void KonqMFIcon::extractCard(DOM::Node node) {
QString name, value;
DOM::NodeList nodes = node.childNodes();
unsigned int n = nodes.length();
value += "BEGIN:VCARD\nVERSION:3.0\n";
for (unsigned int i = 0; i < n; ++i) {
DOM::Node node = nodes.item(i);
DOM::NamedNodeMap map = node.attributes();
for (unsigned int j = 0; j < map.length(); ++j) {
if (map.item(j).nodeName().string() != "class") {
continue;
}
QStringList l = QStringList::split(' ', map.item(j).nodeValue().string());
for (QStringList::ConstIterator it = l.begin(); it != l.end(); ++it) {
if (*it == "photo") {
} else if (*it == "adr") {
value += "ADR:" + extractAddress(node) + "\n";
} else if (*it == "tel") {
value += "TEL;TYPE=VOICE:" + textForNode(node) + "\n";
} else if (*it == "fn") {
name = textForNode(node);
value += "FN:" + name + "\n";
} else if (*it == "url") {
DOM::Node at = node.attributes().getNamedItem("href");
if (!at.isNull()) {
value += "URL:" + at.nodeValue().string().stripWhiteSpace() + "\n";
}
} else if (*it == "email") {
DOM::Node at = node.attributes().getNamedItem("href");
if (!at.isNull()) {
QString v = at.nodeValue().string();
if (v.startsWith("mailto:")) {
v = v.mid(7);
}
value += "EMAIL:" + v.stripWhiteSpace() + "\n";
}
} else if (*it == "org") {
value += "ORG:" + textForNode(node) + "\n";
}
}
}
}
if (!name.isEmpty()) {
value += "END:VCARD\n";
_cards.append(qMakePair(name, value));
}
}
int getAddress (char* url) {
char *stream, *text, *textHighlight, *lenstr;
char *tokens = (char*)malloc(DEF_BUFF_SIZE);
int *positions;// array to record position of each token in the text
cvector addressVector;
Address *adr;
long len;
int MAXLEN = 1805;
int EXTRA = 11;
/* 4 for field name "data", 1 for "=" */
int MAXINPUT = MAXLEN+EXTRA+2;
char input[MAXINPUT];
char* data = input, *p;
int rightOrWrong = -1, numRight, numTotal;
//char* domain_url;
int i;
http_setTimeout(8);//seconds
//fetch web page
int ret = httpFetch (url, &stream);
if (ret == -1) {
printf("%s\n",http_strerror());
exit(0);
}
//printf("ret: %d, strlen: %d\n",ret, strlen(stream));
assert(stream);
text= (char*)malloc(ret+2);
if (!text) {
printf("out of memory when convert text to tokens!\n");
exit(0);
}
strncpy(text, stream, ret);
//append a '\0' to the end of string to make sure it is end with two '\0' for flex to scan
*(text+ret) = '\0';
*(text+ret+1) = '\0';
free(stream);
/* convert text to tokens, remove tags
and convert back to string: tokens
and keep all positions in array "positions"
*/
convertToken2Text(text, tokens, &positions);
// get base domain of given url
//e.g. given http://www.google.com/address, return http://www.google.com to domain_url
/* domain_url = (char*)malloc(strlen(url)+1);
strcpy(domain_url, url);
for (i=strlen(url); i>0; i--) {
if (url[i] == '/') {
if (url[i-1] == '/' ) // is "//"
break;
else // not "//"
domain_url[i] = '\0';
}
}
printf ("<base href=\"%s%s\">\n", GEO_URL, domain_url);
free(domain_url);
*/
VectorNew (&addressVector, sizeof (Address),free_address, DEF_ADDRESS_PER_PAGE);
//extract address,
//get position from positions vector
//and save extracted address, position, country to addressVector
extractAddress(tokens, positions, &addressVector);
//display the parsed text
//printf("tokens: %s\n",tokens);
//printf("url: %s\n", url);
//printf("domain_url: %s\n", domain_url);
//output header
printf("<meta http-equiv=\"Content-Type\" content=\"text/html; charset=ISO-8859-1\">\n");
printf("<table border=1 width=100%%><tr><td><table border=1 bgcolor=#ffffff cellpadding=10 cellspacing=0 width=100%% color=#ffffff><tr><td>\n");
printf("<font face=arial,sans-serif color=black size=-1>\n");
printf("<b><a href='%s'>US, UK & Canadian Addresses</a> extracted by <a href='%s'>Geo Extractor</a> from web page</b> <a href='%s'>%s</a></font><br><br>\n",LIST_FILES_URL, HOME_PAGE,url,url);
//printf("%s,",textHighlight);
//display extracted address
//table header
printf("<table width=100%% border=0 cellpadding=0 cellspacing=0><tr><td bgcolor=#3366cc><img width=1 height=1 alt=''></td></tr></table>\n");
printf("<table width=100%% border=0 cellpadding=0 cellspacing=0 bgcolor=#e5ecf9><tr><td width=10></td><td bgcolor=#e5ecf9 nowrap><br>\n");
printf("<font face=arial,sans-serif color=black size=-1><b>\n");
for (i=0; i<addressVector.ItemsCount; i++) {
adr = (Address*)VectorNth(&addressVector,i);
printf("%s<br>\n", adr->address);
/*printf("%s, start: %d, end: %d<br>\n",adr->address, adr->start, adr->end);
for (j=adr->start; j<=adr->end; j++)
printf("%c",*(text+j));
printf("\n");
*/
}
printf("</b></font>\n");
printf("<br></td></tr></table>\n");
printf("<table width=100%% border=0 cellpadding=0 cellspacing=0><tr><td bgcolor=#3366cc><img width=1 height=1 alt=''></td></tr></table>\n");
textHighlight = (char*)malloc(DEF_BUFF_SIZE);
numRight=numTotal =addressVector.ItemsCount;
/* if there is a user post, we save the user input to get tagged data*/
lenstr = getenv("CONTENT_LENGTH");
if ( !(lenstr == NULL || sscanf(lenstr,"%ld",&len)!=1 || len > MAXLEN) ) {
tagAddress(text, textHighlight, &addressVector);
fgets(input, len+1, stdin);
URLdecode(input);
data = input+EXTRA;
//printf("posted: %s\n",data);
len = strlen("right");
if ( strncmp(data, "right", len)==0 ) {
rightOrWrong = 0; //set flag for right or wrong extraction
}
len = strlen("wrong");
if ( strncmp(data, "wrong", len)==0 ) {
rightOrWrong = 1; //user input "Wrong Extraction"
}
// get user input: numRight, which is number of correct extracted address
data += strlen("right") + strlen("&numRight=");
p = data;
while (*data++ !='&');
*data= '\0';
numRight = atoi(p);
//printf("numRight: %d\n", numRight);
// get user input numTotal, which is number of total address in the page
p = data+strlen("numTotal=");
numTotal = atoi(p);
//printf("numTotal: %d\n", numTotal);
if (rightOrWrong == 0) {
//printf("webpage saved to RIGHT folder\n");
saveTaggedText(url, text, textHighlight, rightOrWrong, numRight, addressVector.ItemsCount, numTotal);
}
if ( rightOrWrong == 1 ) {
//printf("webpage saved to WRONG folder\n");
saveTaggedText(url, text, textHighlight, rightOrWrong, numRight, addressVector.ItemsCount, numTotal);
}
//printf("tagged text: %s\n", textHighlight);
}
// give source text, and addressVector
//highlight all extracted address in the webpage
getHighlight(text, textHighlight, &addressVector);
/* if there is at least one address extracted, show user input to let user
judge where extraction is correct*/
if ( SHOW_COLLECT_DATA_INTERFACE) {
printf("<FORM ACTION=\"%s%s\" METHOD=\"POST\">\n", GEO_URL, url);
printf("<font face=arial,sans-serif color=black size=-1>\n");
printf("<P><input name=\"extraction\" type=\"radio\" value=\"right\" ");
if ((rightOrWrong == 0)||(rightOrWrong == -1)) //if no user input or user input: extracted address all correct
printf("checked");
printf("> All address extracted correctly<br>\n");
printf("<input name=\"extraction\" type=\"radio\" value=\"wrong\" ");
if (rightOrWrong == 1) //user input: extracted address all correct
printf("checked");
printf("> Not all addresses extracted correctly. \n");
printf("<input type=\"text\" name=\"numRight\" size=\"4\" value=\"%d\"> addresses extracted correctly from total <input type=\"text\" name=\"numTotal\" size=\"4\" value=\"%d\"> addresses<BR>\n", numRight, numTotal);
printf("<INPUT TYPE=\"SUBMIT\" VALUE=\"Save Webpage\"></font></FORM>\n");
//show google search
printf("<SCRIPT language=\"JavaScript\">function OnSubmitForm(){ document.g.action =\"%shttp://www.google.com/search?num=100&q=\"+document.g.q.value.replace(\" \",\"%%2B\");}</SCRIPT>\n", GEO_URL);
printf("<table border=0 align=right><tr><td>\n");
printf("<form action=\"\" method=\"post\" name=\"g\" onSubmit=\"return OnSubmitForm();\">\n");
printf("<input size=\"32\" name=\"q\">\n");
printf("<INPUT TYPE=\"SUBMIT\" name=\"Submit\" VALUE=\"Google\"></form>\n");
printf("</td></tr></table>\n");
}
printf("</td></tr></table></td></tr></table>\n");
//extract address from original html text
// extract_address(text);
// printf("Original <hr>%s",text);
printf("<hr>\n");
//printf("%s",textHighlight);
displayHtmlAbsoluteURL(textHighlight, url);
VectorDispose(&addressVector);
free (positions);
free (text);
free (tokens);
return 0;
}
