Set charset to utf-8 when adding headers for certain text content types (#7044)

Fixes #7043
This commit is contained in:
The Stranjer 2020-03-17 09:29:09 -04:00 committed by GitHub
parent 7581d30e9f
commit 5e477c1deb
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 25 additions and 1 deletions

1
changelog.d/7044.bugfix Normal file
View File

@ -0,0 +1 @@
Fix a bug that renders UTF-8 text files incorrectly when loaded from media. Contributed by @TheStranjer.

View File

@ -30,6 +30,22 @@ from synapse.util.stringutils import is_ascii
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
# list all text content types that will have the charset default to UTF-8 when
# none is given
TEXT_CONTENT_TYPES = [
"text/css",
"text/csv",
"text/html",
"text/calendar",
"text/plain",
"text/javascript",
"application/json",
"application/ld+json",
"application/rtf",
"image/svg+xml",
"text/xml",
]
def parse_media_id(request): def parse_media_id(request):
try: try:
@ -96,7 +112,14 @@ def add_file_headers(request, media_type, file_size, upload_name):
def _quote(x): def _quote(x):
return urllib.parse.quote(x.encode("utf-8")) return urllib.parse.quote(x.encode("utf-8"))
request.setHeader(b"Content-Type", media_type.encode("UTF-8")) # Default to a UTF-8 charset for text content types.
# ex, uses UTF-8 for 'text/css' but not 'text/css; charset=UTF-16'
if media_type.lower() in TEXT_CONTENT_TYPES:
content_type = media_type + "; charset=UTF-8"
else:
content_type = media_type
request.setHeader(b"Content-Type", content_type.encode("UTF-8"))
if upload_name: if upload_name:
# RFC6266 section 4.1 [1] defines both `filename` and `filename*`. # RFC6266 section 4.1 [1] defines both `filename` and `filename*`.
# #