trytond: Add binary scanner [PREVIEW]

trytond/doc/topics/configuration.rst

@@ -304,6 +304,24 @@ The name of the similarity function.
 
 Default: ``similarity``
 
+binary_scanner
+~~~~~~~~~~~~~~
+
+The command used to scan the content of Binary fields.
+The command receive the quarantine directory as last argument.
+If the command returns a non zero status, it is considered that at least one
+file in the directory is malicious.
+
+binary_scanner_directory
+~~~~~~~~~~~~~~~~~~~~~~~~
+
+The directory where the binary content are stored before being checked
+by the scanner.
+
+Default: The temporary directory as determined by Python's tempfile_ module.
+
+.. _tempfile: https://docs.python.org/3/library/tempfile.html#tempfile.gettempdir
+
 .. _config-request:
 
 request

trytond/trytond/ir/message.xml

@@ -361,6 +361,9 @@ this repository contains the full copyright notices and license terms. -->
         <record model="ir.message" id="msg_trigger_invalid_condition">
             <field name="text">Condition "%(condition)s" is not a valid PYSON expression for trigger "%(trigger)s".</field>
         </record>
+        <record model="ir.message" id="msg_malicious_binary">
+            <field name="text">The content of field "%(field)s" has been detected as malicious.</field>
+        </record>
         <record model="ir.message" id="msg_html_editor_save_fail">
             <field name="text">Failed to save, please retry.</field>
         </record>

trytond/trytond/model/fields/binary.py

@@ -1,13 +1,22 @@
 # This file is part of Tryton.  The COPYRIGHT file at the top level of
 # this repository contains the full copyright notices and license terms.
+import logging
+import shlex
+import subprocess
+import tempfile
+
 from sql import Column, Null
 
+from trytond.config import config
 from trytond.filestore import filestore
+from trytond.i18n import gettext
 from trytond.tools import cached_property, grouped_slice, reduce_ids
 from trytond.transaction import Transaction
 
 from .field import Field
 
+logger = logging.getLogger(__name__)
+
 
 def caster(d):
     if isinstance(d, bytes):
@@ -17,6 +26,32 @@ def caster(d):
     return bytes(d, encoding='utf8')
 
 
+def check_content(field_name, *binaries):
+    from trytond.model.modelstorage import BinaryScanError
+
+    scanner = config.get('database', 'binary_scanner')
+    scanner_dir = config.get(
+        'database', 'binary_scanner_directory', default=tempfile.gettempdir())
+    if not scanner:
+        return
+
+    with tempfile.TemporaryDirectory(dir=scanner_dir) as tempdir:
+        for binary in binaries:
+            with tempfile.NamedTemporaryFile(dir=tempdir, delete=False) as fd:
+                fd.write(binary)
+
+        try:
+            subprocess.check_call(
+                shlex.split(scanner) + [tempdir],
+                stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+        except subprocess.CalledProcessError as error:
+            logger.critical(
+                "'%s %s' exited with code '%s'",
+                scanner, tempdir, error.returncode)
+            raise BinaryScanError(gettext(
+                    'ir.msg_malicious_binary', field=field_name))
+
+
 class Binary(Field):
     _type = 'binary'
     _sql_type = 'BLOB'
@@ -112,6 +147,9 @@ def set(self, Model, name, ids, value, *args):
         if prefix is None:
             prefix = transaction.database.name
 
+        self._check_contents(
+            Model.__names__(name)['field'], *((ids, value) + args)[1::2])
+
         args = iter((ids, value) + args)
         for ids, value in zip(args, args):
             if self.file_id:
@@ -124,6 +162,10 @@ def set(self, Model, name, ids, value, *args):
             cursor.execute(*table.update(columns, values,
                     where=reduce_ids(table.id, ids)))
 
+    @classmethod
+    def _check_contents(cls, field, *values):
+        check_content(field, *values)
+
     def definition(self, model, language):
         definition = super().definition(model, language)
         definition['searchable'] = False

trytond/trytond/model/modelstorage.py

@@ -98,6 +98,10 @@ def __init__(self, record):
         self.record = record
 
 
+class BinaryScanError(ValidationError):
+    pass
+
+
 def is_leaf(expression):
     return (isinstance(expression, (list, tuple))
         and len(expression) > 2