From 28ddac2ec285b950a17bc42d7c79c329c8573c9e Mon Sep 17 00:00:00 2001
From: Anna Henningsen <anna@addaleax.net>
Date: Sun, 14 May 2017 19:57:52 +0200
Subject: [PATCH] buffer: fix indexOf for empty searches

Make searches for empty subsequences do exactly what
`String.prototype.indexOf()` does.

Fixes: https://github.com/nodejs/node/issues/13023
PR-URL: https://github.com/nodejs/node/pull/13024
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Refael Ackermann <refack@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
---
 doc/api/buffer.md                     |  6 ++++
 lib/buffer.js                         |  2 +-
 src/node_buffer.cc                    | 48 +++++++++++++++++++--------
 test/parallel/test-buffer-includes.js | 16 ++++-----
 test/parallel/test-buffer-indexof.js  | 24 +++++++-------
 5 files changed, 62 insertions(+), 34 deletions(-)

diff --git a/doc/api/buffer.md b/doc/api/buffer.md
index 3a6a048417..9b3f3f8b4b 100644
--- a/doc/api/buffer.md
+++ b/doc/api/buffer.md
@@ -1340,6 +1340,10 @@ console.log(b.indexOf('b', null));
 console.log(b.indexOf('b', []));
 ```
 
+If `value` is an empty string or empty `Buffer` and `byteOffset` is less
+than `buf.length`, `byteOffset` will be returned. If `value` is empty and
+`byteOffset` is at least `buf.length`, `buf.length` will be returned.
+
 ### buf.keys()
 <!-- YAML
 added: v1.1.0
@@ -1450,6 +1454,8 @@ console.log(b.lastIndexOf('b', null));
 console.log(b.lastIndexOf('b', []));
 ```
 
+If `value` is an empty string or empty `Buffer`, `byteOffset` will be returned.
+
 ### buf.length
 <!-- YAML
 added: v0.1.90
diff --git a/lib/buffer.js b/lib/buffer.js
index c30cbc2d85..168a607a36 100644
--- a/lib/buffer.js
+++ b/lib/buffer.js
@@ -686,7 +686,7 @@ function bidirectionalIndexOf(buffer, val, byteOffset, encoding, dir) {
   // If the offset is undefined, "foo", {}, coerces to NaN, search whole buffer.
   // `x !== x`-style conditionals are a faster form of `isNaN(x)`
   if (byteOffset !== byteOffset) {
-    byteOffset = dir ? 0 : (buffer.length - 1);
+    byteOffset = dir ? 0 : buffer.length;
   }
   dir = !!dir;  // Cast to bool.
 
diff --git a/src/node_buffer.cc b/src/node_buffer.cc
index 367af6592f..3079d7c5f9 100644
--- a/src/node_buffer.cc
+++ b/src/node_buffer.cc
@@ -919,17 +919,16 @@ void Compare(const FunctionCallbackInfo<Value> &args) {
 // Computes the offset for starting an indexOf or lastIndexOf search.
 // Returns either a valid offset in [0...<length - 1>], ie inside the Buffer,
 // or -1 to signal that there is no possible match.
-int64_t IndexOfOffset(size_t length, int64_t offset_i64, bool is_forward) {
+int64_t IndexOfOffset(size_t length,
+                      int64_t offset_i64,
+                      int64_t needle_length,
+                      bool is_forward) {
   int64_t length_i64 = static_cast<int64_t>(length);
-  if (length_i64 == 0) {
-    // Empty buffer, no match.
-    return -1;
-  }
   if (offset_i64 < 0) {
     if (offset_i64 + length_i64 >= 0) {
       // Negative offsets count backwards from the end of the buffer.
       return length_i64 + offset_i64;
-    } else if (is_forward) {
+    } else if (is_forward || needle_length == 0) {
       // indexOf from before the start of the buffer: search the whole buffer.
       return 0;
     } else {
@@ -937,9 +936,12 @@ int64_t IndexOfOffset(size_t length, int64_t offset_i64, bool is_forward) {
       return -1;
     }
   } else {
-    if (offset_i64 < length_i64) {
+    if (offset_i64 + needle_length <= length_i64) {
       // Valid positive offset.
       return offset_i64;
+    } else if (needle_length == 0) {
+      // Out of buffer bounds, but empty needle: point to end of buffer.
+      return length_i64;
     } else if (is_forward) {
       // indexOf from past the end of the buffer: no match.
       return -1;
@@ -974,11 +976,21 @@ void IndexOfString(const FunctionCallbackInfo<Value>& args) {
   const size_t needle_length =
       StringBytes::Size(args.GetIsolate(), needle, enc);
 
-  if (needle_length == 0 || haystack_length == 0) {
+  int64_t opt_offset = IndexOfOffset(haystack_length,
+                                     offset_i64,
+                                     needle_length,
+                                     is_forward);
+
+  if (needle_length == 0) {
+    // Match String#indexOf() and String#lastIndexOf() behaviour.
+    args.GetReturnValue().Set(static_cast<double>(opt_offset));
+    return;
+  }
+
+  if (haystack_length == 0) {
     return args.GetReturnValue().Set(-1);
   }
 
-  int64_t opt_offset = IndexOfOffset(haystack_length, offset_i64, is_forward);
   if (opt_offset <= -1) {
     return args.GetReturnValue().Set(-1);
   }
@@ -1077,11 +1089,21 @@ void IndexOfBuffer(const FunctionCallbackInfo<Value>& args) {
   const char* needle = buf_data;
   const size_t needle_length = buf_length;
 
-  if (needle_length == 0 || haystack_length == 0) {
+  int64_t opt_offset = IndexOfOffset(haystack_length,
+                                     offset_i64,
+                                     needle_length,
+                                     is_forward);
+
+  if (needle_length == 0) {
+    // Match String#indexOf() and String#lastIndexOf() behaviour.
+    args.GetReturnValue().Set(static_cast<double>(opt_offset));
+    return;
+  }
+
+  if (haystack_length == 0) {
     return args.GetReturnValue().Set(-1);
   }
 
-  int64_t opt_offset = IndexOfOffset(haystack_length, offset_i64, is_forward);
   if (opt_offset <= -1) {
     return args.GetReturnValue().Set(-1);
   }
@@ -1132,8 +1154,8 @@ void IndexOfNumber(const FunctionCallbackInfo<Value>& args) {
   int64_t offset_i64 = args[2]->IntegerValue();
   bool is_forward = args[3]->IsTrue();
 
-  int64_t opt_offset = IndexOfOffset(ts_obj_length, offset_i64, is_forward);
-  if (opt_offset <= -1) {
+  int64_t opt_offset = IndexOfOffset(ts_obj_length, offset_i64, 1, is_forward);
+  if (opt_offset <= -1 || ts_obj_length == 0) {
     return args.GetReturnValue().Set(-1);
   }
   size_t offset = static_cast<size_t>(opt_offset);
diff --git a/test/parallel/test-buffer-includes.js b/test/parallel/test-buffer-includes.js
index 59d6cee94b..44e8fe9671 100644
--- a/test/parallel/test-buffer-includes.js
+++ b/test/parallel/test-buffer-includes.js
@@ -29,10 +29,10 @@ assert(b.includes('bc', -Infinity));
 assert(!b.includes('bc', Infinity));
 assert(b.includes('f'), b.length - 1);
 assert(!b.includes('z'));
-assert(!b.includes(''));
-assert(!b.includes('', 1));
-assert(!b.includes('', b.length + 1));
-assert(!b.includes('', Infinity));
+assert(b.includes(''));
+assert(b.includes('', 1));
+assert(b.includes('', b.length + 1));
+assert(b.includes('', Infinity));
 assert(b.includes(buf_a));
 assert(!b.includes(buf_a, 1));
 assert(!b.includes(buf_a, -1));
@@ -51,10 +51,10 @@ assert(b.includes(buf_bc, -Infinity));
 assert(!b.includes(buf_bc, Infinity));
 assert(b.includes(buf_f), b.length - 1);
 assert(!b.includes(buf_z));
-assert(!b.includes(buf_empty));
-assert(!b.includes(buf_empty, 1));
-assert(!b.includes(buf_empty, b.length + 1));
-assert(!b.includes(buf_empty, Infinity));
+assert(b.includes(buf_empty));
+assert(b.includes(buf_empty, 1));
+assert(b.includes(buf_empty, b.length + 1));
+assert(b.includes(buf_empty, Infinity));
 assert(b.includes(0x61));
 assert(!b.includes(0x61, 1));
 assert(!b.includes(0x61, -1));
diff --git a/test/parallel/test-buffer-indexof.js b/test/parallel/test-buffer-indexof.js
index 680ba1c23d..72505632e3 100644
--- a/test/parallel/test-buffer-indexof.js
+++ b/test/parallel/test-buffer-indexof.js
@@ -31,10 +31,10 @@ assert.strictEqual(b.indexOf('bc', -Infinity), 1);
 assert.strictEqual(b.indexOf('bc', Infinity), -1);
 assert.strictEqual(b.indexOf('f'), b.length - 1);
 assert.strictEqual(b.indexOf('z'), -1);
-assert.strictEqual(b.indexOf(''), -1);
-assert.strictEqual(b.indexOf('', 1), -1);
-assert.strictEqual(b.indexOf('', b.length + 1), -1);
-assert.strictEqual(b.indexOf('', Infinity), -1);
+assert.strictEqual(b.indexOf(''), 0);
+assert.strictEqual(b.indexOf('', 1), 1);
+assert.strictEqual(b.indexOf('', b.length + 1), b.length);
+assert.strictEqual(b.indexOf('', Infinity), b.length);
 assert.strictEqual(b.indexOf(buf_a), 0);
 assert.strictEqual(b.indexOf(buf_a, 1), -1);
 assert.strictEqual(b.indexOf(buf_a, -1), -1);
@@ -53,10 +53,10 @@ assert.strictEqual(b.indexOf(buf_bc, -Infinity), 1);
 assert.strictEqual(b.indexOf(buf_bc, Infinity), -1);
 assert.strictEqual(b.indexOf(buf_f), b.length - 1);
 assert.strictEqual(b.indexOf(buf_z), -1);
-assert.strictEqual(b.indexOf(buf_empty), -1);
-assert.strictEqual(b.indexOf(buf_empty, 1), -1);
-assert.strictEqual(b.indexOf(buf_empty, b.length + 1), -1);
-assert.strictEqual(b.indexOf(buf_empty, Infinity), -1);
+assert.strictEqual(b.indexOf(buf_empty), 0);
+assert.strictEqual(b.indexOf(buf_empty, 1), 1);
+assert.strictEqual(b.indexOf(buf_empty, b.length + 1), b.length);
+assert.strictEqual(b.indexOf(buf_empty, Infinity), b.length);
 assert.strictEqual(b.indexOf(0x61), 0);
 assert.strictEqual(b.indexOf(0x61, 1), -1);
 assert.strictEqual(b.indexOf(0x61, -1), -1);
@@ -429,10 +429,10 @@ assert.strictEqual(b.lastIndexOf(buf_bc, -Infinity), -1);
 assert.strictEqual(b.lastIndexOf(buf_bc, Infinity), 1);
 assert.strictEqual(b.lastIndexOf(buf_f), b.length - 1);
 assert.strictEqual(b.lastIndexOf(buf_z), -1);
-assert.strictEqual(b.lastIndexOf(buf_empty), -1);
-assert.strictEqual(b.lastIndexOf(buf_empty, 1), -1);
-assert.strictEqual(b.lastIndexOf(buf_empty, b.length + 1), -1);
-assert.strictEqual(b.lastIndexOf(buf_empty, Infinity), -1);
+assert.strictEqual(b.lastIndexOf(buf_empty), b.length);
+assert.strictEqual(b.lastIndexOf(buf_empty, 1), 1);
+assert.strictEqual(b.lastIndexOf(buf_empty, b.length + 1), b.length);
+assert.strictEqual(b.lastIndexOf(buf_empty, Infinity), b.length);
 // lastIndexOf number:
 assert.strictEqual(b.lastIndexOf(0x61), 0);
 assert.strictEqual(b.lastIndexOf(0x61, 1), 0);