From b9a62d798af14ea799ae5fceab1ed7a537d1cbdd Mon Sep 17 00:00:00 2001
From: jvoisin <julien.voisin@dustri.org>
Date: Sun, 3 Feb 2019 22:55:15 +0100
Subject: [PATCH] Refactor a bit office get_meta handling

This should make easier to get more metadata from
archive-based file formats.
---
 libmat2/archive.py    | 24 +++++++++++++++++++-----
 libmat2/office.py     | 29 ++++++++++++-----------------
 tests/test_libmat2.py |  6 +++---
 3 files changed, 34 insertions(+), 25 deletions(-)

diff --git a/libmat2/archive.py b/libmat2/archive.py
index b2483fc..d155664 100644
--- a/libmat2/archive.py
+++ b/libmat2/archive.py
@@ -4,7 +4,7 @@ import tempfile
 import os
 import logging
 import shutil
-from typing import Dict, Set, Pattern, Union
+from typing import Dict, Set, Pattern, Union, Any
 
 from . import abstract, UnknownMemberPolicy, parser_factory
 
@@ -42,6 +42,12 @@ class ArchiveBasedAbstractParser(abstract.AbstractParser):
         # pylint: disable=unused-argument,no-self-use
         return True  # pragma: no cover
 
+    def _specific_get_meta(self, full_path: str, file_path: str) -> Dict[str, Any]:
+        """ This method can be used to extract specific metadata
+        from files present in the archive."""
+        # pylint: disable=unused-argument,no-self-use
+        return {}  # pragma: no cover
+
     @staticmethod
     def _clean_zipinfo(zipinfo: zipfile.ZipInfo) -> zipfile.ZipInfo:
         zipinfo.create_system = 3  # Linux
@@ -74,6 +80,10 @@ class ArchiveBasedAbstractParser(abstract.AbstractParser):
             temp_folder = tempfile.mkdtemp()
 
             for item in zin.infolist():
+                local_meta = dict()  # type: Dict[str, Union[str, Dict]]
+                for k, v in self._get_zipinfo_meta(item).items():
+                    local_meta[k] = v
+
                 if item.filename[-1] == '/':  # pragma: no cover
                     # `is_dir` is added in Python3.6
                     continue  # don't keep empty folders
@@ -81,11 +91,15 @@ class ArchiveBasedAbstractParser(abstract.AbstractParser):
                 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
+                specific_meta = self._specific_get_meta(full_path, item.filename)
+                for (k, v) in specific_meta.items():
+                    local_meta[k] = v
+
+                tmp_parser, _ = parser_factory.get_parser(full_path)  # type: ignore
+                if tmp_parser:
+                    for k, v in tmp_parser.get_meta().items():
+                        local_meta[k] = v
 
-                local_meta = tmp_parser.get_meta()
                 if local_meta:
                     meta[item.filename] = local_meta
 
diff --git a/libmat2/office.py b/libmat2/office.py
index 365c230..dfad3b3 100644
--- a/libmat2/office.py
+++ b/libmat2/office.py
@@ -2,7 +2,7 @@ import logging
 import os
 import re
 import zipfile
-from typing import Dict, Set, Pattern, Tuple, Union
+from typing import Dict, Set, Pattern, Tuple, Union, Any
 
 import xml.etree.ElementTree as ET  # type: ignore
 
@@ -295,26 +295,21 @@ class MSOfficeParser(ArchiveBasedAbstractParser):
 
         return True
 
-    def get_meta(self) -> Dict[str, Union[str, dict]]:
+    def _specific_get_meta(self, full_path: str, file_path: str) -> Dict[str, Any]:
         """
         Yes, I know that parsing xml with regexp ain't pretty,
         be my guest and fix it if you want.
         """
-        metadata = super().get_meta()
-        zipin = zipfile.ZipFile(self.filename)
-        for item in zipin.infolist():
-            if item.filename.startswith('docProps/') and item.filename.endswith('.xml'):
-                try:
-                    content = zipin.read(item).decode('utf-8')
-                    results = re.findall(r"<(.+)>(.+)</\1>", content, re.I|re.M)
-                    for (key, value) in results:
-                        metadata[key] = value
-                except (TypeError, UnicodeDecodeError):  # We didn't manage to parse the xml file
-                    metadata[item.filename] = 'harmful content'
-            for key, value in self._get_zipinfo_meta(item).items():
-                metadata[key] = value
-        zipin.close()
-        return metadata
+        if not file_path.startswith('docProps/') or not file_path.endswith('.xml'):
+            return {}
+
+        with open(full_path, encoding='utf-8') as f:
+            try:
+                results = re.findall(r"<(.+)>(.+)</\1>", f.read(), re.I|re.M)
+                return {k:v for (k, v) in results}
+            except (TypeError, UnicodeDecodeError):
+                # We didn't manage to parse the xml file
+                return {file_path: 'harmful content', }
 
 
 class LibreOfficeParser(ArchiveBasedAbstractParser):
diff --git a/tests/test_libmat2.py b/tests/test_libmat2.py
index 9354286..d692181 100644
--- a/tests/test_libmat2.py
+++ b/tests/test_libmat2.py
@@ -131,9 +131,9 @@ class TestGetMeta(unittest.TestCase):
     def test_docx(self):
         p = office.MSOfficeParser('./tests/data/dirty.docx')
         meta = p.get_meta()
-        self.assertEqual(meta['cp:lastModifiedBy'], 'Julien Voisin')
-        self.assertEqual(meta['dc:creator'], 'julien voisin')
-        self.assertEqual(meta['Application'], 'LibreOffice/5.4.5.1$Linux_X86_64 LibreOffice_project/40m0$Build-1')
+        self.assertEqual(meta['docProps/core.xml']['cp:lastModifiedBy'], 'Julien Voisin')
+        self.assertEqual(meta['docProps/core.xml']['dc:creator'], 'julien voisin')
+        self.assertEqual(meta['docProps/app.xml']['Application'], 'LibreOffice/5.4.5.1$Linux_X86_64 LibreOffice_project/40m0$Build-1')
 
     def test_libreoffice(self):
         p = office.LibreOfficeParser('./tests/data/dirty.odt')