kc3-lang/libxml2/fuzz/html.c

• Show log

  Commit

• Hash : ec808a44
  Author :
  Date : 2021-02-07T13:57:49
  

  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.
  

Download

• fuzz/html.c

• /*
   * html.c: a libFuzzer target to test several HTML parser interfaces.
   *
   * See Copyright for the status of this software.
   */
  
  #include <libxml/HTMLparser.h>
  #include <libxml/HTMLtree.h>
  #include "fuzz.h"
  
  int
  LLVMFuzzerInitialize(int *argc ATTRIBUTE_UNUSED,
                       char ***argv ATTRIBUTE_UNUSED) {
      xmlInitParser();
      xmlSetGenericErrorFunc(NULL, xmlFuzzErrorFunc);
  
      return 0;
  }
  
  int
  LLVMFuzzerTestOneInput(const char *data, size_t size) {
      static const size_t maxChunkSize = 128;
      htmlDocPtr doc;
      htmlParserCtxtPtr ctxt;
      xmlOutputBufferPtr out;
      const char *docBuffer;
      size_t docSize, consumed, chunkSize;
      int opts, outSize;
  
      xmlFuzzDataInit(data, size);
      opts = xmlFuzzReadInt();
  
      docBuffer = xmlFuzzReadRemaining(&docSize);
      if (docBuffer == NULL) {
          xmlFuzzDataCleanup();
          return(0);
      }
  
      /* Pull parser */
  
      doc = htmlReadMemory(docBuffer, docSize, NULL, NULL, opts);
  
      /*
       * 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 */
  
      ctxt = htmlCreatePushParserCtxt(NULL, NULL, NULL, 0, NULL,
                                      XML_CHAR_ENCODING_NONE);
      htmlCtxtUseOptions(ctxt, opts);
  
      for (consumed = 0; consumed < docSize; consumed += chunkSize) {
          chunkSize = docSize - consumed;
          if (chunkSize > maxChunkSize)
              chunkSize = maxChunkSize;
          htmlParseChunk(ctxt, docBuffer + consumed, chunkSize, 0);
      }
  
      htmlParseChunk(ctxt, NULL, 0, 1);
      xmlFreeDoc(ctxt->myDoc);
      htmlFreeParserCtxt(ctxt);
  
      /* Cleanup */
  
      xmlFuzzDataCleanup();
  
      return(0);
  }
  
  

Download