From 4f197877caca0cda7586bd5947c9839a0afda40b Mon Sep 17 00:00:00 2001
From: Jean-Marie Mineau <jean-marie.mineau@inria.fr>
Date: Tue, 5 Nov 2024 16:01:34 +0100
Subject: [PATCH] check smali

---
 android_class_shadowing_scanner/__init__.py | 107 +++++++++++++++++++-
 pyproject.toml                              |   1 +
 2 files changed, 107 insertions(+), 1 deletion(-)

diff --git a/android_class_shadowing_scanner/__init__.py b/android_class_shadowing_scanner/__init__.py
index 8be090a..df2df0d 100644
--- a/android_class_shadowing_scanner/__init__.py
+++ b/android_class_shadowing_scanner/__init__.py
@@ -2,12 +2,16 @@ import zipfile
 import io
 import hashlib
 import json
+import sqlite3
+import tempfile
+import subprocess
 
 from argparse import ArgumentParser
 from pathlib import Path
 from getpass import getpass
 from datetime import datetime
 
+
 from .androzoo import download_apk
 from .data import ApkData, load_from_directory
 from .analysis import analyze
@@ -130,7 +134,7 @@ def main():
             print(entry.to_string())
         else:
             with (args.output_dir / sha256).open("w") as file:
-                file.write(entry)
+                file.write(entry.to_string())
 
     if apks:
         if args.json:
@@ -231,3 +235,104 @@ def collect_to_db():
     )
     args = parser.parse_args()
     load_from_directory(args.dir, args.db, args.androzoo_list)
+
+
+def check_smali():
+    parser = ArgumentParser(
+        prog="Smalli Check",
+        description="Check if duplicated classes have distinct smali",
+    )
+    parser.add_argument(
+        "--db",
+        help="Path to the database storing the results",
+        type=Path,
+        required=True,
+    )
+    key_parser = parser.add_mutually_exclusive_group(required=False)
+    key_parser.add_argument(
+        "--api-key-file",
+        help="The path to a file containing the Androzoo API key",
+        type=Path,
+    )
+    key_parser.add_argument(
+        "--api-key", help="The Androzoo API key (Usage NOT recommanded)", type=str
+    )
+    SECRET_STORAGE_IMPORTED = False
+    try:
+        import secretstorage
+
+        SECRET_STORAGE_IMPORTED = True
+
+        key_parser.add_argument(
+            "--api-key-keyring-id",
+            help="The ID of the Androzoo API key in the secret service storage",
+            type=str,
+        )
+    except ModuleNotFoundError:
+        pass
+    args = parser.parse_args()
+
+    api_key = ""
+    if args.api_key:
+        api_key = args.api_key
+    if args.api_key_file:
+        with args.api_key_file.open("r") as file:
+            api_key = file.read().strip()
+    if SECRET_STORAGE_IMPORTED and not api_key:
+        if args.api_key_keyring_id:
+            key_id = args.api_key_keyring_id
+        else:
+            key_id = "androzoo"
+        try:
+            with secretstorage.dbus_init() as connection:
+                collection = secretstorage.get_default_collection(connection)
+                item = next(collection.search_items({"Title": key_id}))
+                item.unlock()
+                api_key = item.get_secret().decode("utf-8").strip()
+        except:
+            pass
+    if not api_key:
+        api_key = getpass(prompt="Androzoo API key: ").strip()
+
+    with sqlite3.connect(args.db) as conn:
+        apks = list(
+            map(
+                lambda t: t[0],
+                conn.execute("SELECT sha256 FROM data WHERE nb_duplicate_classes >= 1"),
+            )
+        )
+    data = {}
+    for sha256 in apks:
+        with tempfile.TemporaryDirectory() as tmpdirname:
+            d = Path(tmpdirname)
+            apk_bin = download_apk(sha256, api_key, logfile=None)
+            if apk_bin is None:
+                continue
+            with (d / "app.apk").open("wb") as fp:
+                fp.write(apk_bin)
+            with zipfile.ZipFile(io.BytesIO(apk_bin)) as apk:
+                json_data[sha256] = {}
+                entry = analyze(apk, sha256, json_out=json_data[sha256])
+            subprocess.run(["apktool", "d", "app.apk", "-o", "apktool_out"], cwd=d)
+            smalli_dirs = []
+            for dex in json_data[sha256]["class_dex"]:
+                if dex == "classes.dex":
+                    smalli_dirs.append(out / "apktool_out" / "smali")
+                else:
+                    smalli_dirs.append(
+                        out / "apktool_out" / "smali_" + dex.removesuffix(".dex")
+                    )
+            dist_dup_classes = set()
+            for cl in json_data[sha256]["duplicated_classes"]:
+                cl_f = cl.removesuffix(";").removeprefix("L") + ".smali"
+                smali = None
+                for cdir in smalli_dirs:
+                    if (cdir / cl_f).exists():
+                        with (cdir / cl_f).open() as file:
+                            smali_new = file.read()
+                        if smali is None:
+                            smali = smali_new
+                        elif smali != smali_new:
+                            dist_dup_classes.add(cl)
+            json_data[sha256]["redef_classes"] = list(dist_dup_classes)
+    print(json.dumps(json_data))
diff --git a/pyproject.toml b/pyproject.toml
index 4a14ffd..87171bc 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -21,3 +21,4 @@ build-backend = "poetry.core.masonry.api"
 [tool.poetry.scripts]
 scan  = 'android_class_shadowing_scanner.__init__:main'
 collect-scan  = 'android_class_shadowing_scanner.__init__:collect_to_db'
+check-class-redef = 'android_class_shadowing_scanner.__init__:check_smali'