summaryrefslogtreecommitdiff
path: root/ext
diff options
context:
space:
mode:
authorCharles Oliver Nutter <headius@headius.com>2017-09-07 16:25:54 -0500
committerGitHub <noreply@github.com>2017-09-07 16:25:54 -0500
commitb572261900893df0352d3ad71c17dfa81ab17ccc (patch)
tree3bcf8b863240a2dfb2023242fe2a35d71ba4add2 /ext
parent4039a811248de7b0b9ae9e4a97854f00ba985255 (diff)
parent2bbb245edaf80179f4673bf4bf75978a10654c64 (diff)
downloadpsych-b572261900893df0352d3ad71c17dfa81ab17ccc.zip
Merge pull request #328 from enebo/desensitize_io_encoding_on_load
Parse.load with an IO/File which has a non-YAML allowed encoding will…
Diffstat (limited to 'ext')
-rw-r--r--ext/java/PsychParser.java7
1 files changed, 7 insertions, 0 deletions
diff --git a/ext/java/PsychParser.java b/ext/java/PsychParser.java
index b3e747e..f5b6faf 100644
--- a/ext/java/PsychParser.java
+++ b/ext/java/PsychParser.java
@@ -33,6 +33,8 @@ import java.nio.charset.Charset;
import java.util.Map;
import org.jcodings.Encoding;
+import org.jcodings.specific.UTF16BEEncoding;
+import org.jcodings.specific.UTF16LEEncoding;
import org.jcodings.specific.UTF8Encoding;
import org.jcodings.unicode.UnicodeEncoding;
import org.jruby.Ruby;
@@ -162,6 +164,11 @@ public class PsychParser extends RubyObject {
if (yaml instanceof RubyIO) {
Encoding enc = ((RubyIO) yaml).getReadEncoding();
charset = enc.getCharset();
+
+ // libyaml treats non-utf encodings as utf-8 and hopes for the best.
+ if (!(enc instanceof UTF8Encoding) && !(enc instanceof UTF16LEEncoding) && !(enc instanceof UTF16BEEncoding)) {
+ charset = UTF8Encoding.INSTANCE.getCharset();
+ }
}
if (charset == null) {
// If we can't get it from the IO or it doesn't have a charset, fall back on UTF-8