From fa1ef30dcdac83b271ce38c71975df0ed96b08f7 Mon Sep 17 00:00:00 2001
From: Kevin Pullin <kevin.pullin@gmail.com>
Date: Sun, 3 Mar 2013 18:10:28 -0800
Subject: [PATCH] Strip the UTF8 BOM from a UTF encoded response

Some frameworks, such as .NET, often emit the UTF8 byte-order-mark in
the HTTP response.

This commit strips out the BOM when the encoding is set to 'utf8' and
improves compatibility with external services.

(This is a copy of a previous commit for issue #371, which no longer
cleanly applies and requires a refresh).
---
 index.js           | 5 +++++
 tests/test-body.js | 5 +++++
 2 files changed, 10 insertions(+)

diff --git a/index.js b/index.js
index 36b48a708..88a582a63 100755
--- a/index.js
+++ b/index.js
@@ -764,6 +764,11 @@ Request.prototype.start = function () {
               response.body = body.toString(self.encoding)
             }
           } else if (buffer.length) {
+            // The UTF8 BOM [0xEF,0xBB,0xBF] is converted to [0xFE,0xFF] in the JS UTC16/UCS2 representation.
+            // Strip this value out when the encoding is set to 'utf8', as upstream consumers won't expect it and it breaks JSON.parse().
+            if (self.encoding === 'utf8' && buffer[0].length > 0 && buffer[0][0] === "\uFEFF") {
+              buffer[0] = buffer[0].substring(1)
+            }
             response.body = buffer.join('')
           }
 
diff --git a/tests/test-body.js b/tests/test-body.js
index baac5b88a..186de123f 100644
--- a/tests/test-body.js
+++ b/tests/test-body.js
@@ -35,6 +35,11 @@ var tests =
     , encoding: 'hex'
     , expectBody: "efa3bfcea9e29883"
     }
+  , testGetUTF8:
+     { resp: server.createGetResponse(new Buffer([0xEF, 0xBB, 0xBF, 226, 152, 131]))
+     , encoding: "utf8"
+     , expectBody: "☃"
+     }
   , testGetJSON :
      { resp : server.createGetResponse('{"test":true}', 'application/json')
      , json : true