html解析利器Html Tidy-附示例与交叉编译方式

最近用curl模拟一些网站的登录要对html进行解析，而libxml2要求比较严格，很多html页面无法正确解析，为此才接触到一个c库Html Tidy,故引出本文。

此库官方：http://tidy.sourceforge.net/libintro.html

最新代码可通过cvs下载

cvs -d:pserver:[email protected]:/cvsroot/tidy login
cvs -z3 -d:pserver:[email protected]:/cvsroot/tidy co -P tidy

交叉编译

cd tidy
./build/gnuauto/setup.sh

mkdir ../build-tidy
cd ../build-tidy
../tidy/configure --host=arm-linux CC=arm-linux-gcc --prefix=`pwd`/out
mkdir ./out/include/tidy
mv ./out/include/*.h ./out/include/tidy/
make
make install
cd ../

cvs上下来的源码编译安装后include目录下少了tidy目录，而其示例程序用的是#include <tidy/tidy.h> ，不手动添加就无法编译通过

作者：帅得不敢出门 C++哈哈堂<31843264>

废话不说，上示例：

#include <stdio.h>
#include <tidy/tidy.h>
#include <tidy/buffio.h>
#include <curl/curl.h>

/* curl write callback, to fill tidy's input buffer... */
uint write_cb(char *in, uint size, uint nmemb, TidyBuffer *out) {
  uint r;
  r = size * nmemb;
  tidyBufAppend( out, in, r );
  return(r);
}

/* Traverse the document tree */
void dumpNode(TidyDoc doc, TidyNode tnod, int indent ) {
  TidyNode child;
  for ( child = tidyGetChild(tnod); child; child = tidyGetNext(child) )
  {
    ctmbstr name = tidyNodeGetName( child );
    if ( name )
    {
      /* if it has a name, then it's an HTML tag ... */
      TidyAttr attr;
      printf( "%*.*s%s ", indent, indent, "<", name);
      /* walk the attribute list */
      for ( attr=tidyAttrFirst(child); attr; attr=tidyAttrNext(attr) ) {
        printf(tidyAttrName(attr));
        tidyAttrValue(attr)?printf("=\"%s\" ", tidyAttrValue(attr)):printf(" ");
      }
      printf( ">\n");
    } else {
      /* if it doesn't have a name, then it's probably text, cdata, etc... */
      TidyBuffer buf;
      tidyBufInit(&buf);
      tidyNodeGetText(doc, child, &buf);
      printf("%*.*s\n", indent, indent, buf.bp?(char *)buf.bp:"");
      tidyBufFree(&buf);
    }
    dumpNode( doc, child, indent + 4 ); /* recursive */
  }
}

int main(int argc, char **argv ) {
  CURL *curl;
  char curl_errbuf[CURL_ERROR_SIZE];
  TidyDoc tdoc;
  TidyBuffer docbuf = {0};
  TidyBuffer tidy_errbuf = {0};
  int err;
  if ( argc == 2) {
    curl = curl_easy_init();
    curl_easy_setopt(curl, CURLOPT_URL, argv[1]);
    curl_easy_setopt(curl, CURLOPT_ERRORBUFFER, curl_errbuf);
    curl_easy_setopt(curl, CURLOPT_NOPROGRESS, no);
    curl_easy_setopt(curl, CURLOPT_VERBOSE, yes);
    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_cb);

    tdoc = tidyCreate();
    tidyOptSetBool(tdoc, TidyForceOutput, yes); /* try harder */
    tidyOptSetInt(tdoc, TidyWrapLen, 4096);
    tidySetErrorBuffer( tdoc, &tidy_errbuf );
    tidyBufInit(&docbuf);

    curl_easy_setopt(curl, CURLOPT_WRITEDATA, &docbuf);
    err=curl_easy_perform(curl);
    if ( !err ) {
      err = tidyParseBuffer(tdoc, &docbuf); /* parse the input */
      if ( err >= 0 ) {
        err = tidyCleanAndRepair(tdoc); /* fix any problems */
        if ( err >= 0 ) {
          err = tidyRunDiagnostics(tdoc); /* load tidy error buffer */
          if ( err >= 0 ) {
            dumpNode( tdoc, tidyGetRoot(tdoc), 0 ); /* walk the tree */
            fprintf(stderr, "%s\n", tidy_errbuf.bp); /* show errors */
         }
        }
      }
    } else fprintf(stderr, "%s\n", curl_errbuf);

    /* clean-up */
    curl_easy_cleanup(curl);
    tidyBufFree(&docbuf);
    tidyBufFree(&tidy_errbuf);
    tidyRelease(tdoc);
    return(err);

  } else printf( "usage: %s <url>\n", argv[0] );
  return(0);
}