passepartout/test/integration-tui.sh

#!/bin/bash
set -euo pipefail

PASS=0
FAIL=0
WARN=0
TUI_LOG="/tmp/passepartout-tui-test.log"
> "$TUI_LOG"

cleanup() {
  tmux kill-session -t tui-test 2>/dev/null || true
}
trap cleanup EXIT

run_test() {
  local name="$1"; shift
  echo -n "  $name ... "
  if "$@" 2>/dev/null; then
    echo "PASS"
    PASS=$((PASS + 1))
  else
    echo "FAIL"
    FAIL=$((FAIL + 1))
  fi
}

# ---- Setup ----
echo "Starting TUI in tmux (daemon must already be running on port 9105)..."
tmux new-session -d -s tui-test "passepartout tui 2>&1 | tee $TUI_LOG"
for i in $(seq 1 20); do
  sleep 3
  if tmux capture-pane -t tui-test -p 2>/dev/null | grep -q 'Connected'; then
    echo "  TUI ready after $((i*3))s"
    break
  fi
  if [ "$i" -eq 20 ]; then
    echo "  WARNING: TUI did not render after 60s"
  fi
done

# ---- Tests ----

test_cascade_parsing() {
  # Via /eval, load the provider cascade from the daemon's data dir
  # and verify clean keyword parsing (no cl-dotenv quote contamination).
  local data_dir="${PASSEPARTOUT_DATA_DIR:-$(dirname $(dirname $0))}"
  tmux send-keys -t tui-test "/eval (load (format nil \"~alisp/system-model-provider.lisp\" \"$data_dir/\"))" Enter
  sleep 3
  tmux send-keys -t tui-test "/eval *provider-cascade*" Enter
  sleep 3
  local pane
  pane=$(tmux capture-pane -t tui-test -p -S -15 2>/dev/null)
  echo "$pane" | grep -q ':DEEPSEEK\|:OPENROUTER\|:OPENAI\|:ANTHROPIC\|:GROQ\|:GEMINI\|:NVIDIA'
}

test_eval_command() {
  tmux send-keys -t tui-test "/eval (+ 1 2)" Enter
  sleep 3
  tmux capture-pane -t tui-test -p -S -10 2>/dev/null | grep -q '=> 3'
}

test_status_bar() {
  tmux capture-pane -t tui-test -p -S -20 2>/dev/null | grep -q 'msgs:'
}

# ---- Diagnostic: rendering pipeline isolation ----

test_add_msg_render() {
  # Stage A: can the TUI render an agent message at all?
  # Inject a message directly via /eval — bypasses daemon entirely.
  tmux send-keys -t tui-test "/eval (passepartout.gateway-tui:add-msg :agent \"RENDER-TEST-OK\")" Enter
  sleep 2
  tmux capture-pane -t tui-test -p -S -10 2>/dev/null | grep -q 'RENDER-TEST-OK'
}

test_daemon_msg_roundtrip() {
  # Stage B: does the daemon's LLM response reach the TUI's message list?
  # Sends a message, waits, then checks via /eval that an :agent message exists.
  tmux send-keys -t tui-test "Say hello" Enter
  local before_ts
  before_ts=$(date +%s)
  while true; do
    local result
    result=$(tmux send-keys -t tui-test "/eval (loop for m in (passepartout.gateway-tui:st :messages) when (eq :agent (getf m :role)) return t)" Enter 2>/dev/null; sleep 3; tmux capture-pane -t tui-test -p -S -15 2>/dev/null | grep -o '=> [^ ]*' | tail -1)
    if echo "$result" | grep -q '=> T'; then
      return 0
    fi
    local now_ts
    now_ts=$(date +%s)
    if (( now_ts - before_ts > 90 )); then
      echo "TIMEOUT: no :agent msg in message list after 90s" >&2
      return 1
    fi
    sleep 3
  done
}

test_agent_response_renders() {
  # Stage C: full end-to-end — LLM response appears on the rendered screen.
  # Must show actual response text, not a cascade failure.
  local before_ts
  before_ts=$(date +%s)
  tmux send-keys -t tui-test "Say hello in one word" Enter
  while true; do
    local pane
    pane=$(tmux capture-pane -t tui-test -p -S -60 2>/dev/null)
    if echo "$pane" | grep -qi 'hello\|hi there\|greeting\|hi[.!?]\|hey[.!?]'; then
      if echo "$pane" | grep -qi 'cascade.*fail\|exhausted\|neural cascade'; then
        echo "FAIL: agent responded with cascade failure, not LLM content" >&2
        return 1
      fi
      return 0
    fi
    local now_ts
    now_ts=$(date +%s)
    if (( now_ts - before_ts > 90 )); then
      echo "TIMEOUT: no agent response on screen after 90s" >&2
      return 1
    fi
    sleep 3
  done
}

test_connection_drop() {
  sleep 1
  tmux capture-pane -t tui-test -p -S -10 2>/dev/null | grep -qi 'connection.*lost\|ERROR.*Connection\|error.*connect' || true
  return 0
}

run_test "cascade-parsing"         test_cascade_parsing
run_test "eval-command"            test_eval_command
run_test "status-bar"              test_status_bar
run_test "add-msg-render"          test_add_msg_render
run_test "daemon-msg-roundtrip"    test_daemon_msg_roundtrip
run_test "agent-response-renders"   test_agent_response_renders
run_test "connection-drop"         test_connection_drop

# ---- Summary ----
echo ""
echo "===== $PASS passed, $FAIL failed, $WARN warnings ====="
exit $(( FAIL > 0 ? 1 : 0 ))