From 5c6171f2eb58d866e42417b8a73c41f8db2f7664 Mon Sep 17 00:00:00 2001
From: Sebastian Nagel <sebastian@commoncrawl.org>
Date: Fri, 8 Nov 2024 21:05:58 +0100
Subject: [PATCH] Add optional support for brotli content encoding

---
 pom.xml                                    |  7 +++++++
 src/org/netpreserve/jwarc/BrotliUtils.java | 22 ++++++++++++++++++++++
 src/org/netpreserve/jwarc/DecodedBody.java |  7 ++++++-
 src/org/netpreserve/jwarc/HttpMessage.java |  2 ++
 4 files changed, 37 insertions(+), 1 deletion(-)
 create mode 100644 src/org/netpreserve/jwarc/BrotliUtils.java

diff --git a/pom.xml b/pom.xml
index 61994a0..051fb08 100644
--- a/pom.xml
+++ b/pom.xml
@@ -68,6 +68,13 @@
             <version>4.13.2</version>
             <scope>test</scope>
         </dependency>
+        <dependency>
+          <groupId>org.brotli</groupId>
+          <artifactId>dec</artifactId>
+          <version>0.1.2</version>
+          <scope>compile</scope>
+          <optional>true</optional>
+        </dependency>
     </dependencies>
 
     <properties>
diff --git a/src/org/netpreserve/jwarc/BrotliUtils.java b/src/org/netpreserve/jwarc/BrotliUtils.java
new file mode 100644
index 0000000..f080a8c
--- /dev/null
+++ b/src/org/netpreserve/jwarc/BrotliUtils.java
@@ -0,0 +1,22 @@
+/*
+ * SPDX-License-Identifier: Apache-2.0
+ * Copyright (C) 2024 National Library of Australia and the jwarc contributors
+ */
+
+package org.netpreserve.jwarc;
+
+import java.io.IOException;
+import java.nio.channels.Channels;
+import java.nio.channels.ReadableByteChannel;
+
+import org.brotli.dec.BrotliInputStream;
+
+/**
+ * Utility class to read brotli-encoded data, based on org.brotli:dec.
+ */
+public final class BrotliUtils {
+
+    public static ReadableByteChannel brotliChannel(ReadableByteChannel brotli) throws IOException {
+        return Channels.newChannel(new BrotliInputStream(Channels.newInputStream(brotli)));
+    }
+}
diff --git a/src/org/netpreserve/jwarc/DecodedBody.java b/src/org/netpreserve/jwarc/DecodedBody.java
index 4f54e0e..6254be8 100644
--- a/src/org/netpreserve/jwarc/DecodedBody.java
+++ b/src/org/netpreserve/jwarc/DecodedBody.java
@@ -35,7 +35,12 @@ private DecodedBody(ReadableByteChannel channel, Encoding encoding) throws IOExc
             this.channel = IOUtils.gunzipChannel(channel);
             break;
         case BROTLI:
-            throw new IOException("Brotli encoding not yet supported");
+            try {
+                this.channel = BrotliUtils.brotliChannel(channel);
+            } catch (NoClassDefFoundError e) {
+                throw new IOException("Brotli decoder not found, please install org.brotli:dec", e);
+            }
+            break;
         default:
             throw new IOException("Unsupported encoding");
         }
diff --git a/src/org/netpreserve/jwarc/HttpMessage.java b/src/org/netpreserve/jwarc/HttpMessage.java
index 037466c..809a009 100644
--- a/src/org/netpreserve/jwarc/HttpMessage.java
+++ b/src/org/netpreserve/jwarc/HttpMessage.java
@@ -42,6 +42,8 @@ public MessageBody bodyDecoded() throws IOException {
         } else if (contentEncodings.get(0).equalsIgnoreCase("gzip")
                 || contentEncodings.get(0).equalsIgnoreCase("x-gzip")) {
             return DecodedBody.create(payload, DecodedBody.Encoding.GZIP);
+        } else if (contentEncodings.get(0).equalsIgnoreCase("br")) {
+            return DecodedBody.create(payload, DecodedBody.Encoding.BROTLI);
         } else if (contentEncodings.get(0).equalsIgnoreCase("deflate")) {
             return DecodedBody.create(payload, DecodedBody.Encoding.DEFLATE);
         } else {