First working release.

Other files (not .docx) support is not guaranteed

modules/document/document_file.py

@@ -0,0 +1,30 @@
+import os.path
+import tempfile
+import zipfile
+import shutil
+
+
+class ExtractedDocument:
+    def __init__(self, path: str):
+        self.documentroot = None
+        self.app = None
+        self.core = None
+        if zipfile.is_zipfile(path):
+            self.documentroot = tempfile.mkdtemp()
+            zipfile.ZipFile(path).extractall(self.documentroot)
+            self.app = os.path.join(self.documentroot, 'docProps', 'app.xml')
+            self.core = os.path.join(self.documentroot, 'docProps', 'core.xml')
+
+    def pack(self, path):
+        with zipfile.ZipFile(path, "w", compresslevel=9, compression=zipfile.ZIP_DEFLATED) as z:
+            for root, dirs, files in os.walk(self.documentroot):
+                for file in files:
+                    z.write(os.path.join(root, file),
+                            os.path.relpath(os.path.join(root, file),
+                                            self.documentroot))
+
+    def remove(self):
+        try:
+            shutil.rmtree(self.documentroot, True)
+        except Exception as e:
+            print(f'Error while removing {self.documentroot}: {e}, remove it manually if you want')

modules/document/document_properties.py

@@ -0,0 +1,36 @@
+import os.path
+import lxml
+from lxml import etree
+from modules.document.document_file import ExtractedDocument
+from modules.helpers import xml
+from modules.helpers.convert import Int
+from datetime import datetime
+
+
+class DocumentProps:
+    def __init__(self, extracted_document: ExtractedDocument):
+        if not extracted_document.documentroot or not os.path.isdir(extracted_document.documentroot):
+            self.parsed_ = False
+            return
+
+        self.parsed_ = True
+
+        core_xml = etree.parse(extracted_document.core)
+        app_xml = etree.parse(extracted_document.app)
+
+        self.extracted_document = extracted_document
+        self.application = xml.get_value(app_xml, '//Application')
+        self.paragraphs = Int(xml.get_value(app_xml, '//Paragraphs'))
+        self.lines = Int(xml.get_value(app_xml, '//Lines'))
+        self.characters = Int(xml.get_value(app_xml, '//Characters'))
+        self.words = Int(xml.get_value(app_xml, '//Words'))
+        self.pages = Int(xml.get_value(app_xml, '//Pages'))
+        self.total_time = Int(xml.get_value(app_xml, '//TotalTime'))
+        self.template = xml.get_value(app_xml, '//Template')
+        self.modified = datetime.strptime(xml.get_value(core_xml, 'dcterms:modified'), '%Y-%m-%dT%H:%M:%SZ')
+        self.created = datetime.strptime(xml.get_value(core_xml, 'dcterms:created'), '%Y-%m-%dT%H:%M:%SZ')
+        self.revision = xml.get_value(core_xml, 'cp:revision')
+        self.last_modified_by = xml.get_value(core_xml, 'cp:lastModifiedBy')
+        self.creator = xml.get_value(core_xml, 'dc:creator')
+        self.core_xml = core_xml
+        self.app_xml = app_xml

modules/helpers/convert.py

@@ -0,0 +1,6 @@
+def Int(s: str) -> int | None:
+    try:
+        return int(s)
+    except Exception as e:
+        assert e
+        return 0

modules/helpers/xml.py

@@ -0,0 +1,22 @@
+namespaces = {
+    "cp": "http://schemas.openxmlformats.org/package/2006/metadata/core-properties",
+    "dc": "http://purl.org/dc/elements/1.1/",
+    "dcterms": "http://purl.org/dc/terms/",
+    "dcmitype": "http://purl.org/dc/dcmitype/",
+    "xsi": "http://www.w3.org/2001/XMLSchema-instance",
+    "vt": "http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes",
+    "": "http://schemas.openxmlformats.org/officeDocument/2006/extended-properties"
+}
+
+
+def get_value(parsed, name):
+    return str(parsed.findall(name, namespaces)[0].text) \
+        if parsed.findall(name, namespaces) else None
+
+
+def set_value(parsed, name, value):
+    if value:
+        parsed.findall(name, namespaces)[0].text = str(value)
+    else:
+        parsed.findall(name, namespaces)[0].getparent().remove(parsed.findall(name, namespaces)[0]) \
+            if parsed.findall(name, namespaces) else None

modules/time_tools/time_diff.py

@@ -0,0 +1,9 @@
+import datetime
+
+
+def compute_time_diff(date_first, time_first,
+                      date_second, time_second):
+    dt1 = datetime.datetime.combine(date_first, time_first)
+    dt2 = datetime.datetime.combine(date_second, time_second)
+
+    return int((max(dt1, dt2) - min(dt1, dt2)).total_seconds() / 60)