Speed up HTML fuzzer

htmlDocDumpMemory uses the "HTML" encoding if no other encoding was
specified in the source HTML. This encoding can be extremely slow
because of an inefficiency in htmlEntityValueLookup. Stop encoding
the output for now.

fuzz/html.c

@@ -22,7 +22,7 @@ LLVMFuzzerTestOneInput(const char *data, size_t size) {
     static const size_t maxChunkSize = 128;
     htmlDocPtr doc;
     htmlParserCtxtPtr ctxt;
-    xmlChar *out;
+    xmlOutputBufferPtr out;
     const char *docBuffer;
     size_t docSize, consumed, chunkSize;
     int opts, outSize;
@@ -39,9 +39,16 @@ LLVMFuzzerTestOneInput(const char *data, size_t size) {
     /* Pull parser */
 
     doc = htmlReadMemory(docBuffer, docSize, NULL, NULL, opts);
-    /* Also test the serializer. */
-    htmlDocDumpMemory(doc, &out, &outSize);
-    xmlFree(out);
+
+    /*
+     * Also test the serializer. Call htmlDocContentDumpOutput with our
+     * own buffer to avoid encoding the output. The HTML encoding is
+     * excruciatingly slow (see htmlEntityValueLookup).
+     */
+    out = xmlAllocOutputBuffer(NULL);
+    htmlDocContentDumpOutput(out, doc, NULL);
+    xmlOutputBufferClose(out);
+
     xmlFreeDoc(doc);
 
     /* Push parser */