fix(appsec): use CharsetDecoder with REPORT for charset fallback in FileItemContentReader

Author: jandro996

dd-java-agent/instrumentation/commons-fileupload-1.5/src/main/java/datadog/trace/instrumentation/commons/fileupload/FileItemContentReader.java

@@ -2,7 +2,10 @@
 
 import java.io.IOException;
 import java.io.InputStream;
+import java.nio.ByteBuffer;
+import java.nio.charset.CharacterCodingException;
 import java.nio.charset.Charset;
+import java.nio.charset.CodingErrorAction;
 import java.nio.charset.StandardCharsets;
 import java.util.ArrayList;
 import java.util.List;
@@ -47,8 +50,13 @@ static String decodeBytes(byte[] buf, int length, String contentType) {
     Charset charset = extractCharset(contentType);
     if (charset == null) charset = StandardCharsets.UTF_8;
     try {
-      return new String(buf, 0, length, charset);
-    } catch (Exception e) {
+      return charset
+          .newDecoder()
+          .onMalformedInput(CodingErrorAction.REPORT)
+          .onUnmappableCharacter(CodingErrorAction.REPORT)
+          .decode(ByteBuffer.wrap(buf, 0, length))
+          .toString();
+    } catch (CharacterCodingException e) {
       return new String(buf, 0, length, StandardCharsets.ISO_8859_1);
     }
   }

dd-java-agent/instrumentation/commons-fileupload-1.5/src/test/groovy/FileItemContentReaderTest.groovy

@@ -66,6 +66,16 @@ class FileItemContentReaderTest extends Specification {
     FileItemContentReader.readContent(item) == text
   }
 
+  void 'readContent falls back to ISO-8859-1 when bytes are invalid UTF-8'() {
+    given:
+    // 0xE9 is 'é' in ISO-8859-1 but an invalid lone UTF-8 byte
+    byte[] iso88591Bytes = 'café'.getBytes('ISO-8859-1')
+    def item = fileItemFromBytes(iso88591Bytes, 'file.txt', null)
+
+    expect:
+    FileItemContentReader.readContent(item) == 'café'
+  }
+
   void 'readContents returns content for each non-form file with a name'() {
     given:
     def items = [fileItem('content-a', 'file-a.txt'), fileItem('content-b', 'file-b.txt'),]