Software /
code /
prosody
File
spec/net_http_parser_spec.lua @ 12180:53e0ae770917
util.xml: Do not allow doctypes, comments or processing instructions
Yes. This is as bad as it sounds. CVE pending.
In Prosody itself, this only affects mod_websocket, which uses util.xml
to parse the <open/> frame, thus allowing unauthenticated remote DoS
using Billion Laughs. However, third-party modules using util.xml may
also be affected by this.
This commit installs handlers which disallow the use of doctype
declarations and processing instructions without any escape hatch. It,
by default, also introduces such a handler for comments, however, there
is a way to enable comments nontheless.
This is because util.xml is used to parse human-facing data, where
comments are generally a desirable feature, and also because comments
are generally harmless.
author | Jonas Schäfer <jonas@wielicki.name> |
---|---|
date | Mon, 10 Jan 2022 18:23:54 +0100 |
parent | 11033:cb5555443852 |
child | 12882:9ed628635dc6 |
line wrap: on
line source
local http_parser = require "net.http.parser"; local sha1 = require "util.hashes".sha1; local parser_input_bytes = 3; local function CRLF(s) return (s:gsub("\n", "\r\n")); end local function test_stream(stream, expect) local success_cb = spy.new(function (packet) assert.is_table(packet); if packet.body ~= false then assert.is_equal(expect.body, packet.body); end end); local parser = http_parser.new(success_cb, error, stream:sub(1,4) == "HTTP" and "client" or "server") for chunk in stream:gmatch("."..string.rep(".?", parser_input_bytes-1)) do parser:feed(chunk); end assert.spy(success_cb).was_called(expect.count or 1); end describe("net.http.parser", function() describe("parser", function() it("should handle requests with no content-length or body", function () test_stream( CRLF[[ GET / HTTP/1.1 Host: example.com ]], { body = ""; } ); end); it("should handle responses with empty body", function () test_stream( CRLF[[ HTTP/1.1 200 OK Content-Length: 0 ]], { body = ""; } ); end); it("should handle simple responses", function () test_stream( CRLF[[ HTTP/1.1 200 OK Content-Length: 7 Hello ]], { body = "Hello\r\n", count = 1; } ); end); it("should handle chunked encoding in responses", function () test_stream( CRLF[[ HTTP/1.1 200 OK Transfer-Encoding: chunked 1 H 1 e 2 ll 1 o 0 ]], { body = "Hello", count = 2; } ); end); it("should handle a stream of responses", function () test_stream( CRLF[[ HTTP/1.1 200 OK Content-Length: 5 Hello HTTP/1.1 200 OK Transfer-Encoding: chunked 1 H 1 e 2 ll 1 o 0 ]], { body = "Hello", count = 3; } ); end); end); it("should handle large chunked responses", function () local data = io.open("spec/inputs/http/httpstream-chunked-test.txt", "rb"):read("*a"); -- Just a sanity check... text editors and things may mess with line endings, etc. assert.equal("25930f021785ae14053a322c2dbc1897c3769720", sha1(data, true), "test data malformed"); test_stream(data, { body = string.rep("~", 11085), count = 2; }); end); end);