-/* $Id: extract.c,v 1.165 2004-10-04 09:27:48 adam Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004
- Index Data Aps
+/* $Id: extract.c,v 1.186 2005-06-14 20:28:54 adam Exp $
+ Copyright (C) 1995-2005
+ Index Data ApS
This file is part of the Zebra server.
#include <ctype.h>
#ifdef WIN32
#include <io.h>
-#else
+#endif
+#if HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <fcntl.h>
++zh->records_processed;
if (!(zh->records_processed % 1000))
{
- logf (LOG_LOG, "Records: "ZINT_FORMAT" i/u/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);
}
}
+static void extract_set_store_data_prepare(struct recExtractCtrl *p);
+
static void extract_init (struct recExtractCtrl *p, RecWord *w)
{
w->zebra_maps = p->zebra_maps;
w->seqno = 1;
w->attrSet = VAL_BIB1;
w->attrUse = 1016;
+ w->attrStr = 0;
w->reg_type = 'w';
w->extractCtrl = p;
+ w->record_id = 0;
+ w->section_id = 0;
}
static const char **searchRecordKey (ZebraHandle zh,
int attrSet, attrUse;
iscz1_decode(decode_handle, &dst, &src);
- assert(key.len < 4 && key.len > 2);
+ assert(key.len <= 4 && key.len > 2);
- attrSet = (int) key.mem[0];
- attrUse = (int) key.mem[1];
- seqno = (int) key.mem[2];
+ attrSet = (int) key.mem[0] >> 16;
+ attrUse = (int) key.mem[0] & 65535;
+ seqno = (int) key.mem[key.len-1];
if (attrUseS == attrUse && attrSetS == attrSet)
{
off_t file_moffset; /* offset of rec/rec boundary */
int file_more;
int fd;
- char *sdrbuf;
- int sdrmax;
};
static struct file_read_info *file_read_start (int fd)
fi->fd = fd;
fi->file_max = 0;
fi->file_moffset = 0;
- fi->sdrbuf = 0;
- fi->sdrmax = 0;
+ fi->file_offset = 0;
+ fi->file_more = 0;
return fi;
}
{
struct file_read_info *p = (struct file_read_info *) handle;
p->file_offset = offset;
- if (p->sdrbuf)
- return offset;
return lseek (p->fd, offset, SEEK_SET);
}
struct file_read_info *p = (struct file_read_info *) handle;
int fd = p->fd;
int r;
- if (p->sdrbuf)
- {
- r = count;
- if (r > p->sdrmax - p->file_offset)
- r = p->sdrmax - p->file_offset;
- if (r)
- memcpy (buf, p->sdrbuf + p->file_offset, r);
- }
- else
- r = read (fd, buf, count);
+ r = read (fd, buf, count);
if (r > 0)
{
p->file_offset += r;
return r;
}
-static void file_begin (void *handle)
-{
- struct file_read_info *p = (struct file_read_info *) handle;
-
- p->file_offset = p->file_moffset;
- if (!p->sdrbuf && p->file_moffset)
- lseek (p->fd, p->file_moffset, SEEK_SET);
- p->file_more = 0;
-}
-
static void file_end (void *handle, off_t offset)
{
struct file_read_info *p = (struct file_read_info *) handle;
- assert (p->file_more == 0);
- p->file_more = 1;
- p->file_moffset = offset;
+ if (offset != p->file_moffset)
+ {
+ p->file_moffset = offset;
+ p->file_more = 1;
+ }
}
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;
}
}
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;
}
}
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;
}
}
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;
}
keys->buf_used = 0;
iscz1_reset(keys->codec_handle);
}
-
+
+static void init_extractCtrl(ZebraHandle zh, struct recExtractCtrl *ctrl)
+{
+ int i;
+ for (i = 0; i<256; i++)
+ {
+ if (zebra_maps_is_positioned(zh->reg->zebra_maps, i))
+ ctrl->seqno[i] = 1;
+ else
+ ctrl->seqno[i] = 0;
+ }
+ ctrl->zebra_maps = zh->reg->zebra_maps;
+ ctrl->flagShowRecords = !zh->m_flag_rw;
+}
+
static int file_extract_record(ZebraHandle zh,
SYSNO *sysno, const char *fname,
int deleteFlag,
struct file_read_info *fi,
- int force_update)
+ int force_update,
+ RecType recType,
+ void *recTypeClientData)
{
RecordAttr *recordAttr;
int r;
- const char *matchStr;
+ const char *matchStr = 0;
SYSNO sysnotmp;
Record rec;
off_t recordOffset = 0;
- RecType recType;
- void *clientData;
- if (!(recType =
- recType_byName (zh->reg->recTypes, zh->res, zh->m_record_type,
- &clientData)))
- {
- logf (LOG_WARN, "No such record type: %s", zh->m_record_type);
- return 0;
- }
-
/* announce database */
if (zebraExplain_curDatabase (zh->reg->zei, zh->basenames[0]))
{
struct recExtractCtrl extractCtrl;
/* we are going to read from a file, so prepare the extraction */
- int i;
-
create_rec_keys_codec(&zh->reg->keys);
zh->reg->sortKeys.buf_used = 0;
recordOffset = fi->file_moffset;
+ extractCtrl.handle = zh;
extractCtrl.offset = fi->file_moffset;
extractCtrl.readf = file_read;
extractCtrl.seekf = file_seek;
extractCtrl.tokenAdd = extract_token_add;
extractCtrl.schemaAdd = extract_schema_add;
extractCtrl.dh = zh->reg->dh;
- extractCtrl.handle = zh;
- for (i = 0; i<256; i++)
- {
- if (zebra_maps_is_positioned(zh->reg->zebra_maps, i))
- extractCtrl.seqno[i] = 1;
- else
- extractCtrl.seqno[i] = 0;
- }
- extractCtrl.zebra_maps = zh->reg->zebra_maps;
- extractCtrl.flagShowRecords = !zh->m_flag_rw;
+ extractCtrl.match_criteria[0] = '\0';
+ extractCtrl.first_record = fi->file_offset ? 0 : 1;
+
+ extract_set_store_data_prepare(&extractCtrl);
+
+ init_extractCtrl(zh, &extractCtrl);
if (!zh->m_flag_rw)
printf ("File: %s " PRINTF_OFF_T "\n", fname, recordOffset);
yaz_log_init_prefix2 (msg);
}
- r = (*recType->extract)(clientData, &extractCtrl);
+ r = (*recType->extract)(recTypeClientData, &extractCtrl);
yaz_log_init_prefix2 (0);
if (r == RECCTRL_EXTRACT_EOF)
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;
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);
}
return 0;
}
- if (zh->reg->keys.buf_used == 0)
- {
- /* the extraction process returned no information - the record
- is probably empty - unless flagShowRecords is in use */
- if (!zh->m_flag_rw)
- return 1;
-
- if (zh->records_processed < zh->m_file_verbose_limit)
- logf (LOG_WARN, "empty %s %s " PRINTF_OFF_T, zh->m_record_type,
- fname, recordOffset);
- return 1;
- }
+ if (extractCtrl.match_criteria[0])
+ matchStr = extractCtrl.match_criteria;
}
/* perform match if sysno not known and if match criteria is specified */
-
- matchStr = NULL;
if (!sysno)
{
sysnotmp = 0;
sysno = &sysnotmp;
- if (zh->m_record_id && *zh->m_record_id)
+
+ if (matchStr == 0 && zh->m_record_id && *zh->m_record_id)
{
- char *rinfo;
matchStr = fileMatchStr (zh, &zh->reg->keys, fname,
zh->m_record_id);
- if (matchStr)
- {
- rinfo = dict_lookup (zh->reg->matchDict, matchStr);
- if (rinfo)
- {
- assert(*rinfo == sizeof(*sysno));
- memcpy (sysno, rinfo+1, sizeof(*sysno));
- }
- }
- else
- {
- logf (LOG_WARN, "Bad match criteria");
- return 0;
- }
- }
+ if (!matchStr)
+ {
+ yaz_log(YLOG_WARN, "Bad match criteria");
+ return 0;
+ }
+ }
+ if (matchStr)
+ {
+ char *rinfo = dict_lookup (zh->reg->matchDict, matchStr);
+ if (rinfo)
+ {
+ assert(*rinfo == sizeof(*sysno));
+ memcpy (sysno, rinfo+1, sizeof(*sysno));
+ }
+ }
+ }
+ if (! *sysno && zh->reg->keys.buf_used == 0)
+ {
+ /* the extraction process returned no information - the record
+ is probably empty - unless flagShowRecords is in use */
+ if (!zh->m_flag_rw)
+ return 1;
+
+ 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;
}
if (! *sysno)
/* 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);
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 = " ZINT_FORMAT,
+ 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);
/* 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)
/* 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);
/* update store data */
xfree (rec->info[recInfo_storeData]);
- if (zh->m_store_data)
+ if (zh->store_data_buf)
+ {
+ rec->size[recInfo_storeData] = zh->store_data_size;
+ rec->info[recInfo_storeData] = zh->store_data_buf;
+ zh->store_data_buf = 0;
+ }
+ else if (zh->m_store_data)
{
rec->size[recInfo_storeData] = 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 " 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);
}
char ext_res[128];
struct file_read_info *fi;
const char *original_record_type = 0;
+ RecType recType;
+ void *recTypeClientData;
if (!zh->m_group || !*zh->m_group)
*gprefix = '\0';
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';
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 */
zh->m_record_id = res_get (zh->res, ext_res);
}
+ if (!(recType =
+ recType_byName (zh->reg->recTypes, zh->res, zh->m_record_type,
+ &recTypeClientData)))
+ {
+ yaz_log(YLOG_WARN, "No such record type: %s", zh->m_record_type);
+ return 0;
+ }
+
+ switch(recType->version)
+ {
+ case 0:
+ break;
+ default:
+ yaz_log(YLOG_WARN, "Bad filter version: %s", zh->m_record_type);
+ }
if (sysno && deleteFlag)
fd = -1;
else
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;
}
fi = file_read_start (fd);
do
{
- file_begin (fi);
- r = file_extract_record (zh, sysno, fname, deleteFlag, fi, 1);
- } while (r && !sysno && fi->file_more);
+ fi->file_moffset = fi->file_offset;
+ fi->file_more = 0; /* file_end not called (yet) */
+ r = file_extract_record (zh, sysno, fname, deleteFlag, fi, 1,
+ recType, recTypeClientData);
+ if (fi->file_more)
+ { /* file_end has been called so reset offset .. */
+ fi->file_offset = fi->file_moffset;
+ lseek(fi->fd, fi->file_moffset, SEEK_SET);
+ }
+ }
+ while (r && !sysno);
file_read_stop (fi);
if (fd != -1)
close (fd);
If not, and a record is provided, then sysno is got from there
*/
-int buffer_extract_record (ZebraHandle zh,
- const char *buf, size_t buf_size,
- int delete_flag,
- int test_mode,
- const char *recordType,
- SYSNO *sysno,
- const char *match_criteria,
- const char *fname,
- int force_update,
- int allow_update)
+ZEBRA_RES buffer_extract_record (ZebraHandle zh,
+ const char *buf, size_t buf_size,
+ int delete_flag,
+ int test_mode,
+ const char *recordType,
+ SYSNO *sysno,
+ const char *match_criteria,
+ const char *fname,
+ int force_update,
+ int allow_update)
{
RecordAttr *recordAttr;
struct recExtractCtrl extractCtrl;
- int i, r;
+ int r;
const char *matchStr = 0;
RecType recType = NULL;
void *clientData;
long recordOffset = 0;
struct zebra_fetch_control fc;
const char *pr_fname = fname; /* filename to print .. */
+ int show_progress = zh->records_processed < zh->m_file_verbose_limit ? 1:0;
if (!pr_fname)
pr_fname = "<no file>"; /* make it printable if file is omitted */
extractCtrl.seekf = zebra_record_int_seek;
extractCtrl.tellf = zebra_record_int_tell;
extractCtrl.endf = zebra_record_int_end;
+ extractCtrl.first_record = 1;
extractCtrl.fh = &fc;
create_rec_keys_codec(&zh->reg->keys);
{
if (zebraExplain_newDatabase (zh->reg->zei, zh->basenames[0],
zh->m_explain_database))
- return 0;
+ return ZEBRA_FAIL;
}
- if (recordType && *recordType) {
- logf (LOG_DEBUG, "Record type explicitly specified: %s", recordType);
+ if (recordType && *recordType)
+ {
+ 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");
- return 0;
+ }
+ else
+ {
+ if (!(zh->m_record_type))
+ {
+ yaz_log (YLOG_WARN, "No such record type defined");
+ return ZEBRA_FAIL;
}
- logf (LOG_DEBUG, "Get record type from rgroup: %s",zh->m_record_type);
+ 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);
- return 0;
+ if (!recType)
+ {
+ yaz_log (YLOG_WARN, "No such record type: %s", zh->m_record_type);
+ return ZEBRA_FAIL;
}
extractCtrl.init = extract_init;
extractCtrl.schemaAdd = extract_schema_add;
extractCtrl.dh = zh->reg->dh;
extractCtrl.handle = zh;
- extractCtrl.zebra_maps = zh->reg->zebra_maps;
- extractCtrl.flagShowRecords = 0;
- for (i = 0; i<256; i++)
- {
- if (zebra_maps_is_positioned(zh->reg->zebra_maps, i))
- extractCtrl.seqno[i] = 1;
- else
- extractCtrl.seqno[i] = 0;
- }
+ extractCtrl.match_criteria[0] = '\0';
+
+ init_extractCtrl(zh, &extractCtrl);
+
+ extract_set_store_data_prepare(&extractCtrl);
r = (*recType->extract)(clientData, &extractCtrl);
if (r == RECCTRL_EXTRACT_EOF)
- return 0;
+ return ZEBRA_FAIL;
else if (r == RECCTRL_EXTRACT_ERROR_GENERIC)
{
/* error occured during extraction ... */
- yaz_log (LOG_WARN, "extract error: generic");
- return 0;
+ yaz_log (YLOG_WARN, "extract error: generic");
+ return ZEBRA_FAIL;
}
else if (r == RECCTRL_EXTRACT_ERROR_NO_SUCH_FILTER)
{
/* error occured during extraction ... */
- yaz_log (LOG_WARN, "extract error: no such filter");
- return 0;
- }
- if (zh->reg->keys.buf_used == 0)
- {
- /* the extraction process returned no information - the record
- 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");
- return 1;
+ yaz_log (YLOG_WARN, "extract error: no such filter");
+ return ZEBRA_FAIL;
}
/* match criteria */
matchStr = NULL;
+ if (extractCtrl.match_criteria[0])
+ match_criteria = extractCtrl.match_criteria;
+
if (! *sysno) {
char *rinfo;
if (match_criteria && *match_criteria) {
zh->m_record_id);
if (!matchStr)
{
- logf (LOG_WARN, "Bad match criteria (recordID)");
- return 1;
+ yaz_log (YLOG_WARN, "Bad match criteria (recordID)");
+ return ZEBRA_FAIL;
}
}
}
}
}
}
+ if (zh->reg->keys.buf_used == 0)
+ {
+ /* the extraction process returned no information - the record
+ is probably empty - unless flagShowRecords is in use */
+ if (test_mode)
+ return ZEBRA_OK;
+ }
if (! *sysno)
{
/* new record */
if (delete_flag)
{
- logf (LOG_LOG, "delete %s %s %ld", recordType,
- pr_fname, (long) recordOffset);
- logf (LOG_WARN, "cannot delete record above (seems new)");
- return 1;
+ if (show_progress)
+ yaz_log (YLOG_LOG, "delete %s %s %ld", recordType,
+ pr_fname, (long) recordOffset);
+ yaz_log (YLOG_WARN, "cannot delete record above (seems new)");
+ return ZEBRA_FAIL;
}
- logf (LOG_LOG, "add %s %s %ld", recordType, pr_fname,
- (long) recordOffset);
+ if (show_progress)
+ yaz_log (YLOG_LOG, "add %s %s %ld", recordType, pr_fname,
+ (long) recordOffset);
rec = rec_new (zh->reg->records);
*sysno = rec->sysno;
struct recKeys delkeys;
struct sortKeys sortKeys;
- if (!allow_update) {
- logf (LOG_LOG, "skipped %s %s %ld",
- recordType, pr_fname, (long) recordOffset);
- logRecord(zh);
- return -1;
+ if (!allow_update)
+ {
+ if (show_progress)
+ yaz_log (YLOG_LOG, "skipped %s %s %ld",
+ recordType, pr_fname, (long) recordOffset);
+ logRecord(zh);
+ return ZEBRA_FAIL;
}
rec = rec_get (zh->reg->records, *sysno);
if (recordAttr->runNumber ==
zebraExplain_runNumberIncrement (zh->reg->zei, 0))
{
- logf (LOG_LOG, "skipped %s %s %ld", recordType,
- pr_fname, (long) recordOffset);
+ if (show_progress)
+ yaz_log (YLOG_LOG, "skipped %s %s %ld", recordType,
+ pr_fname, (long) recordOffset);
extract_flushSortKeys (zh, *sysno, -1, &zh->reg->sortKeys);
rec_rm (&rec);
logRecord(zh);
- return -1;
+ return ZEBRA_FAIL;
}
}
/* record going to be deleted */
if (!delkeys.buf_used)
{
- logf (LOG_LOG, "delete %s %s %ld", recordType,
- pr_fname, (long) recordOffset);
- logf (LOG_WARN, "cannot delete file above, storeKeys false");
- }
+ if (show_progress)
+ {
+ yaz_log (YLOG_LOG, "delete %s %s %ld", recordType,
+ pr_fname, (long) recordOffset);
+ yaz_log (YLOG_WARN, "cannot delete file above, "
+ "storeKeys false");
+ }
+ }
else
{
- logf (LOG_LOG, "delete %s %s %ld", recordType,
- pr_fname, (long) recordOffset);
+ if (show_progress)
+ yaz_log (YLOG_LOG, "delete %s %s %ld", recordType,
+ pr_fname, (long) recordOffset);
zh->records_deleted++;
if (matchStr)
dict_delete (zh->reg->matchDict, matchStr);
}
rec_rm (&rec);
logRecord(zh);
- return 0;
+ return ZEBRA_OK;
}
else
{
/* record going to be updated */
if (!delkeys.buf_used)
{
- logf (LOG_LOG, "update %s %s %ld", recordType,
- pr_fname, (long) recordOffset);
- logf (LOG_WARN, "cannot update file above, storeKeys false");
- }
+ if (show_progress)
+ {
+ yaz_log (YLOG_LOG, "update %s %s %ld", recordType,
+ pr_fname, (long) recordOffset);
+ yaz_log (YLOG_WARN, "cannot update file above, storeKeys false");
+ }
+ }
else
{
- logf (LOG_LOG, "update %s %s %ld", recordType,
- pr_fname, (long) recordOffset);
+ if (show_progress)
+ 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);
zh->records_updated++;
/* update store data */
xfree (rec->info[recInfo_storeData]);
- if (zh->m_store_data)
+
+ /* update store data */
+ if (zh->store_data_buf)
+ {
+ rec->size[recInfo_storeData] = zh->store_data_size;
+ rec->info[recInfo_storeData] = zh->store_data_buf;
+ zh->store_data_buf = 0;
+ }
+ else if (zh->m_store_data)
{
rec->size[recInfo_storeData] = recordAttr->recordSize;
rec->info[recInfo_storeData] = (char *)
/* commit this record */
rec_put (zh->reg->records, &rec);
logRecord(zh);
- return 0;
+ return ZEBRA_OK;
}
int explain_extract (void *handle, Record rec, data1_node *n)
{
ZebraHandle zh = (ZebraHandle) handle;
struct recExtractCtrl extractCtrl;
- int i;
if (zebraExplain_curDatabase (zh->reg->zei,
rec->info[recInfo_databaseName]))
extractCtrl.tokenAdd = extract_token_add;
extractCtrl.schemaAdd = extract_schema_add;
extractCtrl.dh = zh->reg->dh;
- for (i = 0; i<256; i++)
- extractCtrl.seqno[i] = 0;
- extractCtrl.zebra_maps = zh->reg->zebra_maps;
+
+ init_extractCtrl(zh, &extractCtrl);
+
extractCtrl.flagShowRecords = 0;
+ extractCtrl.match_criteria[0] = '\0';
extractCtrl.handle = handle;
+ extractCtrl.first_record = 1;
+
+ extract_set_store_data_prepare(&extractCtrl);
if (n)
grs_extract_tree(&extractCtrl, n);
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 */
const char *src = reckeys->buf + off;
struct it_key key;
char *dst = (char*) &key;
- int attrSet, attrUse;
iscz1_decode(decode_handle, &dst, &src);
- assert(key.len < 4 && key.len > 2);
-
- attrSet = (int) key.mem[0];
- attrUse = (int) key.mem[1]; /* sequence in mem[2] */
+ assert(key.len == 4);
if (zh->reg->key_buf_used + 1024 >
(zh->reg->ptr_top -zh->reg->ptr_i)*sizeof(char*))
(zh->reg->key_buf)[zh->reg->ptr_top - zh->reg->ptr_i] =
(char*)zh->reg->key_buf + zh->reg->key_buf_used;
- ch = zebraExplain_lookupSU (zei, attrSet, attrUse);
- if (ch < 0)
- ch = zebraExplain_addSU (zei, attrSet, attrUse);
+ ch = (int) key.mem[0]; /* ordinal for field/use/attribute */
- assert (ch > 0);
zh->reg->key_buf_used +=
key_SU_encode (ch,((char*)zh->reg->key_buf) +
zh->reg->key_buf_used);
((char*)(zh->reg->key_buf))[(zh->reg->key_buf_used)++] = '\0';
((char*)(zh->reg->key_buf))[(zh->reg->key_buf_used)++] = cmd;
- key.len = 2;
- key.mem[0] = sysno;
- key.mem[1] = key.mem[2]; /* sequence .. */
-
+ key.len = 3;
+ if (key.mem[1]) /* filter specified record ID */
+ key.mem[0] = key.mem[1];
+ else
+ key.mem[0] = sysno;
+ key.mem[1] = key.mem[2]; /* section_id */
+ key.mem[2] = key.mem[3]; /* sequence .. */
+
memcpy ((char*)zh->reg->key_buf + zh->reg->key_buf_used,
&key, sizeof(key));
(zh->reg->key_buf_used) += sizeof(key);
#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,
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';
}
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);
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)
{
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,
#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;
}
-void extract_add_index_string (RecWord *p, const char *str, int length)
+void extract_add_it_key (ZebraHandle zh,
+ int reg_type,
+ const char *str, int slen, struct it_key *key)
{
char *dst;
- ZebraHandle zh = p->extractCtrl->handle;
struct recKeys *keys = &zh->reg->keys;
- struct it_key key;
- const char *src = (char*) &key;
+ const char *src = (char*) key;
if (keys->buf_used+1024 > keys->buf_max)
{
- char *b;
-
- b = (char *) xmalloc (keys->buf_max += 128000);
+ char *b = (char *) xmalloc (keys->buf_max += 128000);
if (keys->buf_used > 0)
memcpy (b, keys->buf, keys->buf_used);
xfree (keys->buf);
}
dst = keys->buf + keys->buf_used;
- key.len = 3;
- key.mem[0] = p->attrSet;
- key.mem[1] = p->attrUse;
- key.mem[2] = p->seqno;
+ iscz1_encode(keys->codec_handle, &dst, &src);
+
+ *dst++ = reg_type;
+ memcpy (dst, str, slen);
+ dst += slen;
+ *dst++ = '\0';
+ keys->buf_used = dst - keys->buf;
+}
+
+ZEBRA_RES zebra_snippets_rec_keys(ZebraHandle zh, struct recKeys *reckeys,
+ zebra_snippets *snippets)
+{
+ void *decode_handle = iscz1_start();
+ int off = 0;
+ int seqno = 0;
+ NMEM nmem = nmem_create();
+
+ yaz_log(YLOG_LOG, "zebra_rec_keys_snippets buf=%p sz=%d", reckeys->buf,
+ reckeys->buf_used);
+ assert(reckeys->buf);
+ while (off < reckeys->buf_used)
+ {
+ const char *src = reckeys->buf + off;
+ struct it_key key;
+ char *dst = (char*) &key;
+ char dst_buf[IT_MAX_WORD];
+ char *dst_term = dst_buf;
+
+ iscz1_decode(decode_handle, &dst, &src);
+ assert(key.len <= 4 && key.len > 2);
+
+ seqno = (int) key.mem[key.len-1];
+
+ zebra_term_untrans_iconv(zh, nmem, src[0], &dst_term, src+1);
+ zebra_snippets_append(snippets, seqno, src[0], key.mem[0], dst_term);
+ while (*src++)
+ ;
+ off = src - reckeys->buf;
+ nmem_reset(nmem);
+ }
+ nmem_destroy(nmem);
+ iscz1_stop(decode_handle);
+ return ZEBRA_OK;
+}
+
+void print_rec_keys(ZebraHandle zh, struct recKeys *reckeys)
+{
+ void *decode_handle = iscz1_start();
+ int off = 0;
+ int seqno = 0;
+ NMEM nmem = nmem_create();
+
+ yaz_log(YLOG_LOG, "print_rec_keys buf=%p sz=%d", reckeys->buf,
+ reckeys->buf_used);
+ assert(reckeys->buf);
+ while (off < reckeys->buf_used)
+ {
+ const char *src = reckeys->buf + off;
+ struct it_key key;
+ char *dst = (char*) &key;
+ int attrSet, attrUse;
+ char dst_buf[IT_MAX_WORD];
+ char *dst_term = dst_buf;
+
+ iscz1_decode(decode_handle, &dst, &src);
+ assert(key.len <= 4 && key.len > 2);
+
+ attrSet = (int) key.mem[0] >> 16;
+ attrUse = (int) key.mem[0] & 65535;
+ seqno = (int) key.mem[key.len-1];
+
+ zebra_term_untrans_iconv(zh, nmem, src[0], &dst_term, src+1);
+
+ yaz_log(YLOG_LOG, "ord=" ZINT_FORMAT " seqno=%d term=%s",
+ key.mem[0], seqno, dst_term);
+ while (*src++)
+ ;
+ off = src - reckeys->buf;
+ nmem_reset(nmem);
+ }
+ nmem_destroy(nmem);
+ iscz1_stop(decode_handle);
+}
+
+void extract_add_index_string (RecWord *p, const char *str, int length)
+{
+ struct it_key key;
+
+ ZebraHandle zh = p->extractCtrl->handle;
+ ZebraExplainInfo zei = zh->reg->zei;
+ int ch;
+
+ if (p->attrStr)
+ {
+ ch = zebraExplain_lookup_attr_str(zei, p->attrStr);
+ if (ch < 0)
+ ch = zebraExplain_add_attr_str(zei, p->attrStr);
+ }
+ else
+ {
+ ch = zebraExplain_lookup_attr_su(zei, p->attrSet, p->attrUse);
+ if (ch < 0)
+ ch = zebraExplain_add_attr_su(zei, p->attrSet, p->attrUse);
+ }
+ key.len = 4;
+ key.mem[0] = ch;
+ key.mem[1] = p->record_id;
+ key.mem[2] = p->section_id;
+ key.mem[3] = p->seqno;
#if 0
/* just for debugging .. */
- yaz_log(LOG_LOG, "set=%d use=%d seqno=%d", p->attrSet, p->attrUse,
- p->seqno);
+ yaz_log(YLOG_LOG, "add: set=%d use=%d "
+ "record_id=%lld section_id=%lld seqno=%lld",
+ p->attrSet, p->attrUse, p->record_id, p->section_id, p->seqno);
#endif
- iscz1_encode(keys->codec_handle, &dst, &src);
-
- *dst++ = p->reg_type;
- memcpy (dst, str, length);
- dst += length;
- *dst++ = '\0';
- keys->buf_used = dst - keys->buf;
+ extract_add_it_key(p->extractCtrl->handle, p->reg_type, str,
+ length, &key);
}
static void extract_add_sort_string (RecWord *p, const char *str,
static void extract_add_incomplete_field (RecWord *p)
{
- const char *b = p->string;
- int remain = p->length;
+ const char *b = p->term_buf;
+ int remain = p->term_len;
const char **map = 0;
+
+ yaz_log(YLOG_DEBUG, "Incomplete field, w='%.*s'", p->term_len, p->term_buf);
if (remain > 0)
map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, 0);
/* Skip spaces */
while (map && *map && **map == *CHR_SPACE)
{
- remain = p->length - (b - p->string);
+ remain = p->term_len - (b - p->term_buf);
if (remain > 0)
map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, 0);
else
while (i < IT_MAX_WORD && *cp)
buf[i++] = *(cp++);
- remain = p->length - (b - p->string);
+ remain = p->term_len - (b - p->term_buf);
if (remain > 0)
map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, 0);
else
static void extract_add_complete_field (RecWord *p)
{
- const char *b = p->string;
+ const char *b = p->term_buf;
char buf[IT_MAX_WORD+1];
const char **map = 0;
- int i = 0, remain = p->length;
- int first; /* first position */
+ int i = 0, remain = p->term_len;
-yaz_log(LOG_DEBUG, "Complete field, w='%.*s'", p->length, p->string);
+ yaz_log(YLOG_DEBUG, "Complete field, w='%.*s'",
+ p->term_len, p->term_buf);
if (remain > 0)
map = zebra_maps_input (p->zebra_maps, p->reg_type, &b, remain, 1);
{
while (map && *map && **map == *CHR_SPACE)
{
- remain = p->length - (b - p->string);
+ remain = p->term_len - (b - p->term_buf);
if (remain > 0)
{
- first = i ? 0 : 1;
+ int first = i ? 0 : 1; /* first position */
map = zebra_maps_input(p->zebra_maps, p->reg_type, &b, remain, first);
}
else
{
if (i >= IT_MAX_WORD)
break;
- yaz_log(LOG_DEBUG, "Adding string to index '%d'", **map);
+ 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);
+ remain = p->term_len - (b - p->term_buf);
if (remain > 0)
{
map = zebra_maps_input (p->zebra_maps, p->reg_type, &b,
{
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);
#endif
if ((wrbuf = zebra_replace(p->zebra_maps, p->reg_type, 0,
- p->string, p->length)))
+ p->term_buf, p->term_len)))
{
- p->string = wrbuf_buf(wrbuf);
- p->length = wrbuf_len(wrbuf);
+ p->term_buf = wrbuf_buf(wrbuf);
+ p->term_len = wrbuf_len(wrbuf);
}
if (zebra_maps_is_complete (p->zebra_maps, p->reg_type))
extract_add_complete_field (p);
extract_add_incomplete_field(p);
}
+static void extract_set_store_data_cb(struct recExtractCtrl *p,
+ void *buf, size_t sz)
+{
+ ZebraHandle zh = (ZebraHandle) p->handle;
+
+ xfree(zh->store_data_buf);
+ zh->store_data_buf = 0;
+ zh->store_data_size = 0;
+ if (buf && sz)
+ {
+ zh->store_data_buf = xmalloc(sz);
+ zh->store_data_size = sz;
+ memcpy(zh->store_data_buf, buf, sz);
+ }
+}
+
+static void extract_set_store_data_prepare(struct recExtractCtrl *p)
+{
+ ZebraHandle zh = (ZebraHandle) p->handle;
+ xfree(zh->store_data_buf);
+ zh->store_data_buf = 0;
+ zh->store_data_size = 0;
+ p->setStoreData = extract_set_store_data_cb;
+}
+
void extract_schema_add (struct recExtractCtrl *p, Odr_oid *oid)
{
- ZebraHandle zh = (ZebraHandle) (p->handle);
+ ZebraHandle zh = (ZebraHandle) p->handle;
zebraExplain_addSchema (zh->reg->zei, oid);
}
*bp0 = (*k * 128) + bp - bp0 - 1; /* length and insert/delete combined */
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->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 */
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 */
projects / idzebra-moved-to-github.git / blobdiff