diff --git a/debian/control b/debian/control
index a339313..51abd71 100644
--- a/debian/control
+++ b/debian/control
@@ -1,24 +1,25 @@
 Source: swh-indexer
 Maintainer: Software Heritage developers <swh-devel@inria.fr>
 Section: python
 Priority: optional
 Build-Depends: debhelper (>= 9),
                dh-python,
                python3-all,
                python3-nose,
                python3-setuptools,
                python3-swh.core (>= 0.0.25~),
                python3-swh.storage (>= 0.0.64~),
                python3-swh.objstorage (>= 0.0.13~),
                python3-swh.scheduler (>= 0.0.9~),
+               python3-chardet (>= 2.3.0~),
                python3-click,
                exuberant-ctags,
                python3-pygments,
                python3-vcversioner
 Standards-Version: 3.9.6
 Homepage: https://forge.softwareheritage.org/diffusion/78/
 
 Package: python3-swh.indexer
 Architecture: all
 Depends: exuberant-ctags, ${misc:Depends}, ${python3:Depends}
 Description: Software Heritage Content Indexer
diff --git a/requirements.txt b/requirements.txt
index 30d3b92..5316e2b 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,11 +1,12 @@
 # Add here external Python modules dependencies, one per line. Module names
 # should match https://pypi.python.org/pypi names. For the full spec or
 # dependency lines, see https://pip.readthedocs.org/en/1.1/requirements.html
 vcversioner
 pygments
 click
+chardet
 
 swh.core >= 0.0.25
 swh.storage >= 0.0.64
 swh.objstorage >= 0.0.13
 swh.scheduler >= 0.0.9
diff --git a/swh/indexer/language.py b/swh/indexer/language.py
index 133d298..2a7855c 100644
--- a/swh/indexer/language.py
+++ b/swh/indexer/language.py
@@ -1,53 +1,48 @@
-# Copyright (C) 2015-2016  The Software Heritage developers
+# Copyright (C) 2016  The Software Heritage developers
 # See the AUTHORS file at the top-level directory of this distribution
 # License: GNU General Public License version 3, or any later version
 # See top-level LICENSE file for more information
 
 
-import json
-import os
-import math
-import sys
-
 from pygments.lexers import guess_lexer
 from pygments.util import ClassNotFound
+from chardet import detect
 
 
 def cleanup_classname(classname):
     """Determine the language from the pygments' lexer names.
 
     """
     return classname.lower().replace(' ', '-')
 
 
-def run_language(path, encoding=None):
-    """Determine mime-type from file at path.
+def run_language(raw_content):
+    """Determine the raw content's language.
 
     Args:
-        path (str): filepath to determine the mime type
-        encoding (str): optional file's encoding
+        raw_content (bytes): content to determine raw content
 
     Returns:
         Dict with keys:
         - lang: None if nothing found or the possible language
         - decoding_failure: True if a decoding failure happened
 
     """
     try:
-        with open(path, 'r', encoding=encoding) as f:
-            try:
-                raw_content = f.read()
-                lang = cleanup_classname(
-                    guess_lexer(raw_content).name)
-                return {
-                    'lang': lang
-                }
-            except ClassNotFound as e:
-                return {
-                    'lang': None
-                }
+        encoding = detect(raw_content)['encoding']
+        content = raw_content.decode(encoding)
+        lang = cleanup_classname(
+            guess_lexer(content).name)
+
+        return {
+            'lang': lang
+        }
+    except ClassNotFound as e:
+        return {
+            'lang': None
+        }
     except LookupError as e:  # Unknown encoding
         return {
             'decoding_failure': True,
             'lang': None
         }