X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=index%2Fextract.c;h=5c17db509cdae90e723bcde6c98a4314dc063428;hp=9561c6e330d02e6a713f42e84d2112b1720e71ec;hb=5437b50633032595afe6f87dc0f989bc92a5aea8;hpb=2e4e9c6def27f1e1463dcb6f205fab6a98054f38 diff --git a/index/extract.c b/index/extract.c index 9561c6e..5c17db5 100644 --- a/index/extract.c +++ b/index/extract.c @@ -1,4 +1,4 @@ -/* $Id: extract.c,v 1.158 2004-08-04 08:35:23 adam Exp $ +/* $Id: extract.c,v 1.166 2004-11-19 10:26:56 heikki Exp $ Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004 Index Data Aps @@ -72,7 +72,8 @@ static void logRecord (ZebraHandle zh) ++zh->records_processed; if (!(zh->records_processed % 1000)) { - logf (LOG_LOG, "Records: %7d i/u/d %d/%d/%d", + yaz_log (YLOG_LOG, "Records: "ZINT_FORMAT" i/u/d " + ZINT_FORMAT"/"ZINT_FORMAT"/"ZINT_FORMAT, zh->records_processed, zh->records_inserted, zh->records_updated, zh->records_deleted); } @@ -92,8 +93,6 @@ static const char **searchRecordKey (ZebraHandle zh, struct recKeys *reckeys, int attrSetS, int attrUseS) { -#if IT_KEY_NEW -/* #error searchRecordKey does not work yet in this mode.. */ static const char *ws[32]; void *decode_handle = iscz1_start(); int off = 0; @@ -114,9 +113,9 @@ static const char **searchRecordKey (ZebraHandle zh, iscz1_decode(decode_handle, &dst, &src); assert(key.len < 4 && key.len > 2); - attrSet = key.mem[0]; - attrUse = key.mem[1]; - seqno = key.mem[2]; + attrSet = (int) key.mem[0]; + attrUse = (int) key.mem[1]; + seqno = (int) key.mem[2]; if (attrUseS == attrUse && attrSetS == attrSet) { @@ -136,86 +135,6 @@ static const char **searchRecordKey (ZebraHandle zh, iscz1_stop(decode_handle); assert (off == reckeys->buf_used); return ws; -#else - static const char *ws[32]; - int off = 0; - int startSeq = -1; - int i; - int seqno = 0; -#if SU_SCHEME - int chS, ch; -#else - short attrUse; - char attrSet; -#endif - - for (i = 0; i<32; i++) - ws[i] = NULL; - -#if SU_SCHEME - chS = zebraExplain_lookupSU (zh->reg->zei, attrSetS, attrUseS); - if (chS < 0) - return ws; -#endif - while (off < reckeys->buf_used) - { - - const char *src = reckeys->buf + off; - const char *wstart; - int lead; - - lead = *src++; -#if SU_SCHEME - if ((lead & 3)<3) - { - memcpy (&ch, src, sizeof(ch)); - src += sizeof(ch); - } -#else - if (!(lead & 1)) - { - memcpy (&attrSet, src, sizeof(attrSet)); - src += sizeof(attrSet); - } - if (!(lead & 2)) - { - memcpy (&attrUse, src, sizeof(attrUse)); - src += sizeof(attrUse); - } -#endif - wstart = src; - while (*src++) - ; - if (lead & 60) - seqno += ((lead>>2) & 15)-1; - else - { - memcpy (&seqno, src, sizeof(seqno)); - src += sizeof(seqno); - } - if ( -#if SU_SCHEME - ch == chS -#else - attrUseS == attrUse && attrSetS == attrSet -#endif - ) - { - int woff; - - - if (startSeq == -1) - startSeq = seqno; - woff = seqno - startSeq; - if (woff >= 0 && woff < 31) - ws[woff] = wstart; - } - - off = src - reckeys->buf; - } - assert (off == reckeys->buf_used); - return ws; -#endif } struct file_read_info { @@ -363,7 +282,7 @@ static char *fileMatchStr (ZebraHandle zh, } else { - logf (LOG_WARN, "Missing ) in match criteria %s in group %s", + yaz_log (YLOG_WARN, "Missing ) in match criteria %s in group %s", spec, zh->m_group ? zh->m_group : "none"); return NULL; } @@ -382,7 +301,7 @@ static char *fileMatchStr (ZebraHandle zh, } if (first) { - logf (LOG_WARN, "Record didn't contain match" + yaz_log (YLOG_WARN, "Record didn't contain match" " fields in (%s,%s)", attset_str, attname_str); return NULL; } @@ -439,7 +358,7 @@ static char *fileMatchStr (ZebraHandle zh, } else { - logf (LOG_WARN, "Syntax error in match criteria %s in group %s", + yaz_log (YLOG_WARN, "Syntax error in match criteria %s in group %s", spec, zh->m_group ? zh->m_group : "none"); return NULL; } @@ -447,7 +366,7 @@ static char *fileMatchStr (ZebraHandle zh, } if (dst == dstBuf) { - logf (LOG_WARN, "No match criteria for record %s in group %s", + yaz_log (YLOG_WARN, "No match criteria for record %s in group %s", fname, zh->m_group ? zh->m_group : "none"); return NULL; } @@ -464,13 +383,7 @@ struct recordLogInfo { void create_rec_keys_codec(struct recKeys *keys) { keys->buf_used = 0; -#if IT_KEY_NEW iscz1_reset(keys->codec_handle); -#else - keys->prevAttrUse = -1; - keys->prevAttrSet = -1; - keys->prevSeqNo = 0; -#endif } static int file_extract_record(ZebraHandle zh, @@ -487,13 +400,12 @@ static int file_extract_record(ZebraHandle zh, off_t recordOffset = 0; RecType recType; void *clientData; - char subType[128]; if (!(recType = - recType_byName (zh->reg->recTypes, zh->m_record_type, subType, + recType_byName (zh->reg->recTypes, zh->res, zh->m_record_type, &clientData))) { - logf (LOG_WARN, "No such record type: %s", zh->m_record_type); + yaz_log (YLOG_WARN, "No such record type: %s", zh->m_record_type); return 0; } @@ -523,7 +435,6 @@ static int file_extract_record(ZebraHandle zh, extractCtrl.tellf = file_tell; extractCtrl.endf = file_end; extractCtrl.fh = fi; - extractCtrl.subType = subType; extractCtrl.init = extract_init; extractCtrl.tokenAdd = extract_token_add; extractCtrl.schemaAdd = extract_schema_add; @@ -559,7 +470,7 @@ static int file_extract_record(ZebraHandle zh, if (zh->m_flag_rw && zh->records_processed < zh->m_file_verbose_limit) { - logf (LOG_WARN, "fail %s %s " PRINTF_OFF_T, zh->m_record_type, + yaz_log (YLOG_WARN, "fail %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); } return 0; @@ -570,7 +481,7 @@ static int file_extract_record(ZebraHandle zh, if (zh->m_flag_rw && zh->records_processed < zh->m_file_verbose_limit) { - logf (LOG_WARN, "no filter for %s %s " + yaz_log (YLOG_WARN, "no filter for %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); } @@ -583,8 +494,9 @@ static int file_extract_record(ZebraHandle zh, if (!zh->m_flag_rw) return 1; - logf (LOG_WARN, "empty %s %s " PRINTF_OFF_T, zh->m_record_type, - fname, recordOffset); + if (zh->records_processed < zh->m_file_verbose_limit) + yaz_log (YLOG_WARN, "empty %s %s " PRINTF_OFF_T, zh->m_record_type, + fname, recordOffset); return 1; } } @@ -613,7 +525,7 @@ static int file_extract_record(ZebraHandle zh, } else { - logf (LOG_WARN, "Bad match criteria"); + yaz_log (YLOG_WARN, "Bad match criteria"); return 0; } } @@ -624,13 +536,13 @@ static int file_extract_record(ZebraHandle zh, /* new record */ if (deleteFlag) { - logf (LOG_LOG, "delete %s %s " PRINTF_OFF_T, zh->m_record_type, + yaz_log (YLOG_LOG, "delete %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); - logf (LOG_WARN, "cannot delete record above (seems new)"); + yaz_log (YLOG_WARN, "cannot delete record above (seems new)"); return 1; } if (zh->records_processed < zh->m_file_verbose_limit) - logf (LOG_LOG, "add %s %s " PRINTF_OFF_T, zh->m_record_type, + yaz_log (YLOG_LOG, "add %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); rec = rec_new (zh->reg->records); @@ -661,8 +573,8 @@ static int file_extract_record(ZebraHandle zh, if (!force_update && recordAttr->runNumber == zebraExplain_runNumberIncrement (zh->reg->zei, 0)) { - yaz_log (LOG_LOG, "run number = %d", recordAttr->runNumber); - yaz_log (LOG_LOG, "skipped %s %s " PRINTF_OFF_T, + yaz_log (YLOG_LOG, "run number = %d", recordAttr->runNumber); + yaz_log (YLOG_LOG, "skipped %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); extract_flushSortKeys (zh, *sysno, -1, &zh->reg->sortKeys); rec_rm (&rec); @@ -682,14 +594,14 @@ static int file_extract_record(ZebraHandle zh, /* record going to be deleted */ if (!delkeys.buf_used) { - logf (LOG_LOG, "delete %s %s " PRINTF_OFF_T, + yaz_log (YLOG_LOG, "delete %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); - logf (LOG_WARN, "cannot delete file above, storeKeys false"); + yaz_log (YLOG_WARN, "cannot delete file above, storeKeys false"); } else { if (zh->records_processed < zh->m_file_verbose_limit) - logf (LOG_LOG, "delete %s %s " PRINTF_OFF_T, + yaz_log (YLOG_LOG, "delete %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); zh->records_deleted++; if (matchStr) @@ -705,14 +617,14 @@ static int file_extract_record(ZebraHandle zh, /* record going to be updated */ if (!delkeys.buf_used) { - logf (LOG_LOG, "update %s %s " PRINTF_OFF_T, + yaz_log (YLOG_LOG, "update %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); - logf (LOG_WARN, "cannot update file above, storeKeys false"); + yaz_log (YLOG_WARN, "cannot update file above, storeKeys false"); } else { if (zh->records_processed < zh->m_file_verbose_limit) - logf (LOG_LOG, "update %s %s " PRINTF_OFF_T, + yaz_log (YLOG_LOG, "update %s %s " PRINTF_OFF_T, zh->m_record_type, fname, recordOffset); extract_flushSortKeys (zh, *sysno, 1, &zh->reg->sortKeys); extract_flushRecordKeys (zh, *sysno, 1, &zh->reg->keys); @@ -775,14 +687,14 @@ static int file_extract_record(ZebraHandle zh, xmalloc (recordAttr->recordSize); if (lseek (fi->fd, recordOffset, SEEK_SET) < 0) { - logf (LOG_ERRNO|LOG_FATAL, "seek to " PRINTF_OFF_T " in %s", + yaz_log (YLOG_ERRNO|YLOG_FATAL, "seek to " PRINTF_OFF_T " in %s", recordOffset, fname); exit (1); } if (read (fi->fd, rec->info[recInfo_storeData], recordAttr->recordSize) < recordAttr->recordSize) { - logf (LOG_ERRNO|LOG_FATAL, "read %d bytes of %s", + yaz_log (YLOG_ERRNO|YLOG_FATAL, "read %d bytes of %s", recordAttr->recordSize, fname); exit (1); } @@ -821,7 +733,7 @@ int fileExtract (ZebraHandle zh, SYSNO *sysno, const char *fname, else sprintf (gprefix, "%s.", zh->m_group); - logf (LOG_DEBUG, "fileExtract %s", fname); + yaz_log (YLOG_DEBUG, "fileExtract %s", fname); /* determine file extension */ *ext = '\0'; @@ -843,7 +755,7 @@ int fileExtract (ZebraHandle zh, SYSNO *sysno, const char *fname, if (!zh->m_record_type) { if (zh->records_processed < zh->m_file_verbose_limit) - logf (LOG_LOG, "? %s", fname); + yaz_log (YLOG_LOG, "? %s", fname); return 0; } /* determine match criteria */ @@ -871,7 +783,7 @@ int fileExtract (ZebraHandle zh, SYSNO *sysno, const char *fname, if ((fd = open (full_rep, O_BINARY|O_RDONLY)) == -1) { - logf (LOG_WARN|LOG_ERRNO, "open %s", full_rep); + yaz_log (YLOG_WARN|YLOG_ERRNO, "open %s", full_rep); zh->m_record_type = original_record_type; return 0; } @@ -911,7 +823,6 @@ int buffer_extract_record (ZebraHandle zh, int i, r; const char *matchStr = 0; RecType recType = NULL; - char subType[1024]; void *clientData; Record rec; long recordOffset = 0; @@ -947,26 +858,25 @@ int buffer_extract_record (ZebraHandle zh, } if (recordType && *recordType) { - logf (LOG_DEBUG, "Record type explicitly specified: %s", recordType); - recType = recType_byName (zh->reg->recTypes, recordType, subType, + yaz_log (YLOG_DEBUG, "Record type explicitly specified: %s", recordType); + recType = recType_byName (zh->reg->recTypes, zh->res, recordType, &clientData); } else { if (!(zh->m_record_type)) { - logf (LOG_WARN, "No such record type defined"); + yaz_log (YLOG_WARN, "No such record type defined"); return 0; } - logf (LOG_DEBUG, "Get record type from rgroup: %s",zh->m_record_type); - recType = recType_byName (zh->reg->recTypes, zh->m_record_type, subType, - &clientData); + yaz_log (YLOG_DEBUG, "Get record type from rgroup: %s",zh->m_record_type); + recType = recType_byName (zh->reg->recTypes, zh->res, + zh->m_record_type, &clientData); recordType = zh->m_record_type; } if (!recType) { - logf (LOG_WARN, "No such record type: %s", zh->m_record_type); + yaz_log (YLOG_WARN, "No such record type: %s", zh->m_record_type); return 0; } - extractCtrl.subType = subType; extractCtrl.init = extract_init; extractCtrl.tokenAdd = extract_token_add; extractCtrl.schemaAdd = extract_schema_add; @@ -989,13 +899,13 @@ int buffer_extract_record (ZebraHandle zh, else if (r == RECCTRL_EXTRACT_ERROR_GENERIC) { /* error occured during extraction ... */ - yaz_log (LOG_WARN, "extract error: generic"); + yaz_log (YLOG_WARN, "extract error: generic"); return 0; } else if (r == RECCTRL_EXTRACT_ERROR_NO_SUCH_FILTER) { /* error occured during extraction ... */ - yaz_log (LOG_WARN, "extract error: no such filter"); + yaz_log (YLOG_WARN, "extract error: no such filter"); return 0; } if (zh->reg->keys.buf_used == 0) @@ -1004,8 +914,8 @@ int buffer_extract_record (ZebraHandle zh, is probably empty - unless flagShowRecords is in use */ if (test_mode) return 1; - logf (LOG_WARN, "No keys generated for record"); - logf (LOG_WARN, " The file is probably empty"); + yaz_log (YLOG_WARN, "No keys generated for record"); + yaz_log (YLOG_WARN, " The file is probably empty"); return 1; } /* match criteria */ @@ -1021,7 +931,7 @@ int buffer_extract_record (ZebraHandle zh, zh->m_record_id); if (!matchStr) { - logf (LOG_WARN, "Bad match criteria (recordID)"); + yaz_log (YLOG_WARN, "Bad match criteria (recordID)"); return 1; } } @@ -1041,12 +951,12 @@ int buffer_extract_record (ZebraHandle zh, /* new record */ if (delete_flag) { - logf (LOG_LOG, "delete %s %s %ld", recordType, + yaz_log (YLOG_LOG, "delete %s %s %ld", recordType, pr_fname, (long) recordOffset); - logf (LOG_WARN, "cannot delete record above (seems new)"); + yaz_log (YLOG_WARN, "cannot delete record above (seems new)"); return 1; } - logf (LOG_LOG, "add %s %s %ld", recordType, pr_fname, + yaz_log (YLOG_LOG, "add %s %s %ld", recordType, pr_fname, (long) recordOffset); rec = rec_new (zh->reg->records); @@ -1071,7 +981,7 @@ int buffer_extract_record (ZebraHandle zh, struct sortKeys sortKeys; if (!allow_update) { - logf (LOG_LOG, "skipped %s %s %ld", + yaz_log (YLOG_LOG, "skipped %s %s %ld", recordType, pr_fname, (long) recordOffset); logRecord(zh); return -1; @@ -1086,7 +996,7 @@ int buffer_extract_record (ZebraHandle zh, if (recordAttr->runNumber == zebraExplain_runNumberIncrement (zh->reg->zei, 0)) { - logf (LOG_LOG, "skipped %s %s %ld", recordType, + yaz_log (YLOG_LOG, "skipped %s %s %ld", recordType, pr_fname, (long) recordOffset); extract_flushSortKeys (zh, *sysno, -1, &zh->reg->sortKeys); rec_rm (&rec); @@ -1108,13 +1018,13 @@ int buffer_extract_record (ZebraHandle zh, /* record going to be deleted */ if (!delkeys.buf_used) { - logf (LOG_LOG, "delete %s %s %ld", recordType, + yaz_log (YLOG_LOG, "delete %s %s %ld", recordType, pr_fname, (long) recordOffset); - logf (LOG_WARN, "cannot delete file above, storeKeys false"); + yaz_log (YLOG_WARN, "cannot delete file above, storeKeys false"); } else { - logf (LOG_LOG, "delete %s %s %ld", recordType, + yaz_log (YLOG_LOG, "delete %s %s %ld", recordType, pr_fname, (long) recordOffset); zh->records_deleted++; if (matchStr) @@ -1130,13 +1040,13 @@ int buffer_extract_record (ZebraHandle zh, /* record going to be updated */ if (!delkeys.buf_used) { - logf (LOG_LOG, "update %s %s %ld", recordType, + yaz_log (YLOG_LOG, "update %s %s %ld", recordType, pr_fname, (long) recordOffset); - logf (LOG_WARN, "cannot update file above, storeKeys false"); + yaz_log (YLOG_WARN, "cannot update file above, storeKeys false"); } else { - logf (LOG_LOG, "update %s %s %ld", recordType, + yaz_log (YLOG_LOG, "update %s %s %ld", recordType, pr_fname, (long) recordOffset); extract_flushSortKeys (zh, *sysno, 1, &zh->reg->sortKeys); extract_flushRecordKeys (zh, *sysno, 1, &zh->reg->keys); @@ -1289,16 +1199,7 @@ int explain_extract (void *handle, Record rec, data1_node *n) void extract_flushRecordKeys (ZebraHandle zh, SYSNO sysno, int cmd, struct recKeys *reckeys) { -#if IT_KEY_NEW void *decode_handle = iscz1_start(); -#else - int seqno = 0; -#if SU_SCHEME -#else - unsigned char attrSet = (unsigned char) -1; - unsigned short attrUse = (unsigned short) -1; -#endif -#endif int off = 0; int ch = 0; ZebraExplainInfo zei = zh->reg->zei; @@ -1308,7 +1209,7 @@ void extract_flushRecordKeys (ZebraHandle zh, SYSNO sysno, int mem= 1024*1024* atoi( res_get_def( zh->res, "memmax", "8")); if (mem <= 0) { - logf(LOG_WARN, "Invalid memory setting, using default 8 MB"); + yaz_log(YLOG_WARN, "Invalid memory setting, using default 8 MB"); mem= 1024*1024*8; } /* FIXME: That "8" should be in a default settings include */ @@ -1320,7 +1221,7 @@ void extract_flushRecordKeys (ZebraHandle zh, SYSNO sysno, zh->reg->key_file_no = 0; } zebraExplain_recordCountIncrement (zei, cmd ? 1 : -1); -#if IT_KEY_NEW + while (off < reckeys->buf_used) { const char *src = reckeys->buf + off; @@ -1331,13 +1232,12 @@ void extract_flushRecordKeys (ZebraHandle zh, SYSNO sysno, iscz1_decode(decode_handle, &dst, &src); assert(key.len < 4 && key.len > 2); - attrSet = key.mem[0]; - attrUse = key.mem[1]; /* sequence in mem[2] */ + attrSet = (int) key.mem[0]; + attrUse = (int) key.mem[1]; /* sequence in mem[2] */ if (zh->reg->key_buf_used + 1024 > (zh->reg->ptr_top -zh->reg->ptr_i)*sizeof(char*)) extract_flushWriteKeys (zh,0); - assert(zh->reg->ptr_i >= 0); ++(zh->reg->ptr_i); assert(zh->reg->ptr_i > 0); (zh->reg->key_buf)[zh->reg->ptr_top - zh->reg->ptr_i] = @@ -1366,76 +1266,8 @@ void extract_flushRecordKeys (ZebraHandle zh, SYSNO sysno, (zh->reg->key_buf_used) += sizeof(key); off = src - reckeys->buf; } -#else - while (off < reckeys->buf_used) - { - const char *src = reckeys->buf + off; - struct it_key key; - int lead; - - lead = *src++; - -#if SU_SCHEME - if ((lead & 3) < 3) - { - memcpy (&ch, src, sizeof(ch)); - src += sizeof(ch); - } -#else - if (!(lead & 1)) - { - memcpy (&attrSet, src, sizeof(attrSet)); - src += sizeof(attrSet); - } - if (!(lead & 2)) - { - memcpy (&attrUse, src, sizeof(attrUse)); - src += sizeof(attrUse); - } -#endif - if (zh->reg->key_buf_used + 1024 > - (zh->reg->ptr_top -zh->reg->ptr_i)*sizeof(char*)) - extract_flushWriteKeys (zh,0); - assert(zh->reg->ptr_i >= 0); - ++(zh->reg->ptr_i); - assert(zh->reg->ptr_i > 0); - (zh->reg->key_buf)[zh->reg->ptr_top - zh->reg->ptr_i] = - (char*)zh->reg->key_buf + zh->reg->key_buf_used; -#if SU_SCHEME -#else - ch = zebraExplain_lookupSU (zei, attrSet, attrUse); - if (ch < 0) - ch = zebraExplain_addSU (zei, attrSet, attrUse); -#endif - assert (ch > 0); - zh->reg->key_buf_used += - key_SU_encode (ch,((char*)zh->reg->key_buf) + - zh->reg->key_buf_used); - - while (*src) - ((char*)zh->reg->key_buf) [(zh->reg->key_buf_used)++] = *src++; - src++; - ((char*)(zh->reg->key_buf))[(zh->reg->key_buf_used)++] = '\0'; - ((char*)(zh->reg->key_buf))[(zh->reg->key_buf_used)++] = cmd; - - if (lead & 60) - seqno += ((lead>>2) & 15)-1; - else - { - memcpy (&seqno, src, sizeof(seqno)); - src += sizeof(seqno); - } - key.seqno = seqno; - key.sysno = sysno; - memcpy ((char*)zh->reg->key_buf + zh->reg->key_buf_used, &key, sizeof(key)); - (zh->reg->key_buf_used) += sizeof(key); - off = src - reckeys->buf; - } -#endif assert (off == reckeys->buf_used); -#if IT_KEY_NEW iscz1_stop(decode_handle); -#endif } void extract_flushWriteKeys (ZebraHandle zh, int final) @@ -1454,19 +1286,19 @@ void extract_flushWriteKeys (ZebraHandle zh, int final) #endif if (!zh->reg->key_buf || ptr_i <= 0) { - logf (LOG_DEBUG, " nothing to flush section=%d buf=%p i=%d", + yaz_log (YLOG_DEBUG, " nothing to flush section=%d buf=%p i=%d", zh->reg->key_file_no, zh->reg->key_buf, ptr_i); - logf (LOG_DEBUG, " buf=%p ", + yaz_log (YLOG_DEBUG, " buf=%p ", zh->reg->key_buf); - logf (LOG_DEBUG, " ptr=%d ",zh->reg->ptr_i); - logf (LOG_DEBUG, " reg=%p ",zh->reg); + yaz_log (YLOG_DEBUG, " ptr=%d ",zh->reg->ptr_i); + yaz_log (YLOG_DEBUG, " reg=%p ",zh->reg); return; } (zh->reg->key_file_no)++; - logf (LOG_LOG, "sorting section %d", (zh->reg->key_file_no)); - logf (LOG_DEBUG, " sort_buff at %p n=%d", + yaz_log (YLOG_LOG, "sorting section %d", (zh->reg->key_file_no)); + yaz_log (YLOG_DEBUG, " sort_buff at %p n=%d", zh->reg->key_buf + zh->reg->ptr_top - ptr_i,ptr_i); #if !SORT_EXTRA qsort (zh->reg->key_buf + zh->reg->ptr_top - ptr_i, ptr_i, @@ -1480,7 +1312,7 @@ void extract_flushWriteKeys (ZebraHandle zh, int final) temp_policy=toupper(res_get_def(zh->res,"tempfiles","auto")[0]); if (temp_policy != 'Y' && temp_policy != 'N' && temp_policy != 'A') { - logf (LOG_WARN, "Illegal tempfiles setting '%c'. using 'Auto' ", + yaz_log (YLOG_WARN, "Illegal tempfiles setting '%c'. using 'Auto' ", temp_policy); temp_policy='A'; } @@ -1502,10 +1334,10 @@ void extract_flushWriteKeys (ZebraHandle zh, int final) if (!(outf = fopen (out_fname, "wb"))) { - logf (LOG_FATAL|LOG_ERRNO, "fopen %s", out_fname); + yaz_log (YLOG_FATAL|YLOG_ERRNO, "fopen %s", out_fname); exit (1); } - logf (LOG_LOG, "writing section %d", zh->reg->key_file_no); + yaz_log (YLOG_LOG, "writing section %d", zh->reg->key_file_no); prevcp = cp = (zh->reg->key_buf)[zh->reg->ptr_top - ptr_i]; encode_key_init (&encode_info); @@ -1531,10 +1363,10 @@ void extract_flushWriteKeys (ZebraHandle zh, int final) if (!(outf = fopen (out_fname, "wb"))) { - logf (LOG_FATAL|LOG_ERRNO, "fopen %s", out_fname); + yaz_log (YLOG_FATAL|YLOG_ERRNO, "fopen %s", out_fname); exit (1); } - logf (LOG_LOG, "writing section %d", key_file_no); + yaz_log (YLOG_LOG, "writing section %d", key_file_no); i = ptr_i; prevcp = key_buf[ptr_top-i]; while (1) @@ -1542,7 +1374,7 @@ void extract_flushWriteKeys (ZebraHandle zh, int final) { key_y_len = strlen(prevcp)+1; #if 0 - logf (LOG_LOG, "key_y_len: %2d %02x %02x %s", + yaz_log (YLOG_LOG, "key_y_len: %2d %02x %02x %s", key_y_len, prevcp[0], prevcp[1], 2+prevcp); #endif qsort (key_buf + ptr_top-ptr_i, ptr_i - i, @@ -1564,10 +1396,10 @@ void extract_flushWriteKeys (ZebraHandle zh, int final) #endif if (fclose (outf)) { - logf (LOG_FATAL|LOG_ERRNO, "fclose %s", out_fname); + yaz_log (YLOG_FATAL|YLOG_ERRNO, "fclose %s", out_fname); exit (1); } - logf (LOG_LOG, "finished section %d", zh->reg->key_file_no); + yaz_log (YLOG_LOG, "finished section %d", zh->reg->key_file_no); zh->reg->ptr_i = 0; zh->reg->key_buf_used = 0; } @@ -1577,17 +1409,8 @@ void extract_add_index_string (RecWord *p, const char *str, int length) char *dst; ZebraHandle zh = p->extractCtrl->handle; struct recKeys *keys = &zh->reg->keys; -#if IT_KEY_NEW struct it_key key; const char *src = (char*) &key; -#else - unsigned char attrSet; - unsigned short attrUse; - int lead = 0; - int diff = 0; - int *pseqno = &p->seqno; - ZebraExplainInfo zei = zh->reg->zei; -#endif if (keys->buf_used+1024 > keys->buf_max) { @@ -1601,7 +1424,6 @@ void extract_add_index_string (RecWord *p, const char *str, int length) } dst = keys->buf + keys->buf_used; -#if IT_KEY_NEW key.len = 3; key.mem[0] = p->attrSet; key.mem[1] = p->attrUse; @@ -1609,7 +1431,7 @@ void extract_add_index_string (RecWord *p, const char *str, int length) #if 0 /* just for debugging .. */ - yaz_log(LOG_LOG, "set=%d use=%d seqno=%d", p->attrSet, p->attrUse, + yaz_log(YLOG_LOG, "set=%d use=%d seqno=%d", p->attrSet, p->attrUse, p->seqno); #endif @@ -1619,69 +1441,6 @@ void extract_add_index_string (RecWord *p, const char *str, int length) memcpy (dst, str, length); dst += length; *dst++ = '\0'; -#else - /* leader byte is encoded as follows: - bit 0 : 1 if attrset is unchanged; 0 if attrset is changed - bit 1 : 1 if attruse is unchanged; 0 if attruse is changed - */ - attrSet = p->attrSet; - if (keys->buf_used > 0 && keys->prevAttrSet == attrSet) - lead |= 1; - else - keys->prevAttrSet = attrSet; - attrUse = p->attrUse; - if (keys->buf_used > 0 && keys->prevAttrUse == attrUse) - lead |= 2; - else - keys->prevAttrUse = attrUse; - - diff = 1 + *pseqno - keys->prevSeqNo; - if (diff >= 1 && diff <= 15) - lead |= (diff << 2); - else - diff = 0; - - keys->prevSeqNo = *pseqno; - - *dst++ = lead; - -#if SU_SCHEME - if ((lead & 3) < 3) - { - int ch = zebraExplain_lookupSU (zei, attrSet, attrUse); - if (ch < 0) - { - ch = zebraExplain_addSU (zei, attrSet, attrUse); - yaz_log (LOG_DEBUG, "addSU set=%d use=%d SU=%d", - attrSet, attrUse, ch); - } - assert (ch > 0); - memcpy (dst, &ch, sizeof(ch)); - dst += sizeof(ch); - } -#else - if (!(lead & 1)) - { - memcpy (dst, &attrSet, sizeof(attrSet)); - dst += sizeof(attrSet); - } - if (!(lead & 2)) - { - memcpy (dst, &attrUse, sizeof(attrUse)); - dst += sizeof(attrUse); - } -#endif - *dst++ = p->reg_type; - memcpy (dst, str, length); - dst += length; - *dst++ = '\0'; - - if (!diff) - { - memcpy (dst, pseqno, sizeof(*pseqno)); - dst += sizeof(*pseqno); - } -#endif keys->buf_used = dst - keys->buf; } @@ -1738,7 +1497,7 @@ static void extract_add_incomplete_field (RecWord *p) const char **map = 0; if (remain > 0) - map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain); + map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, 0); while (map) { @@ -1750,7 +1509,7 @@ static void extract_add_incomplete_field (RecWord *p) { remain = p->length - (b - p->string); if (remain > 0) - map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain); + map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, 0); else map = 0; } @@ -1765,7 +1524,7 @@ static void extract_add_incomplete_field (RecWord *p) buf[i++] = *(cp++); remain = p->length - (b - p->string); if (remain > 0) - map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain); + map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, 0); else map = 0; } @@ -1782,9 +1541,12 @@ static void extract_add_complete_field (RecWord *p) char buf[IT_MAX_WORD+1]; const char **map = 0; int i = 0, remain = p->length; + int first; /* first position */ + +yaz_log(YLOG_DEBUG, "Complete field, w='%.*s'", p->length, p->string); if (remain > 0) - map = zebra_maps_input (p->zebra_maps, p->reg_type, &b, remain); + map = zebra_maps_input (p->zebra_maps, p->reg_type, &b, remain, 1); while (remain > 0 && i < IT_MAX_WORD) { @@ -1793,7 +1555,10 @@ static void extract_add_complete_field (RecWord *p) remain = p->length - (b - p->string); if (remain > 0) - map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain); + { + first = i ? 0 : 1; + map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, first); + } else map = 0; } @@ -1814,13 +1579,16 @@ static void extract_add_complete_field (RecWord *p) { if (i >= IT_MAX_WORD) break; + yaz_log(YLOG_DEBUG, "Adding string to index '%d'", **map); while (i < IT_MAX_WORD && *cp) buf[i++] = *(cp++); } remain = p->length - (b - p->string); if (remain > 0) + { map = zebra_maps_input (p->zebra_maps, p->reg_type, &b, - remain); + remain, 0); + } else map = 0; } @@ -1834,7 +1602,7 @@ void extract_token_add (RecWord *p) { WRBUF wrbuf; #if 0 - yaz_log (LOG_LOG, "token_add " + yaz_log (YLOG_LOG, "token_add " "reg_type=%c attrSet=%d attrUse=%d seqno=%d s=%.*s", p->reg_type, p->attrSet, p->attrUse, p->seqno, p->length, p->string); @@ -1891,39 +1659,9 @@ void encode_key_init (struct encode_info *i) i->prevseq=0; i->prevcmd=-1; i->keylen=0; -#if IT_KEY_NEW i->encode_handle = iscz1_start(); -#endif } -#if IT_KEY_NEW -#else -char *encode_key_int (int d, char *bp) -{ - if (d <= 63) - *bp++ = d; - else if (d <= 16383) - { - *bp++ = 64 + (d>>8); - *bp++ = d & 255; - } - else if (d <= 4194303) - { - *bp++ = 128 + (d>>16); - *bp++ = (d>>8) & 255; - *bp++ = d & 255; - } - else - { - *bp++ = 192 + (d>>24); - *bp++ = (d>>16) & 255; - *bp++ = (d>>8) & 255; - *bp++ = d & 255; - } - return bp; -} -#endif - #define OLDENCODE 1 #ifdef OLDENCODE @@ -1942,35 +1680,20 @@ void encode_key_write (char *k, struct encode_info *i, FILE *outf) ; /* and copy & align key so we can mangle */ memcpy (&key, k+1, sizeof(struct it_key)); /* *k is insert/delete */ -#if IT_KEY_NEW + bp0 = bp++; iscz1_encode(i->encode_handle, &bp, &src); *bp0 = (*k * 128) + bp - bp0 - 1; /* length and insert/delete combined */ -#else - bp = encode_key_int ( (key.sysno - i->sysno) * 2 + *k, bp); - if (i->sysno != key.sysno) - { - 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; -#endif if (fwrite (i->buf, bp - i->buf, 1, outf) != 1) { - logf (LOG_FATAL|LOG_ERRNO, "fwrite"); + yaz_log (YLOG_FATAL|YLOG_ERRNO, "fwrite"); exit (1); } } void encode_key_flush (struct encode_info *i, FILE *outf) { /* dummy routine */ -#if IT_KEY_NEW iscz1_stop(i->encode_handle); -#endif } #else @@ -2032,7 +1755,7 @@ void encode_key_write (char *k, struct encode_info *i, FILE *outf) i->cmd = i->prevcmd; if (fwrite (i->buf, bp - i->buf, 1, outf) != 1) { - logf (LOG_FATAL|LOG_ERRNO, "fwrite"); + yaz_log (YLOG_FATAL|YLOG_ERRNO, "fwrite"); exit (1); } i->keylen=0; /* ok, it's written, forget it */ @@ -2066,7 +1789,7 @@ void encode_key_flush (struct encode_info *i, FILE *outf) i->cmd = i->prevcmd; if (fwrite (i->buf, bp - i->buf, 1, outf) != 1) { - logf (LOG_FATAL|LOG_ERRNO, "fwrite"); + yaz_log (YLOG_FATAL|YLOG_ERRNO, "fwrite"); exit (1); } i->keylen=0; /* ok, it's written, forget it */