Add support for jpeg

src/parsers/jpg.py

+import subprocess
+import json
+
+import gi
+gi.require_version('GdkPixbuf', '2.0')
+from gi.repository import GdkPixbuf
+
+from . import abstract
+
+class JPGParser(abstract.AbstractParser):
+    mimetypes = {'image/jpg', }
+    meta_whitelist = {'SourceFile', 'ExifToolVersion', 'FileName',
+            'Directory', 'FileSize', 'FileModifyDate', 'FileAccessDate',
+            "FileInodeChangeDate", 'FilePermissions', 'FileType',
+            'FileTypeExtension', 'MIMEType', 'ImageWidth',
+            'ImageSize', 'BitsPerSample', 'ColorComponents', 'EncodingProcess',
+            'JFIFVersion', 'ResolutionUnit', 'XResolution', 'YCbCrSubSampling',
+            'YResolution', 'Megapixels', 'ImageHeight'}
+
+    def get_meta(self):
+        out = subprocess.check_output(['exiftool', '-json', self.filename])
+        meta = json.loads(out)[0]
+        for key in self.meta_whitelist:
+            meta.pop(key, None)
+        return meta
+
+    def remove_all(self):
+        pixbuf = GdkPixbuf.Pixbuf.new_from_file(self.filename)
+        pixbuf.savev(self.output_filename, "jpeg", ["quality"], ["100"])
+        return True

tests/test_libmat2.py

@@ -5,7 +5,7 @@ import shutil
 import os
 
 from src import parsers
-from src.parsers import pdf, png
+from src.parsers import pdf, png, jpg
 
 class TestGetMeta(unittest.TestCase):
     def test_pdf(self):
@@ -20,6 +20,11 @@ class TestGetMeta(unittest.TestCase):
         self.assertEqual(meta['Comment'], 'This is a comment, be careful!')
         self.assertEqual(meta['ModifyDate'], "2018:03:20 21:59:25")
 
+    def test_jpg(self):
+        p = jpg.JPGParser('./tests/data/dirty.jpg')
+        meta = p.get_meta()
+        self.assertEqual(meta['Comment'], 'Created with GIMP')
+
 class TestCleaning(unittest.TestCase):
     def test_pdf(self):
         shutil.copy('./tests/data/dirty.pdf', './tests/data/clean.pdf')
@@ -51,3 +56,19 @@ class TestCleaning(unittest.TestCase):
         self.assertEqual(p.get_meta(), {})
 
         os.remove('./tests/data/clean.png')
+
+
+    def test_jpg(self):
+        shutil.copy('./tests/data/dirty.jpg', './tests/data/clean.jpg')
+        p = jpg.JPGParser('./tests/data/clean.jpg')
+
+        meta = p.get_meta()
+        self.assertEqual(meta['Comment'], 'Created with GIMP')
+
+        ret = p.remove_all()
+        self.assertTrue(ret)
+
+        p = jpg.JPGParser('./tests/data/clean.jpg.cleaned')
+        self.assertEqual(p.get_meta(), {})
+
+        os.remove('./tests/data/clean.jpg')