diff --git a/requirements-test.txt b/requirements-test.txt
--- a/requirements-test.txt
+++ b/requirements-test.txt
@@ -12,6 +12,7 @@
 swh.loader.git >= 0.8.0
 swh-scheduler[testing] >= 0.5.0
 swh.storage >= 0.1.1
+types-chardet
 types-docutils
 types-psycopg2
 types-pyyaml
diff --git a/requirements.txt b/requirements.txt
--- a/requirements.txt
+++ b/requirements.txt
@@ -3,6 +3,7 @@
 # dependency lines, see https://pip.readthedocs.org/en/1.1/requirements.html
 
 beautifulsoup4
+chardet
 cryptography
 django < 3
 django-cors-headers
diff --git a/swh/web/browse/utils.py b/swh/web/browse/utils.py
--- a/swh/web/browse/utils.py
+++ b/swh/web/browse/utils.py
@@ -6,7 +6,9 @@
 import base64
 import stat
 import textwrap
+from typing import Tuple
 
+import chardet
 import magic
 import sentry_sdk
 
@@ -91,12 +93,30 @@
 content_display_max_size = get_config()["content_display_max_size"]
 
 
-def _re_encode_content(mimetype, encoding, content_data):
-    # encode textual content to utf-8 if needed
-    if mimetype.startswith("text/"):
-        # probably a malformed UTF-8 content, re-encode it
-        # by replacing invalid chars with a substitution one
-        if encoding == "unknown-8bit":
+def re_encode_content(
+    mimetype: str, encoding: str, content_data: bytes
+) -> Tuple[str, str, bytes]:
+    """Try to re-encode textual content if it is not encoded to UTF-8
+    for proper display in the browse Web UI.
+
+    Args:
+        mimetype: content mimetype as detected by python-magic
+        encoding: content encoding as detected by python-magic
+        content_data: raw content bytes
+
+    Returns:
+        A tuple with 3 members: content mimetype, content encoding (possibly updated
+        after processing), content raw bytes (possibly reencoded to UTF-8)
+    """
+    if mimetype.startswith("text/") and encoding not in ("us-ascii", "utf-8"):
+        # first check if chardet detects an encoding with confidence
+        result = chardet.detect(content_data)
+        if result["confidence"] >= 0.9:
+            encoding = result["encoding"]
+            content_data = content_data.decode(encoding).encode("utf-8")
+        elif encoding == "unknown-8bit":
+            # probably a malformed UTF-8 content, re-encode it
+            # by replacing invalid chars with a substitution one
             content_data = content_data.decode("utf-8", "replace").encode("utf-8")
         elif encoding not in ["utf-8", "binary"]:
             content_data = content_data.decode(encoding, "replace").encode("utf-8")
@@ -176,7 +196,7 @@
                 )
 
             if re_encode:
-                mimetype, encoding, raw_data = _re_encode_content(
+                mimetype, encoding, raw_data = re_encode_content(
                     mimetype, encoding, content_data["raw_data"]
                 )
                 content_data["raw_data"] = raw_data
diff --git a/swh/web/tests/browse/test_utils.py b/swh/web/tests/browse/test_utils.py
--- a/swh/web/tests/browse/test_utils.py
+++ b/swh/web/tests/browse/test_utils.py
@@ -11,6 +11,7 @@
     gen_revision_link,
     get_mimetype_and_encoding_for_content,
     prepare_content_for_display,
+    re_encode_content,
 )
 from swh.web.common.utils import reverse
 
@@ -87,3 +88,14 @@
         content_data=b"", mime_type="", path=path
     )
     assert content_display["language"] == expected_language
+
+
+def test_re_encode_content_for_shift_jis_encoding():
+    data = b"/* \x8a\xd6\x98A\x82\xcc\x95\xb6\x8e\x9a\x83R\x81[\x83h\x95\xcf\x8a\xb7 */"
+    mime_type, encoding = get_mimetype_and_encoding_for_content(data)
+
+    _, encoding, re_encoded_data = re_encode_content(mime_type, encoding, data)
+
+    assert encoding == "SHIFT_JIS"
+    assert data.decode(encoding) == re_encoded_data.decode("utf-8")
+    assert re_encoded_data.decode("utf-8") == "/* 関連の文字コード変換 */"
diff --git a/swh/web/tests/browse/views/test_content.py b/swh/web/tests/browse/views/test_content.py
--- a/swh/web/tests/browse/views/test_content.py
+++ b/swh/web/tests/browse/views/test_content.py
@@ -16,9 +16,9 @@
 from swh.model.swhids import ObjectType
 from swh.web.browse.snapshot_context import process_snapshot_branches
 from swh.web.browse.utils import (
-    _re_encode_content,
     get_mimetype_and_encoding_for_content,
     prepare_content_for_display,
+    re_encode_content,
 )
 from swh.web.common.exc import NotFoundExc
 from swh.web.common.identifiers import gen_swhid
@@ -632,7 +632,7 @@
 
     mime_type, encoding = get_mimetype_and_encoding_for_content(content_data["data"])
 
-    mime_type, encoding, content_data = _re_encode_content(
+    mime_type, encoding, content_data = re_encode_content(
         mime_type, encoding, content_data["data"]
     )
 
diff --git a/swh/web/tests/data.py b/swh/web/tests/data.py
--- a/swh/web/tests/data.py
+++ b/swh/web/tests/data.py
@@ -37,9 +37,9 @@
 from swh.storage.utils import now
 from swh.web import config
 from swh.web.browse.utils import (
-    _re_encode_content,
     get_mimetype_and_encoding_for_content,
     prepare_content_for_display,
+    re_encode_content,
 )
 from swh.web.common import archive
 
@@ -407,7 +407,7 @@
         cnt_data = storage.content_get_data(content.sha1)
         assert cnt_data is not None
         mimetype, encoding = get_mimetype_and_encoding_for_content(cnt_data)
-        _, _, cnt_data = _re_encode_content(mimetype, encoding, cnt_data)
+        _, _, cnt_data = re_encode_content(mimetype, encoding, cnt_data)
 
         content_display_data = prepare_content_for_display(cnt_data, mimetype, path)