.gitlab-ci.yml

@@ -35,8 +35,7 @@ mypy:
   - apt-get -qqy update
   - apt-get -qqy install --no-install-recommends python3-pip
   - pip3 install mypy
-  - mypy mat2 libmat2/*.py --ignore-missing-imports
-  - mypy --ignore-missing-imports ./nautilus/mat2.py
+  - mypy --ignore-missing-imports mat2 libmat2/*.py ./nautilus/mat2.py
 
 tests:debian:
   stage: test

CHANGELOG.md

+# 0.6.0 - 2018-11-10
+
+- Add lightweight cleaning for jpeg
+- Add support for zip files
+- Add support for mp4 files
+- Improve metadata extraction for archives
+- Improve robustness against corrupted embedded files
+- Fix a possible security issue on some terminals (control character
+	injection via --show)
+- Various internal cleanup/improvements
+
 # 0.5.0 - 2018-10-23
 
 - Video (.avi files for now) support, via FFmpeg, optionally
@@ -5,7 +16,7 @@
 - Processing files starting with a dash is now quicker
 - Metadata are now displayed sorted
 - Recursive metadata support for FLAC files
-- Unsupported extensions aren't displayed in `/.mat -l` anymore
+- Unsupported extensions aren't displayed in `./mat2 -l` anymore
 - Improve the display when no metadata are found
 - Update the logo according to the GNOME guidelines
 - The testsuite is now runnable on the installed version of mat2

debian/changelog

+mat2 (0.6.0-1) unstable; urgency=medium
+
+  * New upstream release.
+  * d/control:
+    - Add debhelper-compat as build dependency.
+    - Declare that the build doesn't need root privileges.
+    - Mention '.mp4' support in description.
+  * d/compat:
+    - Drop obsolete compat control file.
+  * d/gitlab-ci.yml:
+    - Drop custom config, rely on the "Salsa CI Team".
+  * d/u/signing-key.asc:
+    - Use clean and minimal key to remove cruft. Thanks, lintian.
+  * d/watch:
+    - Use '@ARCHIVE_EXT@' substitution.
+    - Relax archive name regex, due to inconsistent upstream naming scheme.
+
+ -- Georg Faerber <georg@riseup.net>  Sat, 10 Nov 2018 13:33:52 +0000
+
 mat2 (0.5.0-1) unstable; urgency=medium
 
   * New upstream release.

debian/control

@@ -66,3 +66,4 @@ Description: Metadata anonymisation toolkit v2
     - Free Lossless Audio Codec (.flac)
     - Torrent (.torrent)
     - Audio Video Interleave (.avi)
+    - MPEG-4 (.mp4)

debian/gitlab-ci.yml

-image: debian:sid
+include: https://salsa.debian.org/salsa-ci-team/pipeline/raw/master/salsa-ci.yml
 
 build:
-  script:
-    - export DEBIAN_FRONTEND=noninteractive
-    - apt-get update
-    - apt-get -y upgrade
-    - apt-get -y dist-upgrade
-    - export BUILD_DEPENDS=`perl -ne 'next if /^#/; $p=(s/^Build-Depends:\s*/ / or (/^ / and $p)); s/,|\n|\([^)]+\)//mg; print if $p' < debian/control`
-    - apt-get install -y --no-install-recommends build-essential dpkg-dev fakeroot git-buildpackage lintian pristine-tar $BUILD_DEPENDS
-    - apt-get -y autoremove --purge
-    - apt-get clean
-    - rm -rf /var/lib/apt/lists/*
-    - dpkg-checkbuilddeps
-    - git checkout pristine-tar
-    - echo "make branch current for gbp"
-    - git checkout -B "$CI_COMMIT_REF_NAME" "$CI_COMMIT_SHA"
-    - gbp buildpackage -us -uc --git-ignore-branch --lintian-opts -iIE --pedantic
-    - mkdir results
-    - cp ../*.deb results/
-    - cp ../*.dsc results/
-    - cp ../*.changes results/
-  artifacts:
-    paths:
-    - results
-    expire_in: 3 days
+    extends: .build-unstable
+
+lintian:
+    extends: .test-lintian
+
+piuparts:
+    extends: .test-piuparts
+
+autopkgtest:
+    extends: .test-autopkgtest
+
+reprotest:
+    extends: .test-reprotest

debian/upstream/signing-key.asc

@@ -11,67 +11,67 @@ lQJUCkGFxyIpcDNRnf3ApjT4+QuEaw98tKvgRzCozFx2D94wsSFz858vZrdYj4pt
 BYnPNFBEeGMRWkxuVwV1o+WKNJfwg2UcDghSkJGBCPCAiC2fDlfyk3njjLjxZHP/
 mYNwUkxTlQolzknJZ5wg7vbE6r4rfQX4gTi3mNzYtqUAb17GIczOARZK7qdSapOb
 rXPFGgX3Bd4FZJEaIq3p5xWcWS8fcMveoYO7m9cyaSkSQxAPrPZE3hDF1QARAQAB
-tDFKdWxpZW4gKGp2b2lzaW4pIFZvaXNpbiA8anZvaXNpbkBvcGVubWFpbGJveC5u
-ZXQ+iQI9BBMBCgAnAhsDBQkJZgGABQsJCAcDBRUKCQgLBRYCAwEAAh4BAheABQJV
-vnKvAAoJEATQQegXGQHMRUQP/jr1Wy2x6lBTFZz/bg9b4G+FB98V/OcDmVNEoSnL
-HKEg/X6Z/ePrLOiOxX15owqhWtXZbAC83SkkqZL0/xhZ8G8gVIHej4+Wm8gwza9q
-x/vG9Z4KE2Ed1zrBtOOuVMjTbANaHCEmfbVdoEVkAU+s9P1V/vx2Zff4kFmOqJv3
-btj+pRynjfwie5r53oJGfM5l6/3/VMazMuDSP6a5REBs5b19rgTWI2ocJvZxeemA
-iinuklyPYjCNMMj43d5vSKZX8vV25mtTP1szbKHvaTHId5k4D8hgozkyVp5iAQUj
-vGRgWJHFf+7bKDf8hd94dW6Aoi5gVIMEfLwo7/1HOIuv4wFWehqmg0N96dPj87JN
-qyZbprBLxC1pjqXwhP4e+yp4s7pzPc6HgWQwxZSxodINDlvFlPydDvRwexcsG4Yw
-OhLvuu2HohmjfjyjzsLkOkKML20nHvMSt5aBz2H1zdbGNKAoxpgBL03e06gkWH3R
-fWOy7RBG0iZJrA9/MQH0i1A+BzGqQ8ZmcC1zPQNJHSN0duec4za+7FA820Y9wgEV
-8Sv2dv2iyuouMjlALtyBHY1faJgVHDt6HzTfj6kxqGILGmrKz+VVIH/4QHq2I7VT
-fauIujDFEdtVHojFe2edHaOx3pyfNoVRkLre98AucNS/BDgKLO8UdnviSo3QXRn/
-tJmKtDJKdWxpZW4gKGp2b2lzaW4pIFZvaXNpbiA8anVsaWVuLnZvaXNpbkBkdXN0
-cmkub3JnPokCQAQTAQoAKgIbAwUJCWYBgAULCQgHAwUVCgkICwUWAgMBAAIeAQIX
-gAUCVb5yrwIZAQAKCRAE0EHoFxkBzKOaD/kBQIAeM/zCbwPkTqNjO3OLD1Jr49XD
-lcsuRiBO286SjntWyipPyDy4hc5TDrt5R8wkC4htYdGk+iqogoDc6q1pT6VDvOKp
-RVg4BypxwyRsn81RMNVz5LkqEavLAwpnwow3tHcElmg4SuFyEObZhpmhEsMpS5KS
-zQ3lDnAISQi0hw/xb3HbRDA5rKLs8f57OUgh3Dbl2IskRGVyhBrBaZBfTJVKpodH
-ZsZyH26z8dO19yLlukNVyVgifIiLc/Pd3Jh2LCXJFl4T++i7dmiZZqxWXVIS+EsS
-TZxZHW3lxrRBzU1mVJK/udJbdm3lfz4kwmkUgqjA+3loPIiyd3m1kA3Z4s2DmZmE
-GAR1r50+vQF0uwTluz64qPikt4X2eMQ4Fs3DSZVn9hA1oGU2g3+/V4h5OyVy5OSM
-nvwkpOrV7M+5ie34Ush1fRgIlKyme5W/DKHPU8HZ+C19mLVhd/aGQREP/SJJagrV
-xmAkSbdUGNbP7haDRm6/dxzE3FG5gZ4o5gKSAPfcwNTxPtdbIENfPRHWukkiovPN
-RvJtm+rV3TljJAsHh0tL+Hre2MoTBA0Qi6wccYcW7mleyJ7HLPJA9unFpmATVeyX
-F7pgbTYJ2RgbR1DebaPJ6RNtM57GHWAQ+wY0KUUJZi3luDb0goIZjXMoOPV18lFu
-+IQH/bWic0Y7cbQsSnVsaWVuIChqdm9pc2luKSBWb2lzaW4gPGp2b2lzaW5Acmlz
-ZXVwLm5ldD6JAj0EEwEKACcFAlWzyiYCGwMFCQlmAYAFCwkIBwMFFQoJCAsFFgID
-AQACHgECF4AACgkQBNBB6BcZAczsXA/7BSyvfqxCJ5FP+9348nZu0+FvvcE/C/hf
-LTFeyVJL85+JBi350RRcJmPIcrHZnWHqFG0uPO2UGCDATbTbxbVfL1Wvtqx51FGc
-iZ3bnon0MtEyDZjnQPCDDgQq4SPPlh8gvZcdKVC1ySBYEnikd1ATfzrjpx/jP7dd
-euV8vn1i8f7BbR9ySQJHdKd6/jMeQkvO3l1ZC7msFW41nTEbyuRbyRepWoZqThx8
-PMtDRVH57SPrEew9Komtje+Svq5WD/NRAENjh76FKzId2b3xnQXaP6SwqeOlsFNG
-0/SEkt+8MZh12/EJbOkeWUv6mGQP3hOFmcAc0pzzKv7kC8Irg9u9Fkt0zuZC7uvH
-7B4kPo/whjTR26JUss95HNOeS2XZwZ8kl8xbsXv4gEs/HDo68fl2KEy7ZD6umX3N
-R3Agp8SnPuNpfEwL7ZChtVzKa20StF7AuN4XxxeEpn+j6f8FdGzTSJhwP8IWoK/G
-ihpO07uIagNwJxAuWBcGM9r3j6Df2muwdiSLGkjm8dpdlxi2gfbuUhxiXvv3jDMn
-XCGveGeQoxwCRQEGwZMPYhOdvz4TTbNXNn/Olwj2Iv67xM6ZcvnzE4pq9RaVz9am
-OwIA7fKpGpY8NB5qfMGzLqW9UpUDNMk7IIOKZipUk3LI4CtOr8tt9JUVSg+v3dqc
-UtL8F1+KY/G5Ag0EVbPFoAEQAL+nkRJIiULqpUmfF+G0ilqmmOQeYWMwdaYWsHTk
-qea0YH8icYfR7nLY5VorPRESst1/YIDNTnNv1zIs+eoekTtJqq08GZqFMFfudXMg
-WtuMulT9c8dpasdrozWJ/K7XLaaxd7lTWWeg5WMldNWzKg0C3zDN/vcjhK3b6lju
-RAGW/kBROyxrhPbaFmzcEiZx3g6wuly3ZogZ3c0ah82SyIsHdmIT+8omRdYhvH3+
-Gbye4qVC4caj6weGjxP5BBxz5gKjroFgHOfMP7IS/9QAuOAdZT89fDdFUzySpDda
-WZaxgqpBiCRoDZxU+uRsuXGIeDbxjh167qhlHihKi9eMmTiOPeb9RyRo3R9zjsEb
-VXApUfGOD9zyHLV6W6eEm/RV/Z4sDqlzRR8BylKQlk2Q4uMVn1hE4d2IqJQTikR6
-Pj+jdYTvpMhfnxpRG6i5UBXB7XiwgQPdAug77jBzNNmw5f+Nw6DYPao+q5vLm1o4
-AzrI8/+/t/JMAb1Rh3GrUv3/ArCEfWmZqCROYu7uTlfOBD/qutHg0k9FEwROu2oM
-Uk1vZTGFGWeX8Jq+S3rkErsPRAn2f2mEVYLahMhox3my4VkrcjvVR15NoE6Wuqao
-vO4FWHNB4Ara9okjVzojOqMZGXE8SboG4d9FZp/4suPhnvCu6lbADEuMeA9QIvYK
-GTZnABEBAAGJAiUEGAEKAA8FAlWzxaACGwwFCQlmAYAACgkQBNBB6BcZAczeYxAA
-k14yHOWLV1gA1/nSc5wCcMTNzMnM2UD1BtytpJKQ/SXJcAWHgNk6HlH+pwEF7s13
-9OsZj3GkjJyOIRGYP8ZEdfetrnVlcz41tS2DcITZEbfiRiFIaMBMiCB1CIKeEjHf
-ERiF/Z+1lpCpSEaecwtuJrPySJO4XV2KplgxrVt5no26ahhBnf7kKP/gcaWs2Z1O
-NzplnMFWt7Vq5T2Ayr1p4+Zt/UMov6rRvJ4IcvlHIF6qLhwVi5yHFej2bpB55WQ9
-SceqV1CT1KWJ45I71DyBbPyW1NQUFtza+wyFI0F1ges4d/XvS4sdA3egvjbIdlwY
-Hq42wpGbVB+w2R03fL0n5wgRaSZlQsNAJdf3gFQkNFgBLnXQreTyVJq5TrzTCzNj
-TWIvhmJoFteX+Prh1t3Wyxrj5zYbL4GfN1w0gl5QGpWLPORn/UeoboR9tjDKga2d
-M4d/q/qeLemIZE+box4WFrmv35CZ47TbN7dz1X6nnSNkqUaMtZVqWJMFPyYezmSG
-KwdLl5xP54IQle/EhGoWVFViB//tfAaRTqeXBWK/U/bY+dvEfQZwpeswM2sy8rYg
-CvESHird2rOqtA7Sn/FkDRa23nrtZVJsB/dtBGEc5vNSJ1fI24iAY0UZE+ntcoTj
-INpYmBCcTk9aF3APuer6h65IbSZxlesI660drDlxHe4=
-=Ijk3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+=bpEm
 -----END PGP PUBLIC KEY BLOCK-----

debian/watch

 version=4
 
-opts="pgpmode=next" https://0xacab.org/jvoisin/mat2/tags (?:.*/)mat2-@ANY_VERSION@\.tar\.xz
+opts="pgpmode=next" https://0xacab.org/jvoisin/mat2/tags (?:.*/)mat.+-@ANY_VERSION@@ARCHIVE_EXT@
 
-opts="pgpmode=previous" https://0xacab.org/jvoisin/mat2/tags (?:.*/)mat2-@ANY_VERSION@@SIGNATURE_EXT@
+opts="pgpmode=previous" https://0xacab.org/jvoisin/mat2/tags (?:.*/)mat.+-@ANY_VERSION@@SIGNATURE_EXT@

doc/mat2.1

-.TH MAT2 "1" "October 2018" "MAT2 0.5.0" "User Commands"
+.TH MAT2 "1" "November 2018" "MAT2 0.6.0" "User Commands"
 
 .SH NAME
 mat2 \- the metadata anonymisation toolkit 2

libmat2/archive.py

@@ -67,6 +67,31 @@ class ArchiveBasedAbstractParser(abstract.AbstractParser):
 
         return metadata
 
+    def get_meta(self) -> Dict[str, Union[str, dict]]:
+        meta = dict()  # type: Dict[str, Union[str, dict]]
+
+        with zipfile.ZipFile(self.filename) as zin:
+            temp_folder = tempfile.mkdtemp()
+
+            for item in zin.infolist():
+                if item.filename[-1] == '/':  # pragma: no cover
+                    # `is_dir` is added in Python3.6
+                    continue  # don't keep empty folders
+
+                zin.extract(member=item, path=temp_folder)
+                full_path = os.path.join(temp_folder, item.filename)
+
+                tmp_parser, _ = parser_factory.get_parser(full_path)  # type: ignore
+                if not tmp_parser:
+                    continue
+
+                local_meta = tmp_parser.get_meta()
+                if local_meta:
+                    meta[item.filename] = local_meta
+
+        shutil.rmtree(temp_folder)
+        return meta
+
     def remove_all(self) -> bool:
         # pylint: disable=too-many-branches
 
@@ -113,7 +138,13 @@ class ArchiveBasedAbstractParser(abstract.AbstractParser):
                             abort = True
                             continue
                     if tmp_parser:
-                        tmp_parser.remove_all()
+                        if tmp_parser.remove_all() is False:
+                            logging.warning("In file %s, something went wrong \
+                                             with the cleaning of %s \
+                                             (format: %s)",
+                                            self.filename, item.filename, mtype)
+                            abort = True
+                            continue
                         os.rename(tmp_parser.output_filename, full_path)
 
                 zinfo = zipfile.ZipInfo(item.filename)  # type: ignore
@@ -126,3 +157,8 @@ class ArchiveBasedAbstractParser(abstract.AbstractParser):
             os.remove(self.output_filename)
             return False
         return True
+
+
+
+class ZipParser(ArchiveBasedAbstractParser):
+    mimetypes = {'application/zip'}

libmat2/exiftool.py

@@ -53,15 +53,14 @@ class ExiftoolParser(abstract.AbstractParser):
         return True
 
 def _get_exiftool_path() -> str:  # pragma: no cover
-    exiftool_path = '/usr/bin/exiftool'
-    if os.path.isfile(exiftool_path):
-        if os.access(exiftool_path, os.X_OK):
-            return exiftool_path
+    possible_pathes = {
+        '/usr/bin/exiftool',              # debian/fedora
+        '/usr/bin/vendor_perl/exiftool',  # archlinux
+    }
 
-    # ArchLinux
-    exiftool_path = '/usr/bin/vendor_perl/exiftool'
-    if os.path.isfile(exiftool_path):
-        if os.access(exiftool_path, os.X_OK):
-            return exiftool_path
+    for possible_path in possible_pathes:
+        if os.path.isfile(possible_path):
+            if os.access(possible_path, os.X_OK):
+                return possible_path
 
     raise RuntimeError("Unable to find exiftool")

libmat2/images.py

@@ -6,7 +6,7 @@ import cairo
 
 import gi
 gi.require_version('GdkPixbuf', '2.0')
-from gi.repository import GdkPixbuf
+from gi.repository import GdkPixbuf, GLib
 
 from . import exiftool
 
@@ -50,15 +50,21 @@ class GdkPixbufAbstractParser(exiftool.ExiftoolParser):
 
     def __init__(self, filename):
         super().__init__(filename)
-        if imghdr.what(filename) != self._type:  # better safe than sorry
+        # we can't use imghdr here because of https://bugs.python.org/issue28591
+        try:
+            GdkPixbuf.Pixbuf.new_from_file(self.filename)
+        except GLib.GError:
             raise ValueError
 
     def remove_all(self) -> bool:
+        if self.lightweight_cleaning:
+            return self._lightweight_cleanup()
+
         _, extension = os.path.splitext(self.filename)
         pixbuf = GdkPixbuf.Pixbuf.new_from_file(self.filename)
         if extension.lower() == '.jpg':
             extension = '.jpeg'  # gdk is picky
-        pixbuf.savev(self.output_filename, extension[1:], [], [])
+        pixbuf.savev(self.output_filename, type=extension[1:], option_keys=[], option_values=[])
         return True
 
 

libmat2/office.py

@@ -301,7 +301,7 @@ class MSOfficeParser(ArchiveBasedAbstractParser):
         Yes, I know that parsing xml with regexp ain't pretty,
         be my guest and fix it if you want.
         """
-        metadata = {}
+        metadata = super().get_meta()
         zipin = zipfile.ZipFile(self.filename)
         for item in zipin.infolist():
             if item.filename.startswith('docProps/') and item.filename.endswith('.xml'):

libmat2/video.py

@@ -2,10 +2,37 @@ import os
 import subprocess
 import logging
 
+from typing import Dict, Union
+
 from . import exiftool
 
 
-class AVIParser(exiftool.ExiftoolParser):
+class AbstractFFmpegParser(exiftool.ExiftoolParser):
+    """ Abstract parser for all FFmpeg-based ones, mainly for video. """
+    def remove_all(self) -> bool:
+        cmd = [_get_ffmpeg_path(),
+               '-i', self.filename,      # input file
+               '-y',                     # overwrite existing output file
+               '-map', '0',              # copy everything all streams from input to output
+               '-codec', 'copy',         # don't decode anything, just copy (speed!)
+               '-loglevel', 'panic',     # Don't show log
+               '-hide_banner',           # hide the banner
+               '-map_metadata', '-1',    # remove supperficial metadata
+               '-map_chapters', '-1',    # remove chapters
+               '-disposition', '0',      # Remove dispositions (check ffmpeg's manpage)
+               '-fflags', '+bitexact',   # don't add any metadata
+               '-flags:v', '+bitexact',  # don't add any metadata
+               '-flags:a', '+bitexact',  # don't add any metadata
+               self.output_filename]
+        try:
+            subprocess.check_call(cmd)
+        except subprocess.CalledProcessError as e:
+            logging.error("Something went wrong during the processing of %s: %s", self.filename, e)
+            return False
+        return True
+
+
+class AVIParser(AbstractFFmpegParser):
     mimetypes = {'video/x-msvideo', }
     meta_whitelist = {'SourceFile', 'ExifToolVersion', 'FileName', 'Directory',
                       'FileSize', 'FileModifyDate', 'FileAccessDate',
@@ -24,25 +51,55 @@ class AVIParser(exiftool.ExiftoolParser):
                       'SampleRate', 'AvgBytesPerSec', 'BitsPerSample',
                       'Duration', 'ImageSize', 'Megapixels'}
 
+class MP4Parser(AbstractFFmpegParser):
+    mimetypes = {'video/mp4', }
+    meta_whitelist = {'AudioFormat', 'AvgBitrate', 'Balance', 'TrackDuration',
+                      'XResolution', 'YResolution', 'ExifToolVersion',
+                      'FileAccessDate', 'FileInodeChangeDate', 'FileModifyDate',
+                      'FileName', 'FilePermissions', 'MIMEType', 'FileType',
+                      'FileTypeExtension', 'Directory', 'ImageWidth',
+                      'ImageSize', 'ImageHeight', 'FileSize', 'SourceFile',
+                      'BitDepth', 'Duration', 'AudioChannels',
+                      'AudioBitsPerSample', 'AudioSampleRate', 'Megapixels',
+                      'MovieDataSize', 'VideoFrameRate', 'MediaTimeScale',
+                      'SourceImageHeight', 'SourceImageWidth',
+                      'MatrixStructure', 'MediaDuration'}
+    meta_key_value_whitelist = {  # some metadata are mandatory :/
+        'CreateDate': '0000:00:00 00:00:00',
+        'CurrentTime': '0 s',
+        'MediaCreateDate': '0000:00:00 00:00:00',
+        'MediaLanguageCode': 'und',
+        'MediaModifyDate': '0000:00:00 00:00:00',
+        'ModifyDate': '0000:00:00 00:00:00',
+        'OpColor': '0 0 0',
+        'PosterTime': '0 s',
+        'PreferredRate': '1',
+        'PreferredVolume': '100.00%',
+        'PreviewDuration': '0 s',
+        'PreviewTime': '0 s',
+        'SelectionDuration': '0 s',
+        'SelectionTime': '0 s',
+        'TrackCreateDate': '0000:00:00 00:00:00',
+        'TrackModifyDate': '0000:00:00 00:00:00',
+        'TrackVolume': '0.00%',
+    }
+
     def remove_all(self) -> bool:
-        cmd = [_get_ffmpeg_path(),
-               '-i', self.filename,      # input file
-               '-y',                     # overwrite existing output file
-               '-loglevel', 'panic',     # Don't show log
-               '-hide_banner',           # hide the banner
-               '-codec', 'copy',         # don't decode anything, just copy (speed!)
-               '-map_metadata', '-1',    # remove supperficial metadata
-               '-map_chapters', '-1',    # remove chapters
-               '-fflags', '+bitexact',   # don't add any metadata
-               '-flags:v', '+bitexact',  # don't add any metadata
-               '-flags:a', '+bitexact',  # don't add any metadata
-               self.output_filename]
-        try:
-            subprocess.check_call(cmd)
-        except subprocess.CalledProcessError as e:
-            logging.error("Something went wrong during the processing of %s: %s", self.filename, e)
-            return False
-        return True
+        logging.warning('The format of "%s" (video/mp4) has some mandatory '
+                        'metadata fields; mat2 filled them with standard data.',
+                        self.filename)
+        return super().remove_all()
+
+    def get_meta(self) -> Dict[str, Union[str, dict]]:
+        meta = super().get_meta()
+
+        ret = dict()  # type: Dict[str, Union[str, dict]]
+        for key, value in meta.items():
+            if key in self.meta_key_value_whitelist.keys():
+                if value == self.meta_key_value_whitelist[key]:
+                    continue
+            ret[key] = value
+        return ret
 
 
 def _get_ffmpeg_path() -> str:  # pragma: no cover

mat2

@@ -6,6 +6,7 @@ import sys
 import mimetypes
 import argparse
 import logging
+import unicodedata
 
 try:
     from libmat2 import parser_factory, UNSUPPORTED_EXTENSIONS
@@ -14,12 +15,14 @@ except ValueError as e:
     print(e)
     sys.exit(1)
 
-__version__ = '0.5.0'
+__version__ = '0.6.0'
 
 # Make pyflakes happy
 assert Tuple
 assert Union
 
+logging.basicConfig(format='%(levelname)s: %(message)s', level=logging.WARNING)
+
 
 def __check_file(filename: str, mode: int=os.R_OK) -> bool:
     if not os.path.exists(filename):
@@ -81,6 +84,15 @@ def __print_meta(filename: str, metadata: dict, depth: int=1):
         if isinstance(v, dict):
             __print_meta(k, v, depth+1)
             continue
+
+        # Remove control characters
+        # We might use 'Cc' instead of 'C', but better safe than sorry
+        # https://www.unicode.org/reports/tr44/#GC_Values_Table
+        try:
+            v = ''.join(ch for ch in v if not unicodedata.category(ch).startswith('C'))
+        except TypeError:
+            pass  # for things that aren't iterable
+
         try:  # FIXME this is ugly.
             print(padding + "  %s: %s" % (k, v))
         except UnicodeEncodeError:

nautilus/mat2.py

@@ -14,7 +14,7 @@ thread, so we'll have to resort to using a `queue` to pass "messages" around.
 
 import queue
 import threading
-from typing import Tuple
+from typing import Tuple, Optional, List
 from urllib.parse import unquote
 
 import gi
@@ -25,10 +25,8 @@ from gi.repository import Nautilus, GObject, Gtk, Gio, GLib, GdkPixbuf
 
 from libmat2 import parser_factory
 
-# make pyflakes happy
-assert Tuple
 
-def _remove_metadata(fpath):
+def _remove_metadata(fpath) -> Tuple[bool, Optional[str]]:
     """ This is a simple wrapper around libmat2, because it's
     easier and cleaner this way.
     """
@@ -63,7 +61,7 @@ class ColumnExtension(GObject.GObject, Nautilus.MenuProvider, Nautilus.LocationW
         self.infobar.get_content_area().pack_start(self.infobar_hbox, True, True, 0)
         self.infobar.show_all()
 
-    def get_widget(self, uri, window):
+    def get_widget(self, uri, window) -> Gtk.Widget:
         """ This is the method that we have to implement (because we're
         a LocationWidgetProvider) in order to show our infobar.
         """
@@ -228,7 +226,7 @@ class ColumnExtension(GObject.GObject, Nautilus.MenuProvider, Nautilus.LocationW
         """ https://bugzilla.gnome.org/show_bug.cgi?id=784278 """
         return None
 
-    def get_file_items(self, window, files):
+    def get_file_items(self, window, files) -> Optional[List[Nautilus.MenuItem]]:
         """ This method is the one allowing us to create a menu item.
         """
         # Do not show the menu item if not a single file has a chance to be

setup.py

@@ -5,7 +5,7 @@ with open("README.md", "r") as fh:
 
 setuptools.setup(
     name="mat2",
-    version='0.5.0',
+    version='0.6.0',
     author="Julien (jvoisin) Voisin",
     author_email="julien.voisin+mat2@dustri.org",
     description="A handy tool to trash your metadata",

tests/test_climat2.py

@@ -121,7 +121,7 @@ class TestGetMeta(unittest.TestCase):
         proc = subprocess.Popen(mat2_binary + ['--show', './tests/data/dirty.pdf'],
                 stdout=subprocess.PIPE)
         stdout, _ = proc.communicate()
-        self.assertIn(b'producer: pdfTeX-1.40.14', stdout)
+        self.assertIn(b'Producer: pdfTeX-1.40.14', stdout)
 
     def test_png(self):
         proc = subprocess.Popen(mat2_binary + ['--show', './tests/data/dirty.png'],
@@ -174,3 +174,10 @@ class TestGetMeta(unittest.TestCase):
         self.assertIn(b'genre: Python', stdout)
         self.assertIn(b'i am a : various comment', stdout)
         self.assertIn(b'artist: jvoisin', stdout)
+
+class TestControlCharInjection(unittest.TestCase):
+    def test_jpg(self):
+        proc = subprocess.Popen(mat2_binary + ['--show', './tests/data/control_chars.jpg'],
+                stdout=subprocess.PIPE)
+        stdout, _ = proc.communicate()
+        self.assertIn(b'Comment: GQ\n', stdout)

tests/test_corrupted_files.py

@@ -4,6 +4,7 @@ import unittest
 import shutil
 import os
 import logging
+import zipfile
 
 from libmat2 import pdf, images, audio, office, parser_factory, torrent
 from libmat2 import harmless, video
@@ -222,3 +223,17 @@ class TestCorruptedFiles(unittest.TestCase):
         p = video.AVIParser('./tests/data/--output.avi')
         self.assertFalse(p.remove_all())
         os.remove('./tests/data/--output.avi')
+
+    def test_zip(self):
+        with zipfile.ZipFile('./tests/data/dirty.zip', 'w') as zout:
+            zout.write('./tests/data/dirty.flac')
+            zout.write('./tests/data/dirty.docx')
+            zout.write('./tests/data/dirty.jpg')
+            zout.write('./tests/data/embedded_corrupted.docx')
+        p, mimetype = parser_factory.get_parser('./tests/data/dirty.zip')
+        self.assertEqual(mimetype, 'application/zip')
+        meta = p.get_meta()
+        self.assertEqual(meta['tests/data/dirty.flac']['comments'], 'Thank you for using MAT !')
+        self.assertEqual(meta['tests/data/dirty.docx']['word/media/image1.png']['Comment'], 'This is a comment, be careful!')
+        self.assertFalse(p.remove_all())
+        os.remove('./tests/data/dirty.zip')