From ccbd65f3bb10922b89f381c3141c11a10c68ee5c Mon Sep 17 00:00:00 2001
From: Jonathan Hedley <jonathan@hedley.net>
Date: Tue, 17 May 2022 20:30:06 +1000
Subject: [PATCH] In readToByteBuffer, read to max, not the internal buffer
 size

Fixes #1774. Regressed by #1671.
---
 CHANGES                                       |  6 ++++++
 .../internal/ConstrainableInputStream.java    |  9 +++------
 .../org/jsoup/integration/ConnectTest.java    | 20 +++++++++++++++++++
 3 files changed, 29 insertions(+), 6 deletions(-)

diff --git a/CHANGES b/CHANGES
index 2d1e175149..e49ef25ddb 100644
--- a/CHANGES
+++ b/CHANGES
@@ -1,5 +1,11 @@
 jsoup changelog
 
+*** Release 1.15.2 [PENDING]
+  * Bugfix: when using the readToByteBuffer method, such as in Connection.Response.body(), if the document has not
+    already been parsed and must be read fully, and there is any maximum buffer size being applied, only the default
+    internal buffer size is read.
+    <https://github.com/jhy/jsoup/issues/1774>
+
 *** Release 1.15.1 [2022-May-15]
   * Change: removed previously deprecated methods and classes (including org.jsoup.safety.Whitelist; use
     org.jsoup.safety.Safelist instead).
diff --git a/src/main/java/org/jsoup/internal/ConstrainableInputStream.java b/src/main/java/org/jsoup/internal/ConstrainableInputStream.java
index 0f7e3c3b88..5b6491363e 100644
--- a/src/main/java/org/jsoup/internal/ConstrainableInputStream.java
+++ b/src/main/java/org/jsoup/internal/ConstrainableInputStream.java
@@ -81,17 +81,14 @@ public ByteBuffer readToByteBuffer(int max) throws IOException {
         final ByteArrayOutputStream outStream = new ByteArrayOutputStream(bufferSize);
 
         int read;
-        int remaining = bufferSize;
-
         while (true) {
-            read = read(readBuffer, 0, remaining);
+            read = read(readBuffer, 0, bufferSize);
             if (read == -1) break;
             if (localCapped) { // this local byteBuffer cap may be smaller than the overall maxSize (like when reading first bytes)
-                if (read >= remaining) {
-                    outStream.write(readBuffer, 0, remaining);
+                if (read >= max) {
+                    outStream.write(readBuffer, 0, max);
                     break;
                 }
-                remaining -= read;
             }
             outStream.write(readBuffer, 0, read);
         }
diff --git a/src/test/java/org/jsoup/integration/ConnectTest.java b/src/test/java/org/jsoup/integration/ConnectTest.java
index 28e79dac60..42b4f950e1 100644
--- a/src/test/java/org/jsoup/integration/ConnectTest.java
+++ b/src/test/java/org/jsoup/integration/ConnectTest.java
@@ -671,4 +671,24 @@ public void maxBodySize() throws IOException {
         assertEquals("Large HTML", doc1.title());
         assertEquals("Large HTML", doc2.title());
     }
+
+    @Test
+    public void maxBodySizeInReadToByteBuffer() throws IOException {
+        // https://github.com/jhy/jsoup/issues/1774
+        // when calling readToByteBuffer, contents were not buffered up
+        String url = FileServlet.urlTo("/htmltests/large.html"); // 280 K
+
+        Connection.Response defaultRes = Jsoup.connect(url).execute();
+        Connection.Response smallRes = Jsoup.connect(url).maxBodySize(50 * 1024).execute(); // crops
+        Connection.Response mediumRes = Jsoup.connect(url).maxBodySize(200 * 1024).execute(); // crops
+        Connection.Response largeRes = Jsoup.connect(url).maxBodySize(300 * 1024).execute(); // does not crop
+        Connection.Response unlimitedRes = Jsoup.connect(url).maxBodySize(0).execute();
+
+        int actualDocText = 280735;
+        assertEquals(actualDocText, defaultRes.body().length());
+        assertEquals(50 * 1024, smallRes.body().length());
+        assertEquals(200 * 1024, mediumRes.body().length());
+        assertEquals(actualDocText, largeRes.body().length());
+        assertEquals(actualDocText, unlimitedRes.body().length());
+    }
 }