Fix EncodingHelper#clean blowing up on UTF-16BE strings

Closes gitaly#1101
This commit is contained in:
Ahmad Sherif 2018-03-22 21:22:20 +01:00
parent 5c36e1b96a
commit 1f2cc29c8f
2 changed files with 6 additions and 1 deletions

View file

@ -90,7 +90,7 @@ module Gitlab
end
def clean(message)
message.encode("UTF-16BE", undef: :replace, invalid: :replace, replace: "")
message.encode("UTF-16BE", undef: :replace, invalid: :replace, replace: "".encode("UTF-16BE"))
.encode("UTF-8")
.gsub("\0".encode("UTF-8"), "")
end

View file

@ -161,6 +161,11 @@ describe Gitlab::EncodingHelper do
'removes invalid bytes from ASCII-8bit encoded multibyte string.',
"Lorem ipsum\xC3\n dolor sit amet, xy\xC3\xA0y\xC3\xB9abcd\xC3\xB9efg".force_encoding('ASCII-8BIT'),
"Lorem ipsum\n dolor sit amet, xyàyùabcdùefg"
],
[
'handles UTF-16BE encoded strings',
"\xFE\xFF\x00\x41".force_encoding('ASCII-8BIT'), # An "A" prepended with UTF-16 BOM
"\xEF\xBB\xBFA" # An "A" prepended with UTF-8 BOM
]
].each do |description, test_string, xpect|
it description do