace web UI..."
-    log_info "  Directory: ${WORKSPACE_DIR}"
-    log_info "  Entry:     ${entry}"
-
-    cd "$WORKSPACE_DIR"
-
-    # Start workspace in background so we can trap signals
-    node --max-old-space-size=2048 "${entry}" &
-    _workspace_pid=$!
-    echo "$_workspace_pid" > "$PID_FILE"
-
-    log_info "Workspace started (PID: ${_workspace_pid})"
-
-    # Wait for workspace process
-    wait "$_workspace_pid"
-    local exit_code=$?
-
-    log_info "Workspace exited with code ${exit_code}"
-    return "$exit_code"
-}
-
-# ── Main ───────────────────────────────────────────────────────
-main() {
-    log_info "=== Hermes Workspace Combined Entrypoint ==="
-    log_info "Gateway:   ${GATEWAY_URL}"
-    log_info "Workspace: ${WORKSPACE_DIR}/${WORKSPACE_ENTRY}"
-    log_info "PID file:  ${PID_FILE}"
-
-    # Wait for gateway to be healthy
-    if ! wait_for_gateway; then
-        log_warn "Proceeding without confirmed gateway health..."
-    fi
-
-    # Start the workspace
-    start_workspace
-    local exit_code=$?
-
-    log_info "Entrypoint exiting with code ${exit_code}"
-    return "$exit_code"
-}
-
-# Run main; exit with its return code
-main "$@"

ai/hermes/Dockerfile

@@ -79,6 +79,39 @@ PYEOF
 COPY --chmod=0755 himalaya-ro.sh /usr/local/bin/himalaya-ro
 
 
+# ---------- Install 7-Zip (7zz) for CHM extraction ----------
+RUN /opt/hermes/.venv/bin/python3 /dev/stdin << 'PYEOF'
+import urllib.request, tarfile, os, shutil, re, subprocess
+
+# Scrape 7-zip.org for latest Linux x64 binary link
+url = 'https://7-zip.org/download.html'
+req = urllib.request.Request(url, headers={'User-Agent': 'Mozilla/5.0'})
+r = urllib.request.urlopen(req, timeout=15)
+html = r.read().decode()
+match = re.search(r'href="(a/7z[\d]+-linux-x64\.tar\.xz)"', html)
+if not match:
+    raise RuntimeError('Could not find 7z Linux x64 download link on 7-zip.org')
+dl_url = f'https://7-zip.org/{match.group(1)}'
+
+# Follow Himalaya pattern: download, extract, install, verify
+xz = '/tmp/7z.tar.xz'
+urllib.request.urlretrieve(dl_url, xz)
+os.makedirs('/tmp/7z', exist_ok=True)
+with tarfile.open(xz, 'r:xz') as t:
+    t.extractall('/tmp/7z')
+shutil.move('/tmp/7z/7zz', '/usr/local/bin/7zz')
+os.chmod('/usr/local/bin/7zz', 0o755)
+shutil.rmtree('/tmp/7z', ignore_errors=True)
+os.remove(xz)
+
+# Verify
+result = subprocess.run(['/usr/local/bin/7zz'], capture_output=True, text=True)
+assert result.returncode == 0, f'7zz verify failed: {result.stderr}'
+version = result.stdout.split()[2] if result.stdout else 'unknown'
+print(f'7-Zip {version} installed successfully')
+PYEOF
+
+
 # ---------- Runtime ----------
 USER hermes
 ENV HERMES_HOME=/opt/data