void linkwordadd(struct pagewordsFormat *pagewords, char word[]) {
int wordlLength;
wordlLength = strlen(word);
convert_to_lowercase((unsigned char *)word);
//-1 så vi ikke appender alt på slutten av siste linken hvis vi har mr en IndexerMaxLinks
if ((IndexerMaxLinks)> (*pagewords).nrOfOutLinks) {
if (!(*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].good) {
//dårlig link
}
else if (((*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktextlen + wordlLength +2 ) > sizeof((*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktext)) {
//har ikke mer plass til link tekst
}
else {
//må ha -1 her da det er den nåverende vi jobber på, ikke den neste
strcpy((*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktext + (*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktextlen,word);
(*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktextlen += wordlLength;
strcpy((*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktext + (*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktextlen," ");
(*pagewords).outlinks[(*pagewords).nrOfOutLinks -1].linktextlen += 1;
}
}
else {
#ifdef DEBUG
printf("note: to many links. Has %i sow far\n",(*pagewords).nrOfOutLinks);
#endif
}
}
static int make_key(char *buf, size_t buflen, enum keytype type,
struct mlfi_priv *priv, const char *rcpt)
{
int r = 0;
switch (type) {
case TO:
r = snprintf(buf, buflen, "%s", rcpt);
break;
case TO_IP:
r = snprintf(buf, buflen, "%s:%s", rcpt, priv->priv_ip);
break;
case TO_IP_FROM:
r = snprintf(buf, buflen, "%s:%s:%s", rcpt, priv->priv_ip,
priv->priv_from);
break;
case BOUNCE_TO:
r = snprintf(buf, buflen, "%s:<>", rcpt);
break;
case BOUNCE_TO_IP:
r = snprintf(buf, buflen, "%s:%s:<>", rcpt, priv->priv_ip);
break;
}
if (r >= buflen) {
return 0;
}
convert_to_lowercase (buf, r);
return r;
}
void attribadd(struct IndexerRes_attrib *attrib, char word[]) {
convert_to_lowercase((unsigned char *)word);
#ifdef DEBUG
printf("attribadd: got \"%s\"\n",word);
#endif
if ((*attrib).attribnr > maxAttribForPage){
#ifdef DEBUG
printf("more than maxAttribForPage words\n");
#endif
}
else {
(*attrib).attrib[(*attrib).attribnr].WordID = crc32boitho(word);
(*attrib).attrib[(*attrib).attribnr].position = 0;
++(*attrib).attribnr;
}
}
void acladd(struct IndexerRes_acls *acl, char word[]) {
convert_to_lowercase((unsigned char *)word);
#ifdef DEBUG
printf("acladd: got \"%s\"\n",word);
#endif
if ((*acl).aclnr > maxAclForPage){
#ifdef DEBUG
printf("mor then maxAclForPage words\n");
#endif
}
else {
(*acl).acls[(*acl).aclnr].WordID = crc32boitho(word);
(*acl).acls[(*acl).aclnr].position = 0;
++(*acl).aclnr;
}
}
void
suggest_1(char *host, char *arg, char *user, char *collection)
{
CLIENT *clnt;
numbest_res *result_1;
struct senddata args;
char first[1024];
char suggeston[1024];
char **wordlist;
int splitn;
splitn = split(arg, " ", &wordlist);
if (splitn == 0)
return;
args.word = wordlist[splitn-1];
convert_to_lowercase(args.word);
args.user = user;
args.collection = (collection ? collection : "");
#ifndef DEBUG
clnt = clnt_create (host, SUGGEST, SUGGESTVERS, "udp");
if (clnt == NULL) {
clnt_pcreateerror (host);
exit (1);
}
#endif /* DEBUG */
/* XXX: set lower timeout */
result_1 = get_best_results_2(&args, clnt);
if (!result_1) {
#if 1
//clnt_perror (clnt, "call failed");
#endif
}
else {
if (result_1->_errno == 0) {
namelist nl;
int i;
if (strlen(args.word) > 0) {
for (nl = result_1->numbest_res_u.list;
nl != NULL;
nl = nl->next) {
for (i = 0; i < splitn-1; i++)
printf("%s ", wordlist[i]);
printf("%s\n", nl->name);
}
} else {
for (i = 0; i < splitn-1; i++)
printf("%s ", wordlist[i]);
puts("");
}
}
}
FreeSplitList(wordlist);
#ifndef DEBUG
clnt_destroy (clnt);
#endif /* DEBUG */
}
void fn( char* word, int pos, enum parsed_unit pu, enum parsed_unit_flag puf, void* pagewords )
{
#ifdef DEBUG
printf("\t%s (%i) ", word, pos);
printf("type %i ",pu);
#endif
switch (pu)
{
case pu_word:
switch (puf)
{
case puf_none:
printf(" none");
break;
case puf_title:
printf(" +title");
break;
case puf_h1:
printf(" +h1");
break;
case puf_h2:
printf(" +h2");
break;
case puf_h3:
printf(" +h3");
break;
case puf_h4:
printf(" +h4");
break;
case puf_h5:
printf(" +h5");
break;
case puf_h6:
printf(" +h6");
break;
}
convert_to_lowercase(word);
printf("[word] is now %s (crc32 %u, pos %i)", word, crc32boitho(word), pos);
break;
case pu_linkword:
printf("[linkword]");
break;
case pu_link:
printf("[link]");
break;
case pu_baselink:
printf("[baselink]");
break;
case pu_meta_keywords:
printf("[meta keywords]");
break;
case pu_meta_description:
printf("[meta description]");
break;
case pu_meta_author:
printf("[meta author]");
break;
default:
printf("[...]");
}
printf("\n");
}
int main (int argc, char *argv[]) {
int lotNr;
int i;
unsigned int DocID;
char text[50];
unsigned int radress;
unsigned int rsize;
char **Data;
int Count, TokCount;
unsigned short hits;
unsigned long WordID;
int bucket;
int y;
int nr;
FILE *revindexFilesHa[NrOfDataDirectorys];
unsigned char lang;
FILE *FH;
unsigned int DocIDPlace;
int *nrOfLinkWordsToDocID = malloc(sizeof(int) * NrofDocIDsInLot);
for (i=0;i<NrofDocIDsInLot;i++) {
//begynner på 2000 så det skal være lett og skille de visuelt fra andre hits
nrOfLinkWordsToDocID[i] = 2000;
}
//tester for at vi har fåt hvilken lot vi skal bruke
if (argc < 3) {
printf("Usage: ./anchorread lotnr subname\n\n");
exit(1);
}
lotNr = atoi(argv[1]);
char *subname = argv[2];
if ( (FH = lotOpenFileNoCasheByLotNr(lotNr,"anchors","rb", 's',subname)) == NULL) {
printf("lot dont have a anchors file\n");
exit(1);
}
fclose(FH);
revindexFilesOpenLocal(revindexFilesHa,lotNr,"Anchor","wb",subname);
//int anchorGetNext (int LotNr,unsigned int *DocID,char *text,unsigned int *radress,unsigned int *rsize)
while (anchorGetNext(lotNr,&DocID,text,sizeof(text),&radress,&rsize,subname) ) {
DocIDPlace = (DocID - LotDocIDOfset(rLotForDOCid(DocID)));
++nrOfLinkWordsToDocID[DocIDPlace];
convert_to_lowercase((unsigned char *)text);
#ifdef DEBUG
if (DocID == 4999999) {
printf("DocID %i, text: \"%s\", DocIDPlace %i, nrOfLinkWordsToDocID %i\n",DocID,text,DocIDPlace,nrOfLinkWordsToDocID[DocIDPlace]);
}
#endif
if ((TokCount = split(text, " ", &Data)) == -1) {
printf("canæt splitt \"%s\"\n",text);
}
//for (i=(TokCount-1);i>=0;i--) {
i=0;
while (Data[i] != NULL) {
/*
if (nrOfLinkWordsToDocID[DocIDPlace] > 65505) {
#ifdef DEBUG
if (DocID == 4999999) {
printf("reach max nr of words for DocID %u. Hav %i+ words\n",DocID,nrOfLinkWordsToDocID[DocIDPlace]);
}
#endif
break;
}
*/
if (Data[i][0] == '\0') {
#ifdef DEBUG
if (DocID == 4999999) {
printf("emty data element\n");
}
#endif
}
else if (strcmp(Data[i],"www") == 0) {
#ifdef DEBUG
if (DocID == 4999999) {
printf("www\n");
}
#endif
++nrOfLinkWordsToDocID[DocIDPlace];
}
else if (isStoppWord(Data[i])) {
#ifdef DEBUG
if (DocID == 4999999) {
printf("stopword \"%s\"\n",Data[i]);
}
#endif
//++nrOfLinkWordsToDocID[DocIDPlace];
}
else {
#ifdef DEBUG
if (DocID == 4999999) {
printf("\t\"%s\" %i\n",Data[i],nrOfLinkWordsToDocID[DocIDPlace]);
}
#endif
WordID = crc32boitho(Data[i]);
if (WordID == 0) {
printf("got 0 as word id for \"%s\". Somthing may be wrong.\n",Data[i]);
}
bucket = WordID % NrOfDataDirectorys;
if (nrOfLinkWordsToDocID[DocIDPlace] > 65535) {
hits = 65535;
}
else {
hits = nrOfLinkWordsToDocID[DocIDPlace];
}
#ifdef DEBUG
if (DocID == 4999999) {
printf("\thits %i: \"%s\": %hu, bucket %i\n",i,Data[i],hits,bucket);
}
#endif
if (fwrite(&DocID,sizeof(unsigned int),1,revindexFilesHa[bucket]) != 1) {
perror("fwrite DocID");
}
//runarb: 13 mai 2007. vi har byttet til å bruke et tal for språk.
//burde da dette fra DocumentIndex hvis det finnes, men lagres ikke der
//må si i IndexRes på hvordan vi gjør det der
//fprintf(revindexFilesHa[bucket],"aa ");
lang = 0;
nr = 1;
if(fwrite(&lang,sizeof(unsigned char),1,revindexFilesHa[bucket]) != 1) {
perror("fwrite lang");
}
if(fwrite(&WordID,sizeof(unsigned long),1,revindexFilesHa[bucket]) != 1) {
perror("fwrite WordID");
}
if(fwrite(&nr,sizeof(unsigned long),1,revindexFilesHa[bucket]) != 1) {
perror("fwrite nr");
}
if(fwrite(&hits,sizeof(unsigned short),1,revindexFilesHa[bucket]) != 1) {
perror("fwrite hits");
}
++nrOfLinkWordsToDocID[DocIDPlace];
}
++i;
}
FreeSplitList(Data);
#ifdef DEBUG
if (DocID == 4999999) {
printf("\n");
}
#endif
}
free(nrOfLinkWordsToDocID);
}
void wordsAdd(struct pagewordsFormatPartFormat *wordsPart, char word[],enum parsed_unit_flag puf) {
int wordlLength;
int wordTypeadd = 0;
if (wordsPart->nr > maxWordForPage){
#ifdef DEBUG
printf("mor then maxWordForPage words\n");
#endif
}
else {
switch (puf)
{
case puf_none:
//printf(" +p");
wordTypeadd=1000;break;
case puf_title:
//printf(" +title");
wordTypeadd=100; break;
case puf_h1:
//printf(" +h1");
wordTypeadd=500; break;
case puf_h2:
//printf(" +h2");
wordTypeadd=500; break;
case puf_h3:
//printf(" +h3");
wordTypeadd=500; break;
case puf_h4:
//printf(" +h4");
wordTypeadd=500; break;
case puf_h5:
//printf(" +h5");
wordTypeadd=500; break;
case puf_h6:
//printf(" +h6");
wordTypeadd=500; break;
default:
printf(" no catsh\n"); break;
}
wordlLength = strlen(word);
//gjør om til små bokstaver
convert_to_lowercase((unsigned char *)word);
#ifdef PRESERVE_WORDS
strcpy(wordsPart->words[wordsPart->nr].word,word);
#endif
wordsPart->words[wordsPart->nr].WordID = crc32boitho(word);
#ifdef DEBUG
printf(" (crc %s -> %u) ",word,wordsPart->words[wordsPart->nr].WordID);
#endif
wordsPart->words[wordsPart->nr].position = (wordsPart->nextPosition + wordTypeadd);
// må ha en index posisjon her. Slik at vi kan finne ord før og etter. Posisjon er kodet
wordsPart->words[wordsPart->nr].unsortetIndexPosition = wordsPart->nr;
++wordsPart->nextPosition;
//printf("%s : %u\n",word,wordsPart->words[wordsPart->nr]);
++wordsPart->nr;
}
}
