X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fextract.c;h=4309b60e2a99673b0ce50f5b6997affdda740926;hb=78e20a4e6a704e81bddb2ac37caf83f5a670041d;hp=ebd04ac0c4bc7e590b79b2f0aa183473916fdfd5;hpb=3c5f6226f97612c0d6ac40591f600587c5ffa858;p=idzebra-moved-to-github.git diff --git a/index/extract.c b/index/extract.c index ebd04ac..4309b60 100644 --- a/index/extract.c +++ b/index/extract.c @@ -4,7 +4,25 @@ * Sebastian Hammer, Adam Dickmeiss * * $Log: extract.c,v $ - * Revision 1.90 1999-02-02 14:50:52 adam + * Revision 1.96 1999-05-26 07:49:13 adam + * C++ compilation. + * + * Revision 1.95 1999/05/21 12:00:17 adam + * Better diagnostics for extraction process. + * + * Revision 1.94 1999/05/20 12:57:18 adam + * Implemented TCL filter. Updated recctrl system. + * + * Revision 1.93 1999/05/15 14:36:38 adam + * Updated dictionary. Implemented "compression" of dictionary. + * + * Revision 1.92 1999/03/09 16:27:49 adam + * More work on SDRKit integration. + * + * Revision 1.91 1999/02/12 13:29:22 adam + * Implemented position-flag for registers. + * + * Revision 1.90 1999/02/02 14:50:52 adam * Updated WIN32 code specific sections. Changed header. * * Revision 1.89 1998/10/28 10:54:38 adam @@ -392,14 +410,14 @@ int key_open (struct recordGroup *rGroup, int mem) mem = atoi(res_get_def (common_resource, "memMax", "4"))*1024*1024; if (mem < 50000) mem = 50000; - key_buf = xmalloc (mem); + key_buf = (char **) xmalloc (mem); ptr_top = mem/sizeof(char*); ptr_i = 0; key_buf_used = 0; key_file_no = 0; - if (!(matchDict = dict_open (bfs, GMATCH_DICT, 50, rw))) + if (!(matchDict = dict_open (bfs, GMATCH_DICT, 50, rw, 0))) { logf (LOG_FATAL, "dict_open fail of %s", GMATCH_DICT); return -1; @@ -426,6 +444,7 @@ int key_open (struct recordGroup *rGroup, int mem) struct encode_info { int sysno; int seqno; + int cmd; char buf[768]; }; @@ -433,6 +452,7 @@ void encode_key_init (struct encode_info *i) { i->sysno = 0; i->seqno = 0; + i->cmd = -1; } char *encode_key_int (int d, char *bp) @@ -474,8 +494,11 @@ void encode_key_write (char *k, struct encode_info *i, FILE *outf) i->sysno = key.sysno; i->seqno = 0; } + else if (!i->seqno && !key.seqno && i->cmd == *k) + return; bp = encode_key_int (key.seqno - i->seqno, bp); i->seqno = key.seqno; + i->cmd = *k; if (fwrite (i->buf, bp - i->buf, 1, outf) != 1) { logf (LOG_FATAL|LOG_ERRNO, "fwrite"); @@ -646,7 +669,7 @@ static void addIndexString (RecWord *p, const char *string, int length) { char *b; - b = xmalloc (reckeys.buf_max += 128000); + b = (char *) xmalloc (reckeys.buf_max += 128000); if (reckeys.buf_used > 0) memcpy (b, reckeys.buf, reckeys.buf_used); xfree (reckeys.buf); @@ -696,7 +719,8 @@ static void addIndexString (RecWord *p, const char *string, int length) dst += sizeof(*pseqno); } reckeys.buf_used = dst - reckeys.buf; - (*pseqno)++; + if (*pseqno) + (*pseqno)++; } static void addSortString (RecWord *p, const char *string, int length) @@ -707,11 +731,11 @@ static void addSortString (RecWord *p, const char *string, int length) if (sk->attrSet == p->attrSet && sk->attrUse == p->attrUse) return; - sk = xmalloc (sizeof(*sk)); + sk = (struct sortKey *) xmalloc (sizeof(*sk)); sk->next = sortKeys; sortKeys = sk; - sk->string = xmalloc (length); + sk->string = (char *) xmalloc (length); sk->length = length; memcpy (sk->string, string, length); @@ -975,7 +999,8 @@ struct file_read_info { static struct file_read_info *file_read_start (int fd) { - struct file_read_info *fi = xmalloc (sizeof(*fi)); + struct file_read_info *fi = (struct file_read_info *) + xmalloc (sizeof(*fi)); fi->fd = fd; fi->file_max = 0; @@ -992,7 +1017,7 @@ static void file_read_stop (struct file_read_info *fi) static off_t file_seek (void *handle, off_t offset) { - struct file_read_info *p = handle; + struct file_read_info *p = (struct file_read_info *) handle; p->file_offset = offset; if (p->sdrbuf) return offset; @@ -1001,13 +1026,13 @@ static off_t file_seek (void *handle, off_t offset) static off_t file_tell (void *handle) { - struct file_read_info *p = handle; + struct file_read_info *p = (struct file_read_info *) handle; return p->file_offset; } static int file_read (void *handle, char *buf, size_t count) { - struct file_read_info *p = handle; + struct file_read_info *p = (struct file_read_info *) handle; int fd = p->fd; int r; if (p->sdrbuf) @@ -1031,7 +1056,7 @@ static int file_read (void *handle, char *buf, size_t count) static void file_begin (void *handle) { - struct file_read_info *p = handle; + struct file_read_info *p = (struct file_read_info *) handle; p->file_offset = p->file_moffset; if (!p->sdrbuf && p->file_moffset) @@ -1041,7 +1066,7 @@ static void file_begin (void *handle) static void file_end (void *handle, off_t offset) { - struct file_read_info *p = handle; + struct file_read_info *p = (struct file_read_info *) handle; assert (p->file_more == 0); p->file_more = 1; @@ -1205,7 +1230,7 @@ struct recordLogInfo { static void recordLogPreamble (int level, const char *msg, void *info) { - struct recordLogInfo *p = info; + struct recordLogInfo *p = (struct recordLogInfo *) info; FILE *outf = log_file (); if (level & LOG_LOG) @@ -1223,7 +1248,7 @@ void addSchema (struct recExtractCtrl *p, Odr_oid *oid) static int recordExtract (SYSNO *sysno, const char *fname, struct recordGroup *rGroup, int deleteFlag, struct file_read_info *fi, - RecType recType, char *subType) + RecType recType, char *subType, void *clientData) { RecordAttr *recordAttr; int r; @@ -1258,7 +1283,12 @@ static int recordExtract (SYSNO *sysno, const char *fname, extractCtrl.addSchema = addSchema; extractCtrl.dh = rGroup->dh; for (i = 0; i<256; i++) - extractCtrl.seqno[i] = 0; + { + if (zebra_maps_is_positioned(rGroup->zebra_maps, i)) + extractCtrl.seqno[i] = 1; + else + extractCtrl.seqno[i] = 0; + } extractCtrl.zebra_maps = rGroup->zebra_maps; extractCtrl.flagShowRecords = !rGroup->flagRw; @@ -1270,18 +1300,20 @@ static int recordExtract (SYSNO *sysno, const char *fname, logInfo.rGroup = rGroup; log_event_start (recordLogPreamble, &logInfo); - r = (*recType->extract)(&extractCtrl); + r = (*recType->extract)(clientData, &extractCtrl); log_event_start (NULL, NULL); - if (r) - { + if (r == RECCTRL_EXTRACT_EOF) + return 0; + else if (r == RECCTRL_EXTRACT_ERROR) + { /* error occured during extraction ... */ if (rGroup->flagRw && records_processed < rGroup->fileVerboseLimit) { - logf (LOG_WARN, "fail %s %s %ld code = %d", rGroup->recordType, - fname, (long) recordOffset, r); + logf (LOG_WARN, "fail %s %s %ld", rGroup->recordType, + fname, (long) recordOffset); } return 0; } @@ -1463,7 +1495,8 @@ static int recordExtract (SYSNO *sysno, const char *fname, if (rGroup->flagStoreData == 1) { rec->size[recInfo_storeData] = recordAttr->recordSize; - rec->info[recInfo_storeData] = xmalloc (recordAttr->recordSize); + rec->info[recInfo_storeData] = (char *) + xmalloc (recordAttr->recordSize); if (lseek (fi->fd, recordOffset, SEEK_SET) < 0) { logf (LOG_ERRNO|LOG_FATAL, "seek to %ld in %s", @@ -1509,6 +1542,7 @@ int fileExtract (SYSNO *sysno, const char *fname, struct recordGroup rGroupM; struct recordGroup *rGroup = &rGroupM; struct file_read_info *fi; + void *clientData; memcpy (rGroup, rGroupP, sizeof(*rGroupP)); @@ -1548,7 +1582,8 @@ int fileExtract (SYSNO *sysno, const char *fname, if (!*rGroup->recordType) return 0; if (!(recType = - recType_byName (rGroup->recTypes, rGroup->recordType, subType))) + recType_byName (rGroup->recTypes, rGroup->recordType, subType, + &clientData))) { logf (LOG_WARN, "No such record type: %s", rGroup->recordType); return 0; @@ -1620,14 +1655,17 @@ int fileExtract (SYSNO *sysno, const char *fname, rGroup->flagStoreKeys = 0; #if ZEBRASDR - if (1) + if (rGroup->useSDR) { ZebraSdrHandle h; char xname[128], *xp; strncpy (xname, fname, 127); - if ((xp = strchr (xname, '.'))) - *xp = '\0'; + if (!(xp = strchr (xname, '.'))) + return 0; + *xp = '\0'; + if (strcmp (xp+1, "sdr.bits")) + return 0; h = zebraSdr_open (xname); if (!h) @@ -1643,9 +1681,8 @@ int fileExtract (SYSNO *sysno, const char *fname, segmentno = zebraSdr_segment (h, 0); sprintf (sdr_name, "%%%s.%d", xname, segmentno); - logf (LOG_LOG, "SDR: %s", sdr_name); -#if 1 +#if 0 if (segmentno > 20) break; #endif @@ -1685,7 +1722,7 @@ int fileExtract (SYSNO *sysno, const char *fname, { file_begin (fi); r = recordExtract (sysno, fname, rGroup, deleteFlag, fi, - recType, subType); + recType, subType, clientData); } while (r && !sysno && fi->file_more); file_read_stop (fi); if (fd != -1)