void DpsAppendTarget(DPS_AGENT *Indexer, const char *url, const char *lang, const int hops, int parent) {
DPS_DOCUMENT *Doc, *Save;
size_t i;
TRACE_IN(Indexer, "AppendTarget");
DPS_GETLOCK(Indexer, DPS_LOCK_THREAD);
DPS_GETLOCK(Indexer, DPS_LOCK_CONF);
if (Indexer->Conf->Targets.num_rows > 0) {
for (i = Indexer->Conf->Targets.num_rows - 1; i > 0; i--) {
Doc = &Indexer->Conf->Targets.Doc[i];
if ((strcasecmp(DpsVarListFindStr(&Doc->Sections, "URL", ""), url) == 0)
&& (strcmp(DpsVarListFindStr(&Doc->RequestHeaders, "Accept-Language", ""), lang) == 0)) {
DPS_RELEASELOCK(Indexer, DPS_LOCK_CONF);
DPS_RELEASELOCK(Indexer, DPS_LOCK_THREAD);
TRACE_OUT(Indexer);
return;
}
}
}
if ((Indexer->Conf->Targets.Doc =
DpsRealloc(Save = Indexer->Conf->Targets.Doc, (Indexer->Conf->Targets.num_rows + 1) * sizeof(DPS_DOCUMENT))) == NULL) {
Indexer->Conf->Targets.Doc = Save;
DPS_RELEASELOCK(Indexer, DPS_LOCK_CONF);
DPS_RELEASELOCK(Indexer, DPS_LOCK_THREAD);
TRACE_OUT(Indexer);
return;
}
Doc = &Indexer->Conf->Targets.Doc[Indexer->Conf->Targets.num_rows];
DpsDocInit(Doc);
DpsVarListAddStr(&Doc->Sections, "URL", url);
DpsVarListAddInt(&Doc->Sections, "Hops", hops);
DpsVarListDel(&Doc->Sections, "URL_ID");
DpsVarListReplaceInt(&Doc->Sections, "Referrer-ID", parent);
if (*lang != '\0') DpsVarListAddStr(&Doc->RequestHeaders, "Accept-Language", lang);
if (DPS_OK == DpsURLAction(Indexer, Doc, DPS_URL_ACTION_FINDBYURL)) {
urlid_t url_id = DpsVarListFindInt(&Doc->Sections, "DP_ID", 0);
if (url_id != 0) Indexer->Conf->Targets.num_rows++;
else DpsDocFree(Doc);
}
/* fprintf(stderr, "-- AppandTarget: url:%s URL_ID:%d\n", url, DpsStrHash32(url));*/
DPS_RELEASELOCK(Indexer, DPS_LOCK_CONF);
DpsURLAction(Indexer, Doc, DPS_URL_ACTION_ADD);
DPS_RELEASELOCK(Indexer, DPS_LOCK_THREAD);
TRACE_OUT(Indexer);
return;
}
int DpsCloneListSearchd(DPS_AGENT *Indexer, DPS_DOCUMENT *Doc, DPS_RESULT *Res, DPS_DB *db) {
DPS_SEARCHD_PACKET_HEADER hdr;
ssize_t nsent,nrecv;
char *msg = NULL, *dinfo = NULL;
char *tok, *lt;
char buf[128];
int done = 0;
int rc = DPS_OK;
TRACE_IN(Indexer, "DpsCloneListSearchd");
dps_snprintf(buf, 128, "%s", DpsVarListFindStr(&Doc->Sections, "DP_ID", "0"));
hdr.cmd = DPS_SEARCHD_CMD_CLONES;
hdr.len = dps_strlen(buf);
nsent = DpsSearchdSendPacket(db->searchd, &hdr, buf);
while(!done){
nrecv = DpsRecvall(db->searchd, &hdr, sizeof(hdr), 360);
if(nrecv != sizeof(hdr)){
DpsLog(Indexer, DPS_LOG_ERROR, "Received incomplete header from searchd (%d bytes)", (int)nrecv);
TRACE_OUT(Indexer);
return(DPS_ERROR);
}else{
#ifdef DEBUG_SDP
DpsLog(Indexer, DPS_LOG_DEBUG, "Received header cmd=%d len=%d\n", hdr.cmd, hdr.len);
#endif
}
switch(hdr.cmd){
case DPS_SEARCHD_CMD_ERROR:
msg = (char*)DpsMalloc(hdr.len + 1);
if (msg == NULL) {
done=1;
break;
}
nrecv = DpsRecvall(db->searchd, msg, hdr.len, 360);
msg[(nrecv >= 0) ? nrecv : 0] = '\0';
sprintf(Indexer->Conf->errstr, "Searchd error: '%s'", msg);
rc = DPS_ERROR;
DPS_FREE(msg);
done = 1;
break;
case DPS_SEARCHD_CMD_DOCINFO:
dinfo = (char*)DpsMalloc(hdr.len + 1);
if (dinfo == NULL) {
done=1;
break;
}
nrecv = DpsRecvall(db->searchd, dinfo, hdr.len, 360);
dinfo[(nrecv >= 0) ? nrecv : 0] = '\0';
#ifdef DEBUG_SDP
DpsLog(Indexer, DPS_LOG_DEBUG, "Received DOCINFO size=%d buf=%s\n", hdr.len, dinfo);
#endif
if (strcasecmp(dinfo, "nocloneinfo") != 0) {
tok = dps_strtok_r(dinfo, "\r\n", <, NULL);
while(tok){
DPS_DOCUMENT *D;
size_t nd = Res->num_rows++;
Res->Doc = (DPS_DOCUMENT*)DpsRealloc(Res->Doc, (Res->num_rows + 1) * sizeof(DPS_DOCUMENT));
if (Res->Doc == NULL) {
sprintf(Indexer->Conf->errstr, "Realloc error");
rc = DPS_ERROR;
break;
}
D = &Res->Doc[nd];
DpsDocInit(D);
DpsDocFromTextBuf(D, tok);
tok = dps_strtok_r(NULL, "\r\n", <, NULL);
}
}
DPS_FREE(dinfo);
done = 1;
break;
default:
sprintf(Indexer->Conf->errstr, "Unknown searchd response: cmd=%d len=%d", hdr.cmd, hdr.len);
rc = DPS_ERROR;
done = 1;
break;
}
}
TRACE_OUT(Indexer);
return rc;
}
int __DPSCALL DpsResAddDocInfoSearchd(DPS_AGENT * query,DPS_DB *cl,DPS_RESULT * Res,size_t clnum){
DPS_SEARCHD_PACKET_HEADER hdr;
char * msg=NULL;
size_t i; /* num=0,curnum=0;*/
int done = 0;
ssize_t nsent,nrecv;
char * dinfo=NULL;
int rc=DPS_OK;
char *textbuf;
size_t dlen = 0;
TRACE_IN(query, "DpsResAddDocInfoSearchd");
if(!Res->num_rows) { TRACE_OUT(query); return(DPS_OK); }
for(i=0;i<Res->num_rows;i++){
size_t ulen;
size_t olen;
size_t nsec, r;
DPS_DOCUMENT *D=&Res->Doc[i];
r = (size_t) 's';
for(nsec = 0; nsec < D->Sections.Root[r].nvars; nsec++)
if (strcasecmp(D->Sections.Root[r].Var[nsec].name, "Score") == 0) D->Sections.Root[r].Var[nsec].section = 1;
#ifdef WITH_MULTIDBADDR
if (D->dbnum != cl->dbnum) continue;
#endif
textbuf = DpsDocToTextBuf(D, 1, 0);
if (textbuf == NULL) {TRACE_OUT(query); return DPS_ERROR;}
ulen = dps_strlen(textbuf)+2;
olen = dlen;
dlen = dlen + ulen;
dinfo = (char*)DpsRealloc(dinfo, dlen + 1);
if (dinfo == NULL) {
DpsFree(textbuf);
TRACE_OUT(query);
return DPS_ERROR;
}
dinfo[olen] = '\0';
sprintf(dinfo + olen, "%s\r\n", textbuf);
DpsFree(textbuf);
}
if (dinfo == NULL) {
TRACE_OUT(query);
return DPS_OK;
}
hdr.cmd=DPS_SEARCHD_CMD_DOCINFO;
hdr.len = dps_strlen(dinfo);
nsent = DpsSearchdSendPacket(cl->searchd, &hdr, dinfo);
#ifdef DEBUG_SDP
DpsLog(query, DPS_LOG_ERROR, "Sent DOCINFO size=%d buf=%s\n", hdr.len, dinfo);
#endif
while(!done){
char * tok, * lt;
nrecv = DpsRecvall(cl->searchd, &hdr, sizeof(hdr), 360);
if(nrecv!=sizeof(hdr)){
DpsLog(query, DPS_LOG_ERROR, "Received incomplete header from searchd (%d bytes, errno:%d)", (int)nrecv, errno);
TRACE_OUT(query);
return(DPS_ERROR);
}else{
#ifdef DEBUG_SDP
DpsLog(query, DPS_LOG_ERROR, "Received header cmd=%d len=%d\n",hdr.cmd,hdr.len);
#endif
}
switch(hdr.cmd){
case DPS_SEARCHD_CMD_ERROR:
msg=(char*)DpsMalloc(hdr.len+1);
if (msg == NULL) {
done = 1;
break;
}
nrecv = DpsRecvall(cl->searchd, msg, hdr.len, 360);
msg[(nrecv >= 0) ? nrecv : 0]='\0';
sprintf(query->Conf->errstr,"Searchd error: '%s'",msg);
rc=DPS_ERROR;
DPS_FREE(msg);
done=1;
break;
case DPS_SEARCHD_CMD_MESSAGE:
msg=(char*)DpsMalloc(hdr.len+1);
if (msg == NULL) {
done = 1;
break;
}
nrecv = DpsRecvall(cl->searchd, msg, hdr.len, 360);
msg[(nrecv >= 0) ? nrecv : 0]='\0';
#ifdef DEBUG_SDP
DpsLog(query, DPS_LOG_ERROR, "Message from searchd: '%s'\n",msg);
#endif
DPS_FREE(msg);
break;
case DPS_SEARCHD_CMD_DOCINFO:
dinfo = (char*)DpsRealloc(dinfo, hdr.len + 1);
if (dinfo == NULL) {
done=1;
break;
}
nrecv = DpsRecvall(cl->searchd, dinfo, hdr.len, 360);
dinfo[(nrecv > 0) ? nrecv : 0] = '\0';
#ifdef DEBUG_SDP
DpsLog(query, DPS_LOG_ERROR, "Received DOCINFO size=%d buf=%s\n",hdr.len,dinfo);
#endif
tok = dps_strtok_r(dinfo, "\r\n", <, NULL);
while(tok){
urlid_t Doc_url_id, Res_Doc_url_id;
DPS_DOCUMENT Doc;
DpsDocInit(&Doc);
DpsDocFromTextBuf(&Doc,tok);
Doc_url_id = (urlid_t)DpsVarListFindInt(&Doc.Sections, "DP_ID", 0);
for(i=0;i<Res->num_rows;i++){
#ifdef WITH_MULTIDBADDR
if (Res->Doc[i].dbnum != cl->dbnum) continue;
#endif
Res_Doc_url_id = (urlid_t)DpsVarListFindInt(&Res->Doc[i].Sections, "DP_ID", 0);
if (Res_Doc_url_id == Doc_url_id) {
DpsDocFromTextBuf(&Res->Doc[i], tok);
break;
}
}
tok = dps_strtok_r(NULL, "\r\n", <, NULL);
DpsDocFree(&Doc);
}
DPS_FREE(dinfo);
done=1;
break;
default:
sprintf(query->Conf->errstr,"Unknown searchd response: cmd=%d len=%d",hdr.cmd,hdr.len);
rc=DPS_ERROR;
done=1;
break;
}
}
TRACE_OUT(query);
return rc;
}
void DpsCookiesFind(DPS_AGENT *Indexer, DPS_SERVER *Server, DPS_DOCUMENT *Doc, const char *hostinfo) {
#ifdef HAVE_SQL
DPS_DSTR cookie;
DPS_COOKIES *Cookies = &Indexer->Cookies;
DPS_COOKIE *Coo;
size_t i, blen = dps_strlen(hostinfo), slen;
int have_no_cookies = DpsVarListFindInt(&Doc->Sections, "have_no_cookies", 1);
#ifdef WITH_PARANOIA
void *paran = DpsViolationEnter(paran);
#endif
TRACE_IN(Indexer, "DpsCookiesFind");
DpsDSTRInit(&cookie, 1024);
for(i = 0; i < Cookies->ncookies; i++) {
Coo = &Cookies->Cookie[i];
slen = dps_strlen(Coo->domain);
if (slen > blen) continue;
if (Coo->secure == 'y' && strcasecmp(Doc->CurURL.schema, "https")) continue;
if (strncasecmp(Coo->path, Doc->CurURL.path, dps_strlen(Coo->path))) continue;
if (strcasecmp(Coo->domain, hostinfo + (blen - slen))) continue;
if (Coo->from_config != 1) have_no_cookies = 0;
if (Coo->name[0] == '\0' && Coo->value[0] == '\0') continue;
if (cookie.data_size)
DpsDSTRAppend(&cookie, "; ", 2);
DpsDSTRAppendStr(&cookie, Coo->name);
DpsDSTRAppend(&cookie, "=", 1);
DpsDSTRAppendStr(&cookie, Coo->value);
}
if (have_no_cookies) {
char buf[2*PATH_MAX];
dpshash32_t url_id;
DPS_DB *db;
DPS_SQLRES Res;
size_t rows;
int rc;
if (Server != NULL) {
char *PingData = DpsVarListFindStr(&Server->Vars, "AuthPing", NULL);
if (PingData != NULL) {
char *AuthPing = DpsStrdup(DpsTrim(PingData, " \t\r\n"));
int method = DPS_METHOD_GET;
dps_base64_decode(AuthPing, PingData, dps_strlen(PingData));
if (!strncasecmp(AuthPing, "GET", 3)) {
method = DPS_METHOD_GET;
PingData = DpsTrim(AuthPing + 3, " \t\r\n");
} else if (!strncasecmp(AuthPing, "POST", 4)) {
method = DPS_METHOD_POST;
PingData = DpsTrim(AuthPing + 4, " \t\r\n");
} else {
DpsLog(Indexer, DPS_LOG_ERROR, "AuthPing should be GET or POST: %s", AuthPing);
PingData = NULL;
}
if (PingData != NULL) {
size_t size = dps_strlen(PingData);
{
char PingURL[size + 2];
char PingBody[size];
DPS_DOCUMENT *rDoc;
int result;
rDoc = DpsDocInit(NULL);
DpsSpiderParamInit(&rDoc->Spider);
DpsVarList2Doc(rDoc, Server);
rDoc->Buf.max_size = (size_t)DpsVarListFindInt(&Indexer->Vars, "MaxDocSize", DPS_MAXDOCSIZE);
rDoc->Buf.allocated_size = DPS_NET_BUF_SIZE;
if ((rDoc->Buf.buf = (char*)DpsMalloc(rDoc->Buf.allocated_size + 1)) == NULL) {
DpsDocFree(rDoc);
TRACE_OUT(Indexer);
return;
}
rDoc->Buf.buf[0]='\0';
rDoc->subdoc = Indexer->Flags.SubDocLevel + 1;
#if 1
dps_snprintf(buf, sizeof(buf), "%s://%s/", DPS_NULL2EMPTY(Doc->CurURL.schema), DPS_NULL2EMPTY(Doc->CurURL.hostinfo));
DpsVarListReplaceStr(&rDoc->Sections, "URL", buf);
DpsURLParse(&rDoc->CurURL, buf);
DpsLog(Indexer, DPS_LOG_INFO, "HOME: %s", buf);
rDoc->method = DPS_METHOD_HEAD;
/* DpsVarListFree(&rDoc->RequestHeaders);*/
if (Doc != NULL) {
DpsVarListReplaceLst(&rDoc->RequestHeaders, &Doc->RequestHeaders, NULL, "*");
}
DpsVarListReplaceStr(&rDoc->Sections, "have_no_cookies", "0");
DpsDocAddDocExtraHeaders(Indexer, Server, rDoc);
DpsDocAddConfExtraHeaders(Indexer->Conf, rDoc);
DpsVarListReplaceLst(&rDoc->Sections, &Server->Vars, NULL, "*");
DpsDocAddServExtraHeaders(Server, rDoc);
DpsVarListLog(Indexer, &rDoc->RequestHeaders, DPS_LOG_DEBUG, "HOME.Request");
if (Doc == NULL || Indexer->Flags.cmd == DPS_IND_FILTER) {
DpsDocLookupConn(Indexer, rDoc);
} else {
DPS_FREE(rDoc->connp.connp);
rDoc->connp = Doc->connp;
}
result = DpsGetURL(Indexer, rDoc, NULL); /* Just get headers from the home as we need only Cookies from it */
DpsDocProcessResponseHeaders(Indexer, rDoc);
DpsVarListLog(Indexer, &rDoc->Sections, DPS_LOG_DEBUG, "HOME.Response");
#endif
sscanf(PingData, "%s %s", PingURL, PingBody);
if (rDoc->method == DPS_METHOD_GET) {
dps_strcat(PingURL, "?");
dps_strcat(PingURL, PingBody);
} else {
DpsVarListReplaceStr(&rDoc->Sections, "body", PingBody);
}
DpsVarListReplaceStr(&rDoc->Sections, "URL", PingURL);
DpsURLParse(&rDoc->CurURL, PingURL);
DpsLog(Indexer, DPS_LOG_INFO, "AUTH.PING: %s", PingURL);
rDoc->method = method;
DpsVarListFree(&rDoc->RequestHeaders);
DpsVarListReplaceStr(&rDoc->Sections, "have_no_cookies", "0");
DpsDocAddDocExtraHeaders(Indexer, Server, rDoc);
DpsDocAddConfExtraHeaders(Indexer->Conf, rDoc);
DpsVarListReplaceLst(&rDoc->Sections, &Server->Vars, NULL, "*");
DpsDocAddServExtraHeaders(Server, rDoc);
if (method == DPS_METHOD_POST) {
dps_snprintf(buf, sizeof(buf), "application/x-www-form-urlencoded; charset=%s", DpsVarListFindStr(&Indexer->Conf->Vars, "LocalCharset", "iso-8859-1"));
DpsVarListReplaceStr(&rDoc->RequestHeaders, "Content-Type", buf);
dps_snprintf(buf, sizeof(buf), "%d", dps_strlen(PingBody));
DpsVarListReplaceStr(&rDoc->RequestHeaders, "Content-Length", buf);
}
DpsVarListLog(Indexer, &rDoc->RequestHeaders, DPS_LOG_DEBUG, "AUTHPING.Request");
#if 0
if (Doc == NULL || Indexer->Flags.cmd == DPS_IND_FILTER) {
DpsDocLookupConn(Indexer, rDoc);
} else {
DPS_FREE(rDoc->connp.connp);
rDoc->connp = Doc->connp;
}
#endif
result = DpsGetURL(Indexer, rDoc, NULL); /* Just get it as we need only Cookies from the headers */
DpsDocProcessResponseHeaders(Indexer, rDoc);
DpsVarListDel(&rDoc->Sections, "body");
DpsVarListLog(Indexer, &rDoc->Sections, DPS_LOG_DEBUG, "AUTHPING.Response");
if (Doc != NULL) bzero(&rDoc->connp, sizeof(rDoc->connp));
DpsDocFree(rDoc);
}
}
DpsFree(AuthPing);
}
}
while(hostinfo != NULL) {
url_id = DpsStrHash32(hostinfo);
DpsSQLResInit(&Res);
dps_snprintf(buf, sizeof(buf), "SELECT name,value,path,secure FROM cookies WHERE domain='%s'", hostinfo);
if (Indexer->flags & DPS_FLAG_UNOCON) {
DPS_GETLOCK(Indexer, DPS_LOCK_DB);
db = Indexer->Conf->dbl.db[url_id % Indexer->Conf->dbl.nitems];
} else {
db = Indexer->dbl.db[url_id % Indexer->dbl.nitems];
}
if(DPS_OK == (rc = DpsSQLQuery(db, &Res, buf))) {
rows = DpsSQLNumRows(&Res);
for(i = 0; i < rows; i++) {
DpsCookiesAdd(Indexer, hostinfo, DpsSQLValue(&Res, i, 2), DpsSQLValue(&Res, i, 0), DpsSQLValue(&Res, i, 1),
*DpsSQLValue(&Res, i, 3), 0, 0, 0);
if (*DpsSQLValue(&Res, i, 3) == 'y' && strcasecmp(Doc->CurURL.schema, "https")) continue;
if (strncasecmp(DpsSQLValue(&Res, i, 2), Doc->CurURL.path, dps_strlen(DpsSQLValue(&Res, i, 2)))) continue;
if (cookie.data_size)
DpsDSTRAppend(&cookie, "; ", 2);
DpsDSTRAppendStr(&cookie, DpsSQLValue(&Res, i, 0));
DpsDSTRAppend(&cookie, "=", 1);
DpsDSTRAppendStr(&cookie, DpsSQLValue(&Res, i, 1));
}
if (rows == 0) {
DpsCookiesAdd(Indexer, hostinfo, "/", "", "", 'n', 0, 0, 0);
}
}
DpsSQLFree(&Res);
if (Indexer->flags & DPS_FLAG_UNOCON) {
DPS_RELEASELOCK(Indexer, DPS_LOCK_DB);
}
hostinfo = strchr(hostinfo, '.');
if (hostinfo != NULL) hostinfo++;
}
}
if (cookie.data_size) {
DpsVarListReplaceStr(&Doc->RequestHeaders, "Cookie", cookie.data);
}
DpsDSTRFree(&cookie);
#endif
TRACE_OUT(Indexer);
return;
}
