TidyNetworkReply::~TidyNetworkReply() {
delete nam;
if (!this->attribute(QNetworkRequest::RedirectionTargetAttribute).isValid()) {
tidyBufFree( &output );
tidyBufFree( &errbuf );
tidyRelease( tdoc );
}
}
QString tidyHtml(QString str, bool& ok) {
#ifdef NO_TIDY
ok = true;
return str;
#else
QString res = str;
ok = false;
static bool isTidyWithIntBodyOnly = isTidyWithIntBodyOnlyCheck();
TidyDoc tdoc = tidyCreate();
TidyBuffer output;
TidyBuffer errbuf;
tidyBufInit(&output);
tidyBufInit(&errbuf);
bool configOk =
tidyOptSetBool(tdoc, TidyXhtmlOut, yes) &&
tidyOptSetBool(tdoc, TidyForceOutput, yes) &&
tidyOptSetBool(tdoc, TidyMark, no) &&
(isTidyWithIntBodyOnly
? tidyOptSetInt(tdoc, TidyBodyOnly, 1)
: tidyOptSetBool(tdoc, TidyBodyOnly, yes)) &&
tidyOptSetInt(tdoc, TidyWrapLen, 0) &&
tidyOptSetInt(tdoc, TidyDoctypeMode, TidyDoctypeOmit);
if (configOk &&
(tidySetCharEncoding(tdoc, "utf8") >= 0) &&
(tidySetErrorBuffer(tdoc, &errbuf) >= 0) &&
(tidyParseString(tdoc, str.toUtf8().data()) >= 0) &&
(tidyCleanAndRepair(tdoc) >= 0) &&
(tidyRunDiagnostics(tdoc) >= 0) &&
(tidySaveBuffer(tdoc, &output) >= 0) &&
(output.bp != 0 && output.size > 0)) {
res = QString::fromUtf8((char*)output.bp, output.size);
ok = true;
}
#ifdef DEBUG_MARKUP
if (errbuf.size > 0) {
QString errStr = QString::fromUtf8((char*)errbuf.bp, errbuf.size);
qDebug() << "\n[DEBUG] MARKUP, libtidy errors and warnings:\n" << errStr;
}
#endif
if (output.bp != 0)
tidyBufFree(&output);
if (errbuf.bp != 0)
tidyBufFree(&errbuf);
tidyRelease(tdoc);
return res.trimmed();
#endif
}
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, 0L);
curl_easy_setopt(curl, CURLOPT_VERBOSE, 1L);
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);
}
static html_valid_status_t
html_valid_run (html_valid_t * htv, SV * html,
SV ** output_ptr, SV ** errors_ptr)
{
const char * html_string;
STRLEN html_length;
SV * output;
SV * errors;
TidyBuffer tidy_output = {0};
TidyBuffer tidy_errbuf = {0};
/* First set these up sanely in case the stuff hits the fan. */
* output_ptr = & PL_sv_undef;
* errors_ptr = & PL_sv_undef;
/* Work around bug where allocator sometimes does not get set. */
CopyAllocator (htv->tdoc, & tidy_output);
CopyAllocator (htv->tdoc, & tidy_errbuf);
html_string = SvPV (html, html_length);
CALL_TIDY (tidySetErrorBuffer (htv->tdoc, & tidy_errbuf));
htv->n_mallocs++;
CALL_TIDY (tidyParseString (htv->tdoc, html_string));
CALL_TIDY (tidyCleanAndRepair (htv->tdoc));
CALL_TIDY (tidyRunDiagnostics (htv->tdoc));
CALL_TIDY (tidySaveBuffer (htv->tdoc, & tidy_output));
htv->n_mallocs++;
/* Copy the contents of the buffers into the Perl scalars. */
output = newSVpv ((char *) tidy_output.bp, tidy_output.size);
errors = newSVpv ((char *) tidy_errbuf.bp, tidy_errbuf.size);
/* HTML Tidy randomly segfaults here due to "allocator" not being
set in some cases, hence the above CopyAllocator fix. */
tidyBufFree (& tidy_output);
htv->n_mallocs--;
tidyBufFree (& tidy_errbuf);
htv->n_mallocs--;
/* These are not our mallocs, they are Perl's mallocs, so we don't
increase htv->n_mallocs for these. After we return them, we no
longer take care of these. */
* output_ptr = output;
* errors_ptr = errors;
return html_valid_ok;
}
void TidyReader::closeFile (void)
{
if (m_tidy)
{
tidyBufFree (&m_outbuf);
#ifndef DEBUG
tidyBufFree (&m_errbuf);
#endif
tidyRelease (m_tidy);
m_tidy = 0;
}
memset (&m_outbuf, 0 , sizeof (TidyBuffer));
memset (&m_errbuf, 0 , sizeof (TidyBuffer));
}
QVector<QString> UCHome_Main_SiteConst::find_sys_notice(TidyDoc doc)
{
QVector<QString> sys_notes;
QString note;
char *notes_ids[] = { "mtag_invite", "f_request", NULL };
char *nid = NULL;
TidyNode node = NULL;
TidyNode node2 = NULL;
TidyBuffer tbuf;
for(int i = 0; i < 16 ; i++) {
nid = notes_ids[i];
if(nid == NULL) {
break;
}
node = this->searchNode(doc, NULL, nid, TidyTag_DIV);
if(node != NULL) {
tidyBufInit(&tbuf);
tidyNodeGetText(doc, node, &tbuf);
note = this->u8codec->toUnicode(QByteArray((char*)tbuf.bp));
sys_notes.append(note);
tidyBufFree(&tbuf);
q_debug()<<note;
}else{
q_debug()<<"Warning: "<<nid<<" not found";
}
}
return sys_notes;
}
void CCaHtmlParse::__GetPriceAndRamainTicket(UINT *pPrice, UINT *pRemainTicket, const TidyDoc & tdoc, const TidyNode & tdNode)
{
CStringA straRet;
*pPrice = 0;
*pRemainTicket = 0;
TidyBuffer text = {0};
tidyBufInit(&text);
TidyNodeType type = tidyNodeGetType(tdNode);
tidyNodeGetText(tdoc, tdNode, &text);
straRet.Format("%s",text.bp);
straRet.TrimLeft();
CStringA straKey("</font>");
int iPos = straRet.Find(straKey);
int iEndPos = straRet.Find("</strong>");
int iStartPos = iPos+straKey.GetLength();
CStringA straPrice = straRet.Mid(iStartPos, iEndPos-iStartPos);
straPrice.Remove(0x0d);//去掉回车
straPrice.Remove(0x0a);//去掉换行
*pPrice = atoi(straPrice.GetBuffer(0));
straPrice.ReleaseBuffer();
//剩余座位
straKey = "：";
iPos = straRet.Find(straKey);
iEndPos = straRet.Find("</td>");
iStartPos = iPos+straKey.GetLength();
CStringA straRemainSeat = straRet.Mid(iStartPos, iEndPos-iStartPos);
*pRemainTicket = atoi(straRemainSeat.GetBuffer(0));
straRemainSeat.ReleaseBuffer();
tidyBufFree(&text);
}
QString UCHome_Main_SiteConst::find_sys_notice_by_type(TidyDoc doc, char *type_class)
{
QString note;
char *nid = type_class;
TidyNode node = NULL;
TidyNode node2 = NULL;
TidyBuffer tbuf;
if(nid == NULL) {
q_debug()<<"";
return note;
}
node = this->searchNode(doc, NULL, nid, TidyTag_DIV);
if(node != NULL) {
tidyBufInit(&tbuf);
tidyNodeGetText(doc, node, &tbuf);
note = this->u8codec->toUnicode(QByteArray((char*)tbuf.bp));
tidyBufFree(&tbuf);
q_debug()<<note;
}else{
q_debug()<<"Warning: "<<nid<<" not found";
}
return note;
}
static int tidy_doc_cast_handler(zval *in, zval *out, int type)
{
TidyBuffer output;
PHPTidyObj *obj;
switch (type) {
case IS_LONG:
ZVAL_LONG(out, 0);
break;
case IS_DOUBLE:
ZVAL_DOUBLE(out, 0);
break;
case _IS_BOOL:
ZVAL_TRUE(out);
break;
case IS_STRING:
obj = Z_TIDY_P(in);
tidyBufInit(&output);
tidySaveBuffer (obj->ptdoc->doc, &output);
ZVAL_STRINGL(out, (char *) output.bp, output.size ? output.size-1 : 0);
tidyBufFree(&output);
break;
default:
return FAILURE;
}
return SUCCESS;
}
static void tidy_doc_update_properties(PHPTidyObj *obj)
{
TidyBuffer output;
zval temp;
tidyBufInit(&output);
tidySaveBuffer (obj->ptdoc->doc, &output);
if (output.size) {
if (!obj->std.properties) {
rebuild_object_properties(&obj->std);
}
ZVAL_STRINGL(&temp, (char*)output.bp, output.size-1);
zend_hash_str_update(obj->std.properties, "value", sizeof("value") - 1, &temp);
}
tidyBufFree(&output);
if (obj->ptdoc->errbuf->size) {
if (!obj->std.properties) {
rebuild_object_properties(&obj->std);
}
ZVAL_STRINGL(&temp, (char*)obj->ptdoc->errbuf->bp, obj->ptdoc->errbuf->size-1);
zend_hash_str_update(obj->std.properties, "errorBuffer", sizeof("errorBuffer") - 1, &temp);
}
}
/* 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 */
}
}
void HTMLTidy::run() throw( std::runtime_error ) {
TidyBuffer outputBuffer = { 0 };
TidyBuffer errorBuffer = { 0 };
// try to create valid XHTML document for XML parser:
int tidyResult = -1;
if( tidyOptSetBool( handle, TidyXhtmlOut, yes ) ) {
tidyResult = tidySetErrorBuffer( handle, &errorBuffer );
}
if( tidyResult >= 0 ) {
tidyResult = tidyParseString( handle, document.c_str() );
}
if( tidyResult >= 0 ) {
tidyResult = tidyCleanAndRepair( handle );
}
if( tidyResult >= 0 ) {
tidyResult = tidyRunDiagnostics( handle );
}
if( tidyResult > 1 ) {
if( !tidyOptSetBool( handle, TidyForceOutput, yes ) ) {
tidyResult = -1;
}
}
if( tidyResult >= 0 ) {
tidyResult = tidySaveBuffer( handle, &outputBuffer );
}
if( tidyResult > 0 ) {
std::clog << "*********************************" << std::endl;
std::clog << "HTMLTidy: Diagnostics of libtidy:" << std::endl;
std::clog << errorBuffer.bp;
std::clog << "*********************************" << std::endl;
}
else if( tidyResult < 0 ) {
std::stringstream sstrTidyResult;
sstrTidyResult << tidyResult;
throw std::runtime_error( "HTMLTidy: A severe error occured while tidying up the received document ("
+ sstrTidyResult.str()
+ ")."
);
}
resultDocument.reserve( outputBuffer.size ); // avoid frequent (re-)allocations
for( unsigned int i = 0; i < outputBuffer.size; i++ ) {
resultDocument.insert( resultDocument.end(), static_cast< char >( *(outputBuffer.bp + i) ) );
}
tidyBufFree( &outputBuffer );
tidyBufFree( &errorBuffer );
}
void TIDY_CALL freeFileSource( TidyInputSource* inp, Bool closeIt )
{
FileSource* fin = (FileSource*) inp->sourceData;
if ( closeIt && fin && fin->fp )
fclose( fin->fp );
tidyBufFree( &fin->unget );
MemFree( fin );
}
bool CCFHtmlTidy::TidyMain(const char* pSourceIn, const char* pOptions, std::string &strOut, std::string &strErr)
{
TidyBuffer output;
TidyBuffer errbuf;
int rc = -1;
Bool ok = yes;
TidyDoc tdoc = tidyCreate(); // Initialize "document"
tidyBufInit(&output);
tidyBufInit(&errbuf);
TidyOptionsSet(tidyDocToImpl(tdoc), pOptions);
if (ok)
rc = tidySetErrorBuffer(tdoc, &errbuf); // Capture diagnostics
if (rc >= 0)
rc = tidyParseString(tdoc, pSourceIn); // Parse the input
if (rc >= 0)
rc = tidyCleanAndRepair(tdoc); // Tidy it up!
if (rc >= 0)
rc = tidyRunDiagnostics(tdoc); // Kvetch
//if ( rc > 1 ) // If error, force output.
// rc = ( tidyOptSetBool(tdoc, TidyForceOutput, yes) ? rc : -1 );
if (rc >= 0)
rc = tidySaveBuffer(tdoc, &output); // Pretty Print
if (rc >= 0)
{
if (output.bp)
{
strOut = reinterpret_cast<char const*>(output.bp);
}
}
strErr = reinterpret_cast<char const*>(errbuf.bp);
std::string strEmpty = "No warnings or errors were found.\n\n";
if (0 == strEmpty.compare(strErr))
{
strErr.clear();
}
tidyBufFree(&output);
tidyBufFree(&errbuf);
tidyRelease(tdoc);
return true;
}
QString tidy(QString input)
// take html code and return it converted to xhtml code
{
// the following code is (c) Charles Reitzel and Dave Raggett, see the package tidy
TidyBuffer output = {0};
TidyBuffer errbuf = {0};
QString result;
int rc = -1;
Bool ok;
TidyDoc tdoc = tidyCreate(); // Initialize "document"
ok = tidyOptSetBool( tdoc, TidyXhtmlOut, yes ); // Convert to XHTML
if ( ok ) rc = tidySetErrorBuffer( tdoc, &errbuf ); // Capture diagnostics
tidySetCharEncoding( tdoc, "utf8" );
if ( rc >= 0 ) rc = tidyParseString( tdoc, input.toUtf8().constData() ); // Parse the input
if ( rc >= 0 ) rc = tidyCleanAndRepair( tdoc ); // Tidy it up!
if ( rc >= 0 ) rc = tidyRunDiagnostics( tdoc ); // Kvetch
if ( rc > 1 ) // If error, force output.
rc = ( tidyOptSetBool(tdoc, TidyForceOutput, yes) ? rc : -1 );
if ( rc >= 0 ) rc = tidySaveBuffer( tdoc, &output ); // Pretty Print
if ( rc >= 0 )
{
char* outputstring; // content of the outputfile
// find out length of outputstring
int length=0; // length of outputstring
byte* string=output.bp;
while (*string)
{
string++;
length++;
}
outputstring=(char*)malloc(length);
snprintf(outputstring,length,"%s",output.bp);
result=QString::fromUtf8(outputstring,length);
}
else
printf( "A severe error (\%d) occurred.\\n", rc );
tidyBufFree( &output );
tidyBufFree( &errbuf );
tidyRelease( tdoc );
result=result.replace("ö","ö");
return result;
}
void tidy(std::string &input)
{
TidyBuffer output = {0};
TidyBuffer errbuf = {0};
TidyDoc tdoc = tidyCreate();
tidyOptSetBool(tdoc, TidyXhtmlOut, yes);
tidySetErrorBuffer(tdoc, &errbuf); // Capture diagnostics
tidyParseString(tdoc, input.c_str());
tidyCleanAndRepair(tdoc);
tidySaveBuffer(tdoc, &output);
input = std::string((const char*)output.bp);
tidyBufFree(&output);
tidyBufFree(&errbuf);
tidyRelease(tdoc);
}
bool HTidyInterface::formatSource( const char* textIn, CString &strTidy, CString &strMsg )
{
TidyBuffer output;
TidyBuffer errbuf;
int rc = -1;
Bool ok = yes;
TidyDoc tdoc = tidyCreate(); // Initialize "document"
tidyBufInit(&output);
tidyBufInit(&errbuf);
InitTidyDefault(tdoc);
SetTidyConfig(tdoc);
if ( ok )
rc = tidySetErrorBuffer(tdoc, &errbuf); // Capture diagnostics
if ( rc >= 0 )
rc = tidyParseString(tdoc, textIn); // Parse the input
if ( rc >= 0 )
rc = tidyCleanAndRepair(tdoc); // Tidy it up!
if ( rc >= 0 )
rc = tidyRunDiagnostics(tdoc); // Kvetch
//if ( rc > 1 ) // If error, force output.
// rc = ( tidyOptSetBool(tdoc, TidyForceOutput, yes) ? rc : -1 );
if ( rc >= 0 )
rc = tidySaveBuffer(tdoc, &output); // Pretty Print
if ( rc >= 0 )
{
strTidy = reinterpret_cast< char const* >(output.bp);
}
strMsg = reinterpret_cast< char const* >(errbuf.bp);
CString strEmpty = _T("No warnings or errors were found.\r\n\r\n");
if (0 == strEmpty.Compare(strMsg))
{
strMsg.Empty();
}
tidyBufFree(&output);
tidyBufFree(&errbuf);
tidyRelease(tdoc);
return true;
}
// formats the 'dom' tree into HTML output.
int lua_tidy_saveString(lua_State *L)
{
pTidy t = toTidy(L,1);
TidyBuffer output;
tidyBufInit(&output);
lua_pushnumber(L, tidySaveBuffer(t->tdoc, &output));
lua_pushlstring(L, (char*)output.bp, output.size);
tidyBufFree( &output );
return 1;
}
void CCaHtmlParse::__GetAirPortCode(CStringA & straDCode, CStringA & straACode, const TidyDoc & tdoc, const TidyNode & tdNode)
{
CStringA straRet;
CStringA straSha("上海虹桥");
CStringA straPvg("上海浦东");
CStringA straPek("北京首都");
CStringA straNay("北京南苑");
straDCode = "";
straACode = "";
TidyBuffer text = {0};
tidyBufInit(&text);
TidyNodeType type = tidyNodeGetType(tdNode);
tidyNodeGetText(tdoc, tdNode, &text);
straRet.Format("%s",text.bp);
straRet.TrimLeft();
CStringA straKey("<br />");//<br />后有回车换行符
int iPos = straRet.Find(straKey);
CStringA straDCity = straRet.Mid(4, iPos-4);
if (-1 != straDCity.Find(straSha))
straDCode = "SHA";
else if (-1 != straDCity.Find(straPvg))
straDCode = "PVG";
else if (-1 != straDCity.Find(straPek))
straDCode = "PEK";
else if(-1 != straDCity.Find(straNay))
straDCode = "NAY";
else
{
}
//取到达机场，<br />后有回车换行符
straRet = straRet.Mid(iPos+straKey.GetLength());
straRet.Remove(0x0d);//去掉回车
straRet.Remove(0x0a);//去掉换行
int iEndPos = straRet.Find("</td>");
CStringA straACity = straRet.Left(iEndPos);
if (-1 != straACity.Find(straSha))
straACode = "SHA";
else if (-1 != straACity.Find(straPvg))
straACode = "PVG";
else if (-1 != straACity.Find(straPek))
straACode = "PEK";
else if(-1 != straACity.Find(straNay))
straACode = "NAY";
else
{
}
tidyBufFree(&text);
}
/* {{{ proto string tidy_get_output()
Return a string representing the parsed tidy markup */
static PHP_FUNCTION(tidy_get_output)
{
TidyBuffer output;
TIDY_FETCH_OBJECT;
tidyBufInit(&output);
tidySaveBuffer(obj->ptdoc->doc, &output);
FIX_BUFFER(&output);
RETVAL_STRINGL((char *) output.bp, output.size ? output.size-1 : 0);
tidyBufFree(&output);
}
int main(int argc, char **argv )
{
const char* input = "<title>Hello</title><p>World!";
TidyBuffer output = {0};
TidyBuffer errbuf = {0};
int rc = -1;
Bool ok;
// Initialize "document"
TidyDoc tdoc = tidyCreate();
printf( "Tidying:\t%s\n", input );
// Convert to XHTML
ok = tidyOptSetBool( tdoc, TidyXhtmlOut, yes );
if ( ok )
rc = tidySetErrorBuffer( tdoc, &errbuf ); // Capture diagnostics
if ( rc >= 0 )
rc = tidyParseString( tdoc, input ); // Parse the input
if ( rc >= 0 )
rc = tidyCleanAndRepair( tdoc ); // Tidy it up!
if ( rc >= 0 )
rc = tidyRunDiagnostics( tdoc ); // Kvetch
if ( rc > 1 ) // If error, force output.
rc = ( tidyOptSetBool(tdoc, TidyForceOutput, yes) ? rc : -1 );
if ( rc >= 0 )
rc = tidySaveBuffer( tdoc, &output ); // Pretty Print
if ( rc >= 0 )
{
if ( rc > 0 )
printf( "\nDiagnostics:\n\n%s", errbuf.bp );
printf( "\nAnd here is the result:\n\n%s", output.bp );
}
else
printf( "A severe error (%d) occurred.\n", rc );
tidyBufFree( &output );
tidyBufFree( &errbuf );
tidyRelease( tdoc );
return rc;
}
// garbage collection, make sure we mark all opened nodes as expired.
int lua_tidy_gc ( lua_State *L )
{
pTidy t = toTidy(L,1);
doc_clear_all_nodes(t); // all nodes are marked as expired.
tidyRelease( t->tdoc );
tidyBufFree( &t->errbuf );
t->tdoc = NULL;
return 0;
}
// quick and dirty shortcut function.
int lua_tidy_easyClean ( lua_State *L )
{
TidyBuffer output;
TidyBuffer errbuf;
int rc;
pTidy t;
const char * input;
tidyBufInit(&output);
tidyBufInit(&errbuf);
rc = -1;
t = toTidy(L,1);
input = lua_tostring(L,2);
rc = tidySetErrorBuffer( t->tdoc, &errbuf );
if ( rc >= 0 )
rc = tidyParseString( t->tdoc, input );
if ( rc >= 0 )
rc = tidyCleanAndRepair( t->tdoc );
if ( rc >= 0 )
rc = tidyRunDiagnostics( t->tdoc );
if ( rc >= 0 )
rc = tidySaveBuffer( t->tdoc, &output );
lua_pushlstring(L, (char*)output.bp,output.size);
if ( rc != 0 )
lua_pushlstring(L, (char*)errbuf.bp,errbuf.size);
else
lua_pushnil(L);
lua_pushnumber(L, rc);
tidyBufFree( &output );
tidyBufFree( &errbuf );
return 3;
}
char *copyTidyNodeValue(TidyNode tnode, TidyDoc tdoc)
{
TidyBuffer buf;
tidyBufInit(&buf);
tidyNodeGetValue(tdoc,tnode,&buf);
char *str = (char *)xmalloc(buf.size+1);
memcpy(str,buf.bp,buf.size);
str[buf.size] = '\0';
tidyBufFree(&buf);
return str;
}
std::string cleanHTML (std::string html)
{
TidyDoc tidyDoc = tidyCreate();
TidyBuffer tidyOutputBuffer = {0};
// Configure Tidy
// The flags tell Tidy to output XML and disable showing warnings
bool configSuccess = tidyOptSetBool(tidyDoc, TidyXmlOut, yes)
&& tidyOptSetBool(tidyDoc, TidyQuiet, yes)
&& tidyOptSetBool(tidyDoc, TidyNumEntities, yes)
&& tidyOptSetBool(tidyDoc, TidyShowWarnings, no);
tidyOptSetValue(tidyDoc,TidyForceOutput,"true");
int tidyResponseCode = -1;
// Parse input
if (configSuccess)
tidyResponseCode = tidyParseString(tidyDoc, html.c_str());
// Process HTML
if (tidyResponseCode >= 0)
tidyResponseCode = tidyCleanAndRepair(tidyDoc);
// Output the HTML to our buffer
if (tidyResponseCode >= 0)
tidyResponseCode = tidySaveBuffer(tidyDoc, &tidyOutputBuffer);
// Any errors from Tidy?
if (tidyResponseCode < 0)
throw ("Tidy encountered an error while parsing an HTML response. Tidy response code: " + tidyResponseCode);
// Grab the result from the buffer and then free Tidy's memory
std::string tidyResult = (char*)tidyOutputBuffer.bp;
tidyBufFree(&tidyOutputBuffer);
tidyRelease(tidyDoc);
return tidyResult;
}
static void tidy_object_free_storage(zend_object *object)
{
PHPTidyObj *intern = php_tidy_fetch_object(object);
zend_object_std_dtor(&intern->std);
if (intern->ptdoc) {
intern->ptdoc->ref_count--;
if (intern->ptdoc->ref_count <= 0) {
tidyBufFree(intern->ptdoc->errbuf);
efree(intern->ptdoc->errbuf);
tidyRelease(intern->ptdoc->doc);
efree(intern->ptdoc);
}
}
}
FeedRecord* UCHome_Main_SiteConst::parse_note(void *state_data, TidyDoc doc, TidyNode node)
{
FeedRecord * rec = NULL;
QString note;
TidyBuffer tbuf;
tidyBufInit(&tbuf);
tidyNodeGetText(doc, node, &tbuf);
note = this->u8codec->toUnicode(QByteArray((char*)tbuf.bp));
rec = new FeedRecord();
rec->content = this->rewrite_relative_link(note);
md5CheckSum((char*)tbuf.bp, tbuf.size, rec->md5sum);
tidyBufFree(&tbuf);
//q_debug()<<"Orig:"<<note;
return rec;
}
int tidyDocSaveString( TidyDocImpl* doc, tmbstr buffer, uint* buflen )
{
uint outenc = cfg( doc, TidyOutCharEncoding );
uint nl = cfg( doc, TidyNewline );
TidyBuffer outbuf = {0};
StreamOut* out = TY_(BufferOutput)( &outbuf, outenc, nl );
int status = tidyDocSaveStream( doc, out );
if ( outbuf.size > *buflen )
status = -ENOMEM;
else
memcpy( buffer, outbuf.bp, outbuf.size );
*buflen = outbuf.size;
tidyBufFree( &outbuf );
MemFree( out );
return status;
}
void CCaHtmlParse::__GetFlightStartTime(CStringA & strFlightStartTime, const TidyDoc & tdoc, const TidyNode & tdNode)
{
CStringA straRet;
strFlightStartTime = "";
TidyBuffer text = {0};
tidyBufInit(&text);
TidyNodeType type = tidyNodeGetType(tdNode);
tidyNodeGetText(tdoc, tdNode, &text);
straRet.Format("%s",text.bp);
CStringA straKey("\">");
int iStartPos = straRet.Find(straKey);
int iEndPos = straRet.Find("</strong>");
iStartPos = iStartPos+straKey.GetLength();
strFlightStartTime = straRet.Mid(iStartPos, iEndPos-iStartPos);
tidyBufFree(&text);
}
static zend_object *tidy_object_new(zend_class_entry *class_type, zend_object_handlers *handlers, tidy_obj_type objtype)
{
PHPTidyObj *intern;
intern = ecalloc(1, sizeof(PHPTidyObj) + zend_object_properties_size(class_type));
zend_object_std_init(&intern->std, class_type);
object_properties_init(&intern->std, class_type);
switch(objtype) {
case is_node:
break;
case is_doc:
intern->ptdoc = emalloc(sizeof(PHPTidyDoc));
intern->ptdoc->doc = tidyCreate();
intern->ptdoc->ref_count = 1;
intern->ptdoc->initialized = 0;
intern->ptdoc->errbuf = emalloc(sizeof(TidyBuffer));
tidyBufInit(intern->ptdoc->errbuf);
if (tidySetErrorBuffer(intern->ptdoc->doc, intern->ptdoc->errbuf) != 0) {
tidyBufFree(intern->ptdoc->errbuf);
efree(intern->ptdoc->errbuf);
tidyRelease(intern->ptdoc->doc);
efree(intern->ptdoc);
efree(intern);
php_error_docref(NULL, E_ERROR, "Could not set Tidy error buffer");
}
tidyOptSetBool(intern->ptdoc->doc, TidyForceOutput, yes);
tidyOptSetBool(intern->ptdoc->doc, TidyMark, no);
TIDY_SET_DEFAULT_CONFIG(intern->ptdoc->doc);
tidy_add_default_properties(intern, is_doc);
break;
}
intern->std.handlers = handlers;
return &intern->std;
}
