ginskill-init 1.0.0

package/skills/review-code/scripts/dep-check.sh

@@ -0,0 +1,522 @@
+#!/usr/bin/env bash
+# ─────────────────────────────────────────────────────────────
+# Dependency & Import Analyzer
+#
+# Detects circular dependencies, unused exports, orphan files,
+# and import boundary violations. Pure bash — no npm deps.
+#
+# Usage:
+#   ./dep-check.sh                    # scan all projects
+#   ./dep-check.sh backend            # scan backend only
+#   ./dep-check.sh frontend           # scan frontend only
+#   ./dep-check.sh mobile             # scan mobile only
+#   ./dep-check.sh --check circular   # only circular deps
+#   ./dep-check.sh --check unused     # only unused exports
+#   ./dep-check.sh --check orphans    # only orphan files
+#   ./dep-check.sh --check boundaries # only import boundary violations
+#   ./dep-check.sh --summary          # summary counts only
+#
+# Output: JSON to stdout
+# Exit codes: 0 = clean, 1 = issues found, 2 = critical issues
+# ─────────────────────────────────────────────────────────────
+
+set -uo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+REPO_ROOT="$(cd "$SCRIPT_DIR/../../../../.." && pwd)"
+
+# ─── Defaults ────────────────────────────────────────────────
+TARGET="all"
+CHECK="all"
+SUMMARY_ONLY=false
+
+# ─── Auto-detect project directories ────────────────────────
+detect_src() {
+  local kind="$1"; shift
+  for candidate in "$@"; do
+    for dir in $REPO_ROOT/$candidate; do
+      if [ -d "$dir/src" ]; then echo "$dir/src"; return; fi
+      if [ -d "$dir/app" ]; then echo "$dir/app"; return; fi
+    done
+  done
+  echo ""
+}
+
+detect_root() {
+  local kind="$1"; shift
+  for candidate in "$@"; do
+    for dir in $REPO_ROOT/$candidate; do
+      if [ -d "$dir" ] && [ -f "$dir/package.json" ]; then echo "$dir"; return; fi
+    done
+  done
+  echo ""
+}
+
+BE_SRC="$(detect_src backend be-* backend server api)"
+FE_SRC="$(detect_src frontend web-* frontend client)"
+MB_SRC="$(detect_src mobile styai-mobile mobile)"
+AUTH_SRC=""
+[ -d "$REPO_ROOT/auth-package/src" ] && AUTH_SRC="$REPO_ROOT/auth-package/src"
+
+BE_ROOT="$(detect_root backend be-* backend server api)"
+FE_ROOT="$(detect_root frontend web-* frontend client)"
+MB_ROOT="$(detect_root mobile styai-mobile mobile)"
+
+# ─── Parse arguments ────────────────────────────────────────
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    backend)  TARGET="backend"; shift ;;
+    frontend) TARGET="frontend"; shift ;;
+    mobile)   TARGET="mobile"; shift ;;
+    auth)     TARGET="auth"; shift ;;
+    --check)  CHECK="$2"; shift 2 ;;
+    --summary) SUMMARY_ONLY=true; shift ;;
+    *) shift ;;
+  esac
+done
+
+# ─── Build file list ────────────────────────────────────────
+TMPDIR_WORK=$(mktemp -d)
+trap 'rm -rf "$TMPDIR_WORK"' EXIT
+
+FILES_LIST="$TMPDIR_WORK/files.txt"
+FINDINGS_FILE="$TMPDIR_WORK/findings.jsonl"
+touch "$FINDINGS_FILE"
+
+build_file_list() {
+  local dirs=""
+  case "$TARGET" in
+    backend)  [ -n "$BE_SRC" ] && dirs="$BE_SRC" ;;
+    frontend) [ -n "$FE_SRC" ] && dirs="$FE_SRC" ;;
+    mobile)   [ -n "$MB_SRC" ] && dirs="$MB_SRC" ;;
+    auth)     [ -n "$AUTH_SRC" ] && dirs="$AUTH_SRC" ;;
+    all)
+      [ -n "$BE_SRC" ] && dirs="$BE_SRC"
+      [ -n "$FE_SRC" ] && dirs="$dirs $FE_SRC"
+      [ -n "$MB_SRC" ] && dirs="$dirs $MB_SRC"
+      [ -n "$AUTH_SRC" ] && dirs="$dirs $AUTH_SRC"
+      ;;
+  esac
+  dirs=$(echo "$dirs" | xargs)
+  [ -z "$dirs" ] && return
+  for dir in $dirs; do
+    find "$dir" \( -name "*.ts" -o -name "*.tsx" \) \
+      -not -path "*/node_modules/*" \
+      -not -path "*/.next/*" \
+      -not -path "*/dist/*" \
+      -not -path "*.d.ts" \
+      2>/dev/null
+  done
+}
+
+build_file_list | sort > "$FILES_LIST"
+FILE_COUNT=$(wc -l < "$FILES_LIST" | tr -d ' ')
+
+if [ "$FILE_COUNT" -eq 0 ]; then
+  echo '{"scan":"dep-check","findings":[],"summary":{"total":0}}'
+  exit 0
+fi
+
+# ─── Helpers ────────────────────────────────────────────────
+relpath() { echo "${1#$REPO_ROOT/}"; }
+
+json_escape() {
+  printf '%s' "$1" | python3 -c 'import json,sys; print(json.dumps(sys.stdin.read()), end="")' 2>/dev/null || \
+  printf '"%s"' "$(printf '%s' "$1" | sed 's/\\/\\\\/g; s/"/\\"/g; s/\t/\\t/g')"
+}
+
+count_matches() {
+  local result
+  result=$(grep -c "$1" "$2" 2>/dev/null) || true
+  echo "${result:-0}"
+}
+
+add_finding() {
+  local check="$1" severity="$2" rule="$3" file="$4" line="$5" message="$6" suggestion="$7"
+  if [ "$CHECK" != "all" ] && [ "$CHECK" != "$check" ]; then return; fi
+  echo "{\"check\":\"$check\",\"severity\":\"$severity\",\"rule\":\"$rule\",\"file\":\"$(relpath "$file")\",\"line\":$line,\"message\":$(json_escape "$message"),\"suggestion\":$(json_escape "$suggestion")}" >> "$FINDINGS_FILE"
+}
+
+# ═══════════════════════════════════════════════════════════════
+# 1. CIRCULAR DEPENDENCY DETECTION
+# ═══════════════════════════════════════════════════════════════
+check_circular() {
+  local import_map="$TMPDIR_WORK/import_map.txt"
+
+  # Build import map: file -> imported_file
+  while IFS= read -r file; do
+    local dir
+    dir=$(dirname "$file")
+    # Extract import paths (relative imports only)
+    grep -n "^import.*from ['\"]\./" "$file" 2>/dev/null | while IFS=: read -r lineno content; do
+      # Extract the path
+      local import_path
+      import_path=$(echo "$content" | sed "s/.*from ['\"]//; s/['\"].*//" )
+      [ -z "$import_path" ] && continue
+
+      # Resolve to absolute file path
+      local resolved=""
+      for ext in "" ".ts" ".tsx" "/index.ts" "/index.tsx"; do
+        local candidate="$dir/$import_path$ext"
+        candidate=$(cd "$(dirname "$candidate")" 2>/dev/null && echo "$(pwd)/$(basename "$candidate")")
+        if [ -f "$candidate" ]; then
+          resolved="$candidate"
+          break
+        fi
+      done
+
+      [ -n "$resolved" ] && echo "$(relpath "$file")	$(relpath "$resolved")"
+    done
+  done < "$FILES_LIST" > "$import_map"
+
+  # Detect cycles: A->B->A (direct circular)
+  sort "$import_map" | awk -F'\t' '
+    {
+      from = $1; to = $2
+      edges[from] = edges[from] " " to
+      # Check for direct cycle: A imports B AND B imports A
+      if (from == to) {
+        printf "self\t%s\t%s\n", from, from
+      }
+    }
+    END {
+      # Check bidirectional imports
+      for (a in edges) {
+        n = split(edges[a], targets, " ")
+        for (i = 1; i <= n; i++) {
+          b = targets[i]
+          if (b == "" || b == a) continue
+          # Check if b also imports a
+          m = split(edges[b], btargets, " ")
+          for (j = 1; j <= m; j++) {
+            if (btargets[j] == a) {
+              # Only print once (alphabetically first)
+              if (a < b) {
+                printf "mutual\t%s\t%s\n", a, b
+              }
+            }
+          }
+        }
+      }
+    }
+  ' | sort -u | while IFS=$'\t' read -r kind file_a file_b; do
+    if [ "$kind" = "self" ]; then
+      add_finding "circular" "critical" "self_import" "$REPO_ROOT/$file_a" 1 \
+        "File imports itself" \
+        "Remove the self-referencing import"
+    else
+      add_finding "circular" "warning" "mutual_import" "$REPO_ROOT/$file_a" 1 \
+        "Circular dependency: $file_a <-> $file_b" \
+        "Break the cycle by extracting shared logic into a third module, or use dependency inversion"
+    fi
+  done
+
+  # Detect feature-to-feature cross-imports (NestJS modules importing each other)
+  if [ -n "$BE_SRC" ] && [ -d "$BE_SRC/features" ]; then
+    grep "features/" "$import_map" 2>/dev/null | awk -F'\t' '
+      {
+        from = $1; to = $2
+        # Extract feature names
+        split(from, from_parts, "/")
+        split(to, to_parts, "/")
+        from_feature = ""; to_feature = ""
+        for (i = 1; i <= length(from_parts); i++) {
+          if (from_parts[i] == "features" && i < length(from_parts)) { from_feature = from_parts[i+1]; break }
+        }
+        for (i = 1; i <= length(to_parts); i++) {
+          if (to_parts[i] == "features" && i < length(to_parts)) { to_feature = to_parts[i+1]; break }
+        }
+        if (from_feature != "" && to_feature != "" && from_feature != to_feature) {
+          key = from_feature "\t" to_feature
+          if (!(key in seen)) {
+            seen[key] = 1
+            printf "%s\t%s\t%s\n", from_feature, to_feature, from
+          }
+        }
+      }
+    ' | while IFS=$'\t' read -r from_feat to_feat file; do
+      add_finding "circular" "info" "cross_feature_import" "$REPO_ROOT/$file" 1 \
+        "Feature '$from_feat' imports from feature '$to_feat'" \
+        "Feature modules should communicate through shared interfaces, not import directly from each other"
+    done
+  fi
+}
+
+# ═══════════════════════════════════════════════════════════════
+# 2. UNUSED EXPORTS DETECTION
+# ═══════════════════════════════════════════════════════════════
+check_unused() {
+  local exports_file="$TMPDIR_WORK/exports.txt"
+  local imports_file="$TMPDIR_WORK/all_imports.txt"
+
+  # Collect all exported symbols
+  while IFS= read -r file; do
+    # Named exports: export const/function/class/type/interface/enum NAME
+    grep -n "^export " "$file" 2>/dev/null | \
+      grep -v "^export default\|^export \*\|^export {" | \
+      sed 's/.*export \(const\|let\|var\|function\|class\|type\|interface\|enum\|async function\) //' | \
+      sed 's/[^a-zA-Z0-9_].*//' | while read -r name; do
+        [ -n "$name" ] && echo "$name	$(relpath "$file")"
+      done
+
+    # Re-exports from index files are not counted as "unused"
+  done < "$FILES_LIST" > "$exports_file"
+
+  # Collect all import references across all files
+  while IFS= read -r file; do
+    grep "^import " "$file" 2>/dev/null | \
+      sed 's/.*{//; s/}.*//' | tr ',' '\n' | \
+      sed 's/[[:space:]]*as[[:space:]].*//; s/^[[:space:]]*//; s/[[:space:]]*$//' | \
+      grep -v '^$' | while read -r name; do
+        echo "$name"
+      done
+  done < "$FILES_LIST" | sort -u > "$imports_file"
+
+  # Find exports that are never imported anywhere
+  while IFS=$'\t' read -r name file; do
+    [ -z "$name" ] && continue
+    # Skip common framework exports (decorators, modules, etc.)
+    case "$name" in
+      Module|Controller|Service|Guard|Pipe|Filter|Interceptor|Gateway) continue ;;
+      App*|Main|bootstrap) continue ;;
+    esac
+
+    # Check if this export is referenced in any other file
+    if ! grep -q "^${name}$" "$imports_file" 2>/dev/null; then
+      # Double check: is it used in a dynamic import, decorator, or non-standard way?
+      local used
+      used=$(grep -rl "\b${name}\b" $(cat "$FILES_LIST") 2>/dev/null | grep -v "^${REPO_ROOT}/${file}$" | head -1)
+      if [ -z "$used" ]; then
+        local lineno
+        lineno=$(grep -n "export.*\b${name}\b" "$REPO_ROOT/$file" 2>/dev/null | head -1 | cut -d: -f1)
+        lineno=${lineno:-1}
+        add_finding "unused" "info" "unused_export" "$REPO_ROOT/$file" "$lineno" \
+          "Exported symbol '$name' is not imported by any other file" \
+          "Remove the export keyword if only used internally, or delete if truly unused"
+      fi
+    fi
+  done < "$exports_file"
+}
+
+# ═══════════════════════════════════════════════════════════════
+# 3. ORPHAN FILE DETECTION
+# ═══════════════════════════════════════════════════════════════
+check_orphans() {
+  # Files that are never imported by any other file (potential dead code)
+  local all_imports="$TMPDIR_WORK/all_import_paths.txt"
+
+  # Collect all import targets
+  while IFS= read -r file; do
+    grep "^import.*from ['\"]" "$file" 2>/dev/null | \
+      sed "s/.*from ['\"]//; s/['\"].*//" | while read -r import_path; do
+        echo "$import_path"
+      done
+  done < "$FILES_LIST" | sort -u > "$all_imports"
+
+  # Check each file
+  while IFS= read -r file; do
+    local rel
+    rel=$(relpath "$file")
+    local basename_no_ext
+    basename_no_ext=$(basename "$file" | sed 's/\.[^.]*$//')
+
+    # Skip entry points and special files
+    case "$basename_no_ext" in
+      main|index|app|App|layout|page|loading|error|not-found) continue ;;
+      *.module|*.spec|*.test|*.config|*.d) continue ;;
+      _*) continue ;; # Convention: _ prefix = internal
+    esac
+    [[ "$file" == *"/index."* ]] && continue
+    [[ "$file" == *".module."* ]] && continue
+    [[ "$file" == *".spec."* || "$file" == *".test."* ]] && continue
+    [[ "$file" == *".config."* ]] && continue
+
+    # Check if this file is imported by anything
+    # Try various import path patterns
+    local found=false
+    local rel_no_ext
+    rel_no_ext=$(echo "$rel" | sed 's/\.[^.]*$//')
+
+    # Check if any import path matches this file
+    if grep -qF "$basename_no_ext" "$all_imports" 2>/dev/null; then
+      found=true
+    fi
+
+    # Also check for directory-relative imports that could point to this file
+    if [ "$found" = false ]; then
+      local dir_name
+      dir_name=$(basename "$(dirname "$file")")
+      if grep -qF "$dir_name/$basename_no_ext" "$all_imports" 2>/dev/null; then
+        found=true
+      fi
+    fi
+
+    if [ "$found" = false ]; then
+      add_finding "orphans" "info" "orphan_file" "$file" 1 \
+        "File '$(basename "$file")' is not imported by any other file" \
+        "Verify this file is used (may be dynamically loaded, or could be dead code to remove)"
+    fi
+  done < "$FILES_LIST"
+}
+
+# ═══════════════════════════════════════════════════════════════
+# 4. IMPORT BOUNDARY VIOLATIONS
+# ═══════════════════════════════════════════════════════════════
+check_boundaries() {
+  while IFS= read -r file; do
+    local rel
+    rel=$(relpath "$file")
+
+    # Rule 1: Don't import from node_modules internals (deep imports)
+    grep -n "from '[^']*node_modules" "$file" 2>/dev/null | while IFS=: read -r lineno content; do
+      add_finding "boundaries" "warning" "node_modules_deep_import" "$file" "$lineno" \
+        "Direct import from node_modules path" \
+        "Import from the package name, not from node_modules internal path"
+    done
+
+    # Rule 2: Don't import from parent project (e.g., mobile importing from backend)
+    if [[ "$file" == *"/styai-mobile/"* ]]; then
+      grep -n "from '.*be-easycloset\|from '.*web-ui-easycloset" "$file" 2>/dev/null | while IFS=: read -r lineno content; do
+        add_finding "boundaries" "critical" "cross_project_import" "$file" "$lineno" \
+          "Mobile app imports directly from another project" \
+          "Use shared packages (auth-package) or API calls instead of direct imports"
+      done
+    fi
+
+    if [[ "$file" == *"/web-ui-easycloset/"* ]]; then
+      grep -n "from '.*be-easycloset\|from '.*styai-mobile" "$file" 2>/dev/null | while IFS=: read -r lineno content; do
+        add_finding "boundaries" "critical" "cross_project_import" "$file" "$lineno" \
+          "Frontend imports directly from another project" \
+          "Use shared packages or API calls instead of direct imports"
+      done
+    fi
+
+    # Rule 3: Don't import test utilities in production code
+    [[ "$file" == *".spec."* || "$file" == *".test."* ]] && continue
+    grep -n "from '.*__tests__\|from '.*\.spec\|from '.*\.test\|from '@testing-library\|from 'jest'" "$file" 2>/dev/null | while IFS=: read -r lineno content; do
+      add_finding "boundaries" "warning" "test_import_in_prod" "$file" "$lineno" \
+        "Test utility imported in production code" \
+        "Move test-related code to test files"
+    done
+
+    # Rule 4: Don't use relative imports that go too far up (../../../..)
+    grep -n "from '\.\./\.\./\.\./\.\." "$file" 2>/dev/null | while IFS=: read -r lineno content; do
+      add_finding "boundaries" "info" "deep_relative_import" "$file" "$lineno" \
+        "Deep relative import (4+ levels up)" \
+        "Use path aliases (@/, @features/, @shared/) instead of deep relative imports"
+    done
+
+    # Rule 5: Backend internal imports should use path aliases
+    if [[ "$file" == *"/be-easycloset/src/"* ]]; then
+      grep -n "from '\.\./\.\./\.\." "$file" 2>/dev/null | \
+        grep -v "\.spec\.\|\.test\." | while IFS=: read -r lineno content; do
+          add_finding "boundaries" "info" "missing_path_alias" "$file" "$lineno" \
+            "Deep relative import — consider using path alias" \
+            "Use @features/, @shared/, @core/ path aliases instead of ../../.."
+        done
+    fi
+
+  done < "$FILES_LIST"
+}
+
+# ═══════════════════════════════════════════════════════════════
+# RUN CHECKS
+# ═══════════════════════════════════════════════════════════════
+
+if [ "$CHECK" = "all" ] || [ "$CHECK" = "circular" ]; then
+  check_circular
+fi
+
+if [ "$CHECK" = "all" ] || [ "$CHECK" = "unused" ]; then
+  check_unused
+fi
+
+if [ "$CHECK" = "all" ] || [ "$CHECK" = "orphans" ]; then
+  check_orphans
+fi
+
+if [ "$CHECK" = "all" ] || [ "$CHECK" = "boundaries" ]; then
+  check_boundaries
+fi
+
+# ═══════════════════════════════════════════════════════════════
+# OUTPUT
+# ═══════════════════════════════════════════════════════════════
+
+TOTAL=$(wc -l < "$FINDINGS_FILE" | tr -d ' ')
+CRITICAL=$(count_matches '"severity":"critical"' "$FINDINGS_FILE")
+WARNINGS=$(count_matches '"severity":"warning"' "$FINDINGS_FILE")
+INFO=$(count_matches '"severity":"info"' "$FINDINGS_FILE")
+
+CIRCULAR_COUNT=$(count_matches '"check":"circular"' "$FINDINGS_FILE")
+UNUSED_COUNT=$(count_matches '"check":"unused"' "$FINDINGS_FILE")
+ORPHAN_COUNT=$(count_matches '"check":"orphans"' "$FINDINGS_FILE")
+BOUNDARY_COUNT=$(count_matches '"check":"boundaries"' "$FINDINGS_FILE")
+
+if [ "$SUMMARY_ONLY" = true ]; then
+  cat <<EOF
+{
+  "scan": "dep-check",
+  "timestamp": "$(date -u +%Y-%m-%dT%H:%M:%SZ)",
+  "target": "$TARGET",
+  "files_scanned": $FILE_COUNT,
+  "summary": {
+    "total": $TOTAL,
+    "critical": $CRITICAL,
+    "warning": $WARNINGS,
+    "info": $INFO,
+    "by_check": {
+      "circular": $CIRCULAR_COUNT,
+      "unused": $UNUSED_COUNT,
+      "orphans": $ORPHAN_COUNT,
+      "boundaries": $BOUNDARY_COUNT
+    }
+  }
+}
+EOF
+else
+  {
+    echo '{'
+    echo '  "scan": "dep-check",'
+    echo '  "timestamp": "'"$(date -u +%Y-%m-%dT%H:%M:%SZ)"'",'
+    echo '  "target": "'"$TARGET"'",'
+    echo '  "files_scanned": '"$FILE_COUNT"','
+    echo '  "summary": {'
+    echo '    "total": '"$TOTAL"','
+    echo '    "critical": '"$CRITICAL"','
+    echo '    "warning": '"$WARNINGS"','
+    echo '    "info": '"$INFO"','
+    echo '    "by_check": {'
+    echo '      "circular": '"$CIRCULAR_COUNT"','
+    echo '      "unused": '"$UNUSED_COUNT"','
+    echo '      "orphans": '"$ORPHAN_COUNT"','
+    echo '      "boundaries": '"$BOUNDARY_COUNT"
+    echo '    }'
+    echo '  },'
+    echo '  "findings": ['
+
+    first=true
+    while IFS= read -r line; do
+      [ -z "$line" ] && continue
+      if [ "$first" = true ]; then
+        first=false
+      else
+        echo ','
+      fi
+      printf '    %s' "$line"
+    done < "$FINDINGS_FILE"
+
+    echo ''
+    echo '  ]'
+    echo '}'
+  }
+fi
+
+# ─── Exit code ──────────────────────────────────────────────
+if [ "$CRITICAL" -gt 0 ]; then
+  exit 2
+elif [ "$TOTAL" -gt 0 ]; then
+  exit 1
+else
+  exit 0
+fi