autonomous-coding-toolkit 1.0.0

package/scripts/tests/test-run-plan-headless.sh

@@ -0,0 +1,424 @@
+#!/usr/bin/env bash
+# Test run-plan-headless.sh extraction
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+RP="$SCRIPT_DIR/../run-plan.sh"
+RPH="$SCRIPT_DIR/../lib/run-plan-headless.sh"
+RPEB="$SCRIPT_DIR/../lib/run-plan-echo-back.sh"
+RPS="$SCRIPT_DIR/../lib/run-plan-sampling.sh"
+
+FAILURES=0
+TESTS=0
+
+assert_eq() {
+    local desc="$1" expected="$2" actual="$3"
+    TESTS=$((TESTS + 1))
+    if [[ "$expected" != "$actual" ]]; then
+        echo "FAIL: $desc"
+        echo "  expected: $expected"
+        echo "  actual:   $actual"
+        FAILURES=$((FAILURES + 1))
+    else
+        echo "PASS: $desc"
+    fi
+}
+
+# === Extracted file exists ===
+
+TESTS=$((TESTS + 1))
+if [[ -f "$RPH" ]]; then
+    echo "PASS: run-plan-headless.sh exists"
+else
+    echo "FAIL: run-plan-headless.sh should exist at scripts/lib/"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Extracted echo-back file exists ===
+TESTS=$((TESTS + 1))
+if [[ -f "$RPEB" ]]; then
+    echo "PASS: run-plan-echo-back.sh exists"
+else
+    echo "FAIL: run-plan-echo-back.sh should exist at scripts/lib/"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Extracted sampling file exists ===
+TESTS=$((TESTS + 1))
+if [[ -f "$RPS" ]]; then
+    echo "PASS: run-plan-sampling.sh exists"
+else
+    echo "FAIL: run-plan-sampling.sh should exist at scripts/lib/"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === run-plan.sh sources it ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'source.*lib/run-plan-headless.sh' "$RP"; then
+    echo "PASS: run-plan.sh sources lib/run-plan-headless.sh"
+else
+    echo "FAIL: run-plan.sh should source lib/run-plan-headless.sh"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === run-plan.sh sources new modules ===
+TESTS=$((TESTS + 1))
+if grep -q 'source.*lib/run-plan-echo-back.sh' "$RP"; then
+    echo "PASS: run-plan.sh sources lib/run-plan-echo-back.sh"
+else
+    echo "FAIL: run-plan.sh should source lib/run-plan-echo-back.sh"
+    FAILURES=$((FAILURES + 1))
+fi
+
+TESTS=$((TESTS + 1))
+if grep -q 'source.*lib/run-plan-sampling.sh' "$RP"; then
+    echo "PASS: run-plan.sh sources lib/run-plan-sampling.sh"
+else
+    echo "FAIL: run-plan.sh should source lib/run-plan-sampling.sh"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === run-plan.sh no longer has inline run_mode_headless body ===
+# The function definition should be in the extracted file, not in run-plan.sh
+
+TESTS=$((TESTS + 1))
+# Count lines of run_mode_headless in run-plan.sh — should be 0 (no function body)
+if grep -q 'run_mode_headless()' "$RP"; then
+    echo "FAIL: run-plan.sh should not define run_mode_headless()"
+    FAILURES=$((FAILURES + 1))
+else
+    echo "PASS: run_mode_headless() not defined in run-plan.sh"
+fi
+
+# === Extracted file defines the function ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'run_mode_headless()' "$RPH"; then
+    echo "PASS: run-plan-headless.sh defines run_mode_headless()"
+else
+    echo "FAIL: run-plan-headless.sh should define run_mode_headless()"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === run-plan.sh is under 330 lines ===
+
+line_count=$(wc -l < "$RP")
+TESTS=$((TESTS + 1))
+if [[ $line_count -le 330 ]]; then
+    echo "PASS: run-plan.sh is $line_count lines (<=330)"
+else
+    echo "FAIL: run-plan.sh is $line_count lines (should be <=330)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Extracted file has the key logic markers ===
+
+TESTS=$((TESTS + 1))
+if grep -q 'mkdir -p.*logs' "$RPH"; then
+    echo "PASS: headless file creates logs directory"
+else
+    echo "FAIL: headless file should create logs directory"
+    FAILURES=$((FAILURES + 1))
+fi
+
+TESTS=$((TESTS + 1))
+if grep -q 'claude -p' "$RPH"; then
+    echo "PASS: headless file calls claude -p"
+else
+    echo "FAIL: headless file should call claude -p"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Empty batch detection ===
+# The parser should return empty text for empty batches, and headless mode should skip them.
+# This test verifies the parser side (headless mode integration is tested separately).
+
+WORK=$(mktemp -d)
+trap 'rm -rf "$WORK"' EXIT
+
+source "$SCRIPT_DIR/../lib/run-plan-parser.sh"
+
+# Create a plan with 2 real batches and 1 empty trailing match
+cat > "$WORK/plan-empty.md" << 'PLAN'
+## Batch 1: Real Batch
+### Task 1: Do something
+Write some code.
+
+## Batch 2: Also Real
+### Task 2: Do more
+Write more code.
+
+## Batch 3:
+PLAN
+
+# get_batch_text should return empty for batch 3
+val=$(get_batch_text "$WORK/plan-empty.md" 3)
+assert_eq "get_batch_text: empty batch returns empty" "" "$val"
+
+# count_batches should count all 3 (parser counts headers)
+val=$(count_batches "$WORK/plan-empty.md")
+assert_eq "count_batches: counts all headers including empty" "3" "$val"
+
+# === Bug #4: CLAUDE.md Run-Plan section removal uses awk not sed ===
+# The sed range deletion pattern '/^## Run-Plan:/,/^## [^R]/' has no terminating
+# anchor when Run-Plan is the last section, so it deletes from Run-Plan to EOF.
+# The fix replaces it with awk which handles last-section correctly.
+
+TESTS=$((TESTS + 1))
+# awk should be used for section removal; the old sed range pattern should not be present
+if grep -q "awk" "$RPH" && grep -q 'in_section' "$RPH"; then
+    echo "PASS: CLAUDE.md section removal uses awk (last-section safe, bug #4)"
+else
+    echo "FAIL: CLAUDE.md section removal should use awk to handle last section correctly (bug #4)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+TESTS=$((TESTS + 1))
+# The old broken sed range pattern must not be present
+if grep -q "sed '/\^## Run-Plan:/,/\^## \[^R\]/" "$RPH" 2>/dev/null || \
+   grep -q "sed '/\^\#\# Run-Plan:/,/\^\#\# \[^R\]" "$RPH" 2>/dev/null; then
+    echo "FAIL: Old sed range deletion pattern still present (unbounded at last section, bug #4)"
+    FAILURES=$((FAILURES + 1))
+else
+    echo "PASS: Old unbounded sed range deletion pattern removed"
+fi
+
+# === Bug #16/#28: SAMPLE_COUNT resets at top of batch loop using SAMPLE_DEFAULT ===
+
+# User's --sample value must be preserved into SAMPLE_DEFAULT before the loop
+TESTS=$((TESTS + 1))
+if grep -q 'SAMPLE_DEFAULT=.*SAMPLE_COUNT' "$RPH"; then
+    echo "PASS: SAMPLE_DEFAULT saves user's --sample value before batch loop"
+else
+    echo "FAIL: SAMPLE_DEFAULT should save user's --sample value before batch loop (bug #16/#28)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# The reset inside the loop must use SAMPLE_DEFAULT, not hardcoded 0
+TESTS=$((TESTS + 1))
+batch_loop_region=$(sed -n '/for ((batch = START_BATCH/,/SAMPLE_ON_RETRY/p' "$RPH")
+if echo "$batch_loop_region" | grep -q 'SAMPLE_COUNT=\$SAMPLE_DEFAULT'; then
+    echo "PASS: SAMPLE_COUNT resets to SAMPLE_DEFAULT at start of each batch iteration"
+else
+    echo "FAIL: SAMPLE_COUNT should reset to SAMPLE_DEFAULT (not 0) at start of each batch iteration (bug #16/#28)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Bug #2/#27: Sampling block uses patch files instead of stash ===
+# The fix replaced git stash/pop with git diff > patch + git apply to
+# eliminate LIFO ordering issues. These tests verify the new approach.
+
+# Baseline state must be saved as a patch file (not stash)
+TESTS=$((TESTS + 1))
+if grep -q '_baseline_patch' "$RPS"; then
+    echo "PASS: run-plan-sampling.sh saves baseline state as a patch file"
+else
+    echo "FAIL: run-plan-sampling.sh should save baseline state as a patch file (bug #2/#27)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# Winner state must be saved as a patch file (not stash)
+TESTS=$((TESTS + 1))
+if grep -q '_winner_patch\|run-plan-winner' "$RPS"; then
+    echo "PASS: run-plan-sampling.sh saves winner state as a patch file"
+else
+    echo "FAIL: run-plan-sampling.sh should save winner state as a patch file (bug #2/#27)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# No executable git stash usage remaining in sampling module (patch approach replaces it).
+# Filter out comment lines (lines starting with optional whitespace + #).
+TESTS=$((TESTS + 1))
+sampling_block=$(sed -n '/^run_sampling_candidates()/,/^}/p' "$RPS")
+# Strip comment-only lines before counting stash calls
+stash_uses=$(echo "$sampling_block" | grep -v '^\s*#' | grep -c 'git stash' || true)
+if [[ "$stash_uses" -eq 0 ]]; then
+    echo "PASS: No git stash calls in run_sampling_candidates (replaced by patch approach)"
+else
+    echo "FAIL: Found $stash_uses git stash call(s) in run_sampling_candidates — should use patch files (bug #2/#27)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# Restore of winner uses git apply (patch approach)
+TESTS=$((TESTS + 1))
+if echo "$sampling_block" | grep -q 'git apply'; then
+    echo "PASS: run_sampling_candidates uses git apply to restore winner state"
+else
+    echo "FAIL: run_sampling_candidates should use git apply to restore winner state (bug #2/#27)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Bug #30: Echo-back gate behavior ===
+
+# run-plan.sh must accept --skip-echo-back without error
+TESTS=$((TESTS + 1))
+if grep -q '\-\-skip-echo-back' "$RP"; then
+    echo "PASS: run-plan.sh accepts --skip-echo-back flag"
+else
+    echo "FAIL: run-plan.sh should define --skip-echo-back flag (bug #30)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# run-plan.sh must accept --strict-echo-back without error
+TESTS=$((TESTS + 1))
+if grep -q '\-\-strict-echo-back' "$RP"; then
+    echo "PASS: run-plan.sh accepts --strict-echo-back flag"
+else
+    echo "FAIL: run-plan.sh should define --strict-echo-back flag (bug #30)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# _echo_back_check function must exist in headless file
+TESTS=$((TESTS + 1))
+if grep -q '_echo_back_check()' "$RPEB"; then
+    echo "PASS: _echo_back_check() is defined in run-plan-echo-back.sh"
+else
+    echo "FAIL: _echo_back_check() should be defined in run-plan-echo-back.sh (bug #30)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# Echo-back gate must be non-blocking by default (no early return when STRICT_ECHO_BACK not set)
+TESTS=$((TESTS + 1))
+if grep -q 'STRICT_ECHO_BACK' "$RPEB"; then
+    echo "PASS: STRICT_ECHO_BACK controls blocking behavior in echo-back gate"
+else
+    echo "FAIL: echo-back gate should check STRICT_ECHO_BACK for blocking mode (bug #30)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# echo-back gate is documented as non-blocking by default
+TESTS=$((TESTS + 1))
+if grep -q 'NON-BLOCKING' "$RPEB"; then
+    echo "PASS: run-plan-echo-back.sh documents NON-BLOCKING default behavior"
+else
+    echo "FAIL: run-plan-echo-back.sh should document NON-BLOCKING default (bug #30)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# _echo_back_check: SKIP_ECHO_BACK=true must cause early return without error
+TESTS=$((TESTS + 1))
+(
+    source "$RPEB" 2>/dev/null || true
+    SKIP_ECHO_BACK=true
+    STRICT_ECHO_BACK=false
+    _echo_back_check "some batch text here" "/nonexistent/log" 2>/dev/null
+) && echo "PASS: _echo_back_check returns 0 when SKIP_ECHO_BACK=true" \
+  || {
+    echo "FAIL: _echo_back_check should return 0 when SKIP_ECHO_BACK=true (bug #30)"
+    FAILURES=$((FAILURES + 1))
+}
+
+# _echo_back_check: missing log file does not crash
+TESTS=$((TESTS + 1))
+(
+    source "$RPEB" 2>/dev/null || true
+    SKIP_ECHO_BACK=false
+    STRICT_ECHO_BACK=false
+    _echo_back_check "some batch text here" "/nonexistent/log" 2>/dev/null
+) && echo "PASS: _echo_back_check handles missing log file gracefully" \
+  || {
+    echo "FAIL: _echo_back_check should handle missing log file gracefully (bug #30)"
+    FAILURES=$((FAILURES + 1))
+}
+
+# _echo_back_check: empty batch text does not crash
+TESTS=$((TESTS + 1))
+tmplog=$(mktemp)
+echo "some agent output here" > "$tmplog"
+(
+    source "$RPEB" 2>/dev/null || true
+    SKIP_ECHO_BACK=false
+    STRICT_ECHO_BACK=false
+    _echo_back_check "" "$tmplog" 2>/dev/null
+)
+ec=$?
+rm -f "$tmplog"
+if [[ $ec -eq 0 ]]; then
+    echo "PASS: _echo_back_check handles empty batch text gracefully"
+else
+    echo "FAIL: _echo_back_check should handle empty batch text without error (bug #30)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Bug #4 BEHAVIORAL: awk removes Run-Plan even when it's the last section ===
+# The old sed range pattern '/^## Run-Plan:/,/^## [^R]/' had no closing anchor
+# when Run-Plan was the last section, eating the file from Run-Plan to EOF.
+# This behavioral test exercises the actual awk code path with a CLAUDE.md
+# where "## Run-Plan:" is the last section and verifies other sections survive.
+
+WORK_AWK=$(mktemp -d)
+# Create a CLAUDE.md where Run-Plan is the LAST section
+cat > "$WORK_AWK/CLAUDE.md" << 'CLAUDE_EOF'
+# Project Config
+
+## Conventions
+
+- Use pytest
+- Stage specific files
+
+## Run-Plan: Batch 3
+
+### Recent Commits
+abc1234 fix: something
+
+### Progress Notes
+Batch 2 done.
+CLAUDE_EOF
+
+# Run the same awk logic that run-plan-headless.sh uses inline
+awk '
+    /^## Run-Plan:/ { in_section=1; next }
+    in_section && /^## / { in_section=0 }
+    !in_section { print }
+' "$WORK_AWK/CLAUDE.md" > "$WORK_AWK/CLAUDE.md.tmp"
+mv "$WORK_AWK/CLAUDE.md.tmp" "$WORK_AWK/CLAUDE.md"
+
+# Verify: Run-Plan section is gone
+TESTS=$((TESTS + 1))
+if grep -q "## Run-Plan:" "$WORK_AWK/CLAUDE.md"; then
+    echo "FAIL: awk last-section: Run-Plan section should be removed"
+    FAILURES=$((FAILURES + 1))
+else
+    echo "PASS: awk last-section: Run-Plan section removed"
+fi
+
+# Verify: Conventions section still exists (not eaten by unbounded deletion)
+TESTS=$((TESTS + 1))
+if grep -q "## Conventions" "$WORK_AWK/CLAUDE.md"; then
+    echo "PASS: awk last-section: Conventions section preserved"
+else
+    echo "FAIL: awk last-section: Conventions section should survive Run-Plan removal"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# Verify: content before Run-Plan is preserved
+TESTS=$((TESTS + 1))
+if grep -q "Use pytest" "$WORK_AWK/CLAUDE.md"; then
+    echo "PASS: awk last-section: content before Run-Plan preserved"
+else
+    echo "FAIL: awk last-section: content before Run-Plan should be preserved"
+    FAILURES=$((FAILURES + 1))
+fi
+
+rm -rf "$WORK_AWK"
+
+# === Bug #38: Empty claude output diagnostic ===
+
+# Must check for empty log file after claude invocation
+TESTS=$((TESTS + 1))
+if grep -q 'claude produced no output' "$RPH"; then
+    echo "PASS: Empty claude output is diagnosed with a warning message (#38)"
+else
+    echo "FAIL: Should diagnose empty claude output (crash/no output case) (bug #38)"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# === Summary ===
+echo ""
+echo "Results: $((TESTS - FAILURES))/$TESTS passed"
+if [[ $FAILURES -gt 0 ]]; then
+    echo "FAILURES: $FAILURES"
+    exit 1
+fi
+echo "ALL PASSED"

package/scripts/tests/test-run-plan-notify.sh

@@ -0,0 +1,124 @@
+#!/usr/bin/env bash
+# Test notification format functions (no actual Telegram sending)
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+source "$SCRIPT_DIR/../lib/run-plan-notify.sh"
+
+FAILURES=0
+TESTS=0
+
+assert_contains() {
+    local desc="$1" needle="$2" haystack="$3"
+    TESTS=$((TESTS + 1))
+    if [[ "$haystack" == *"$needle"* ]]; then
+        echo "PASS: $desc"
+    else
+        echo "FAIL: $desc"
+        echo "  expected to contain: $needle"
+        echo "  actual:              $haystack"
+        FAILURES=$((FAILURES + 1))
+    fi
+}
+
+assert_eq() {
+    local desc="$1" expected="$2" actual="$3"
+    TESTS=$((TESTS + 1))
+    if [[ "$expected" == "$actual" ]]; then
+        echo "PASS: $desc"
+    else
+        echo "FAIL: $desc"
+        echo "  expected: $expected"
+        echo "  actual:   $actual"
+        FAILURES=$((FAILURES + 1))
+    fi
+}
+
+# --- Test: format_success_message includes plan name ---
+msg=$(format_success_message "my-feature" 3 10 "Context Assembler" 2003 1953 "4m12s" "headless" "")
+assert_contains "success includes plan name" "my-feature" "$msg"
+
+# --- Test: format_success_message includes batch X/Y ---
+assert_contains "success includes batch X/Y" "Batch 3/10" "$msg"
+
+# --- Test: format_success_message includes batch title ---
+assert_contains "success includes batch title" "Context Assembler" "$msg"
+
+# --- Test: format_success_message includes check mark ---
+assert_contains "success includes check mark" "✓" "$msg"
+
+# --- Test: format_success_message includes test count ---
+assert_contains "success includes test count" "2003" "$msg"
+
+# --- Test: format_success_message includes delta with up arrow ---
+assert_contains "success includes delta" "↑50" "$msg"
+
+# --- Test: format_success_message delta calculation: 2003 - 1953 = 50 ---
+TESTS=$((TESTS + 1))
+if [[ "$msg" == *"↑50"* ]] && [[ "$msg" != *"↑500"* ]]; then
+    echo "PASS: delta is exactly 50"
+else
+    echo "FAIL: delta should be exactly 50 (2003 - 1953)"
+    echo "  message: $msg"
+    FAILURES=$((FAILURES + 1))
+fi
+
+# --- Test: format_success_message includes duration ---
+assert_contains "success includes duration" "4m12s" "$msg"
+
+# --- Test: format_success_message includes mode ---
+assert_contains "success includes mode" "headless" "$msg"
+
+# --- Test: format_success_message with summary ---
+msg=$(format_success_message "my-feature" 1 5 "Quick Fixes" 100 90 "2m30s" "headless" "Added 3 tests, fixed parser")
+assert_contains "success includes summary" "Added 3 tests, fixed parser" "$msg"
+
+# --- Test: format_failure_message includes plan name ---
+msg=$(format_failure_message "my-feature" 2 8 "ast-grep Rules" 45 3 "pytest failed" "Fix test_auth.py")
+assert_contains "failure includes plan name" "my-feature" "$msg"
+
+# --- Test: format_failure_message includes batch X/Y ---
+assert_contains "failure includes batch X/Y" "Batch 2/8" "$msg"
+
+# --- Test: format_failure_message includes batch title ---
+assert_contains "failure includes batch title" "ast-grep Rules" "$msg"
+
+# --- Test: format_failure_message includes cross mark ---
+assert_contains "failure includes cross mark" "✗" "$msg"
+
+# --- Test: format_failure_message includes test count ---
+assert_contains "failure includes test count" "45" "$msg"
+
+# --- Test: format_failure_message includes failing count ---
+assert_contains "failure includes failing count" "3 failing" "$msg"
+
+# --- Test: format_failure_message includes error as Issue ---
+assert_contains "failure includes issue text" "pytest failed" "$msg"
+
+# --- Test: format_failure_message includes action ---
+assert_contains "failure includes action" "Fix test_auth.py" "$msg"
+
+# --- Test: _load_telegram_env warns on missing file ---
+WORK=$(mktemp -d)
+trap 'rm -rf "$WORK"' EXIT
+
+msg=$(_load_telegram_env "$WORK/.env-nonexistent" 2>&1 || true)
+assert_contains "warns on missing env file" "warn" "$(echo "$msg" | tr '[:upper:]' '[:lower:]')"
+
+# --- Test: _send_telegram warns on missing token ---
+unset TELEGRAM_BOT_TOKEN 2>/dev/null || true
+unset TELEGRAM_CHAT_ID 2>/dev/null || true
+msg=$(_send_telegram "test message" 2>&1 || true)
+assert_contains "send warns on missing credentials" "warn" "$(echo "$msg" | tr '[:upper:]' '[:lower:]')"
+
+# --- Test: format_success_message with zero delta ---
+msg=$(format_success_message "zero-delta" 1 1 "Single Batch" 100 100 "1m00s" "team" "")
+assert_contains "zero delta shows ↑0" "↑0" "$msg"
+
+echo ""
+echo "Results: $((TESTS - FAILURES))/$TESTS passed"
+if [[ $FAILURES -gt 0 ]]; then
+    echo "FAILURES: $FAILURES"
+    exit 1
+fi
+echo "ALL PASSED"