expand .jars in the xpi and whitelist .css and .jar (bug 563389)

apps/files/helpers.py

@@ -50,11 +50,11 @@ class FileViewer:
         Raises error on nasty files.
         """
         try:
-            os.makedirs(self.dest)
-        except OSError:
-            return
+            os.makedirs(os.path.dirname(self.dest))
+        except OSError, err:
+            pass
         try:
-            extract_xpi(self.src, self.dest)
+            extract_xpi(self.src, self.dest, expand=True)
         except Exception, err:
             task_log.error('Error (%s) extracting %s' % (err, self.src))
 
@@ -71,7 +71,7 @@ class FileViewer:
         """Uses the filename to see if the file can be shown in HTML or not."""
         if mimetype:
             major, minor = mimetype.split('/')
-            if major == 'text' and minor in ['plain', 'html']:
+            if major == 'text' and minor in ['plain', 'html', 'css']:
                 return False
             elif minor in ['xml', 'rdf+xml', 'javascript', 'x-javascript',
                          'xml-dtd', 'vnd.mozilla.xul+xml']:

apps/files/tests/test_helpers.py

@@ -12,23 +12,25 @@ import test_utils
 
 from amo.urlresolvers import reverse
 from files.helpers import FileViewer, DiffHelper
+from files.utils import extract_zip
+
+root = os.path.join(settings.ROOT, 'apps/files/fixtures/files')
+dictionary = '%s/dictionary-test.xpi' % root
+recurse = '%s/recurse.xpi' % root
 
 
 class TestFileHelper(test_utils.TestCase):
 
     def setUp(self):
-        dictionary = 'apps/files/fixtures/files/dictionary-test.xpi'
-        src = os.path.join(settings.ROOT, dictionary)
-
         file_obj = Mock()
         file_obj.id = file_obj.pk = 1
-        file_obj.file_path = src
-
-        self.viewer = FileViewer(file_obj)
+        file_obj.file_path = dictionary
 
         self.old_tmp = settings.TMP_PATH
         settings.TMP_PATH = tempfile.mkdtemp()
 
+        self.viewer = FileViewer(file_obj)
+
     def tearDown(self):
         self.viewer.cleanup()
         settings.TMP_PATH = self.old_tmp
@@ -40,6 +42,20 @@ class TestFileHelper(test_utils.TestCase):
         self.viewer.extract()
         eq_(self.viewer.is_extracted, True)
 
+    def test_recurse_extract(self):
+        self.viewer.src = recurse
+        self.viewer.extract()
+        eq_(self.viewer.is_extracted, True)
+
+    def test_recurse_contents(self):
+        self.viewer.src = recurse
+        self.viewer.extract()
+        files = self.viewer.get_files()
+        for name in ['chrome/test-root.txt', 'chrome/test.jar',
+                     'chrome/test.jar/test/test.text']:
+            eq_(name in files, True, 'File %r not extracted' % name)
+        eq_(files['chrome/test.jar/test']['directory'], True)
+
     def test_cleanup(self):
         self.viewer.extract()
         self.viewer.cleanup()

apps/files/utils.py

@@ -135,13 +135,53 @@ def parse_search(filename, addon=None):
             'version': datetime.now().strftime('%Y%m%d')}
 
 
-def extract_xpi(xpi, path):
-    zip = zipfile.ZipFile(xpi)
+def extract_zip(source, remove=False):
+    """Extracts the zip file. If remove is given, removes the source file."""
+    tempdir = tempfile.mkdtemp()
+    zip = zipfile.ZipFile(source)
     for f in zip.namelist():
         if '..' in f or f.startswith('/'):
-            log.error('Extraction error, Invalid archive: %s' % xpi)
+            log.error('Extraction error, Invalid archive: %s' % source)
             raise forms.ValidationError(_('Invalid archive.'))
-    zip.extractall(path)
+    zip.extractall(tempdir)
+    if remove:
+        os.remove(source)
+    return tempdir
+
+
+def copy_over(source, dest):
+    """
+    Copies from the source to the destination, removing the destination
+    if it exists and is a directory.
+    """
+    if os.path.exists(dest) and os.path.isdir(dest):
+        shutil.rmtree(dest)
+    shutil.copytree(source, dest)
+    shutil.rmtree(source)
+
+
+def extract_xpi(xpi, path, expand=False):
+    """
+    If expand is given, will look inside the expanded file
+    and find anything in the whitelist and try and expand it as well.
+    Currently only does one iteration of this.
+
+    It will replace the expanded file with a directory and the expanded
+    contents. If you have 'foo.jar', that contains 'some-image.jpg', then
+    it will create a folder, foo.jar, with an image inside.
+    """
+    expand_whitelist = ['.jar']
+    tempdir = extract_zip(xpi)
+
+    if expand:
+        for root, dirs, files in os.walk(tempdir):
+            for name in files:
+                if os.path.splitext(name)[1] in expand_whitelist:
+                    src = os.path.join(root, name)
+                    dest = extract_zip(src, remove=True)
+                    copy_over(dest, src)
+
+    copy_over(tempdir, path)
 
 
 def parse_xpi(xpi, addon=None):