pre-commit

#!/usr/bin/env bash
# Codechu shared pre-commit hook — Python AST tabanlı i18n leak guard +
# conflict marker + large-file warn + .mo sync hint.
#
# Kurulum (her ürün repo'sunda):
#   cp /path/to/codechu-org/hooks/pre-commit .git/hooks/pre-commit
#   chmod +x .git/hooks/pre-commit
#
# Atlama: git commit --no-verify (sadece acil)

set -e
FAILED=0

# ─── 1. i18n leak guard (Python AST) ─────────────────────────────────
# Source kodda Türkçe-spesifik karakter içeren string literal varsa,
# docstring değilse, _()-wrapped değilse fail.
# Bilinen kasıtlı literal'lar exemption list'inde.
SRC_DIR=${SRC_DIR:-}
if [ -z "$SRC_DIR" ]; then
    for candidate in disk_cleaner src lib; do
        [ -d "$candidate" ] && SRC_DIR="$candidate" && break
    done
fi

if [ -n "$SRC_DIR" ] && [ -d "$SRC_DIR" ] && command -v python3 >/dev/null; then
    LEAKS=$(python3 - "$SRC_DIR" <<'PY'
import ast
import os
import re
import sys

TURKISH = re.compile(r'[şŞıİğĞçÇöÖüÜ]')

# Kasıtlı Türkçe literal'lar (UI dışı, source code mantığı için)
EXEMPT_LITERALS = {
    "Türkçe",       # native language name for picker
    "AKTİF",        # backward-compat data marker
    "İndirilenler", # standard Turkish locale dir name (file path)
    "Resimler",     # standard Turkish locale dir name
}

EXEMPT_FILES = {"i18n.py"}  # i18n wrapper has Turkish examples in comments

src = sys.argv[1]
problems = []

for root, _, files in os.walk(src):
    for f in files:
        if not f.endswith(".py"):
            continue
        if f in EXEMPT_FILES:
            continue
        path = os.path.join(root, f)
        try:
            with open(path, encoding="utf-8") as fp:
                source = fp.read()
            tree = ast.parse(source)
        except SyntaxError:
            continue

        # Map: docstring node IDs (skip them)
        docstring_ids = set()
        for node in ast.walk(tree):
            if isinstance(node, (ast.Module, ast.FunctionDef, ast.AsyncFunctionDef, ast.ClassDef)):
                ds = ast.get_docstring(node, clean=False)
                if ds and node.body:
                    first = node.body[0]
                    if isinstance(first, ast.Expr) and isinstance(first.value, ast.Constant):
                        docstring_ids.add(id(first.value))

        # _()-wrapped IDs (gettext call)
        wrapped_ids = set()
        for node in ast.walk(tree):
            if isinstance(node, ast.Call):
                fname = ""
                if isinstance(node.func, ast.Name):
                    fname = node.func.id
                elif isinstance(node.func, ast.Attribute):
                    fname = node.func.attr
                if fname in ("_", "gettext", "ngettext", "pgettext"):
                    for arg in node.args:
                        if isinstance(arg, ast.Constant):
                            wrapped_ids.add(id(arg))

        # Walk string Constants
        for node in ast.walk(tree):
            if not isinstance(node, ast.Constant):
                continue
            if not isinstance(node.value, str):
                continue
            if id(node) in docstring_ids or id(node) in wrapped_ids:
                continue
            text = node.value
            if not TURKISH.search(text):
                continue
            if text in EXEMPT_LITERALS:
                continue
            # Substring of exempt literal? (örn. "AKTİF proje" — kısmi compat)
            if any(lit in text for lit in EXEMPT_LITERALS):
                continue
            problems.append(f"{path}:{node.lineno}: {text!r}")

for p in problems:
    print(p)
PY
)
    if [ -n "$LEAKS" ]; then
        echo "✗ i18n leak: Türkçe string literal — _() ile sar veya EXEMPT_LITERALS'a ekle:"
        echo "$LEAKS"
        FAILED=1
    fi
fi

# ─── 2. Conflict marker / whitespace check ──────────────────────────
if ! git diff --cached --check >/dev/null 2>&1; then
    echo "✗ Whitespace error or conflict marker:"
    git diff --cached --check
    FAILED=1
fi

# ─── 3. Large file warn (>1MB) ─────────────────────────────────────
big=$(git diff --cached --name-only --diff-filter=A | \
      xargs -I{} sh -c 'test -f "{}" && find "{}" -size +1M 2>/dev/null' 2>/dev/null | head -5)
if [ -n "$big" ]; then
    echo "⚠ Büyük dosya eklenmiş (>1MB) — git LFS düşün:"
    echo "$big"
fi

# ─── 4. .po → .mo sync hint ─────────────────────────────────────────
if [ -d "po" ] && [ -f "po/messages.pot" ]; then
    if git diff --cached --name-only | grep -q "po/.*\.po$"; then
        if ! git diff --cached --name-only | grep -q "locale/.*\.mo$"; then
            echo "⚠ .po değişti ama .mo regenerate edilmedi — 'cd po && make compile'"
        fi
    fi
fi

if [ "$FAILED" = "1" ]; then
    echo
    echo "Commit reddedildi. Düzeltip tekrar deneyin (veya --no-verify acil durumda)."
    exit 1
fi

exit 0