-/* $Id: extract.c,v 1.244 2006-12-05 08:14:47 adam Exp $
- Copyright (C) 1995-2006
+/* $Id: extract.c,v 1.252 2007-03-14 11:48:32 adam Exp $
+ Copyright (C) 1995-2007
Index Data ApS
This file is part of the Zebra server.
static void extract_schema_add (struct recExtractCtrl *p, Odr_oid *oid);
static void extract_token_add (RecWord *p);
+static void check_log_limit(ZebraHandle zh)
+{
+ if (zh->records_processed + zh->records_skipped == zh->m_file_verbose_limit)
+ {
+ yaz_log(YLOG_LOG, "More than %d file log entries. Omitting rest",
+ zh->m_file_verbose_limit);
+ }
+}
+
static void logRecord (ZebraHandle zh)
{
+ check_log_limit(zh);
++zh->records_processed;
if (!(zh->records_processed % 1000))
{
"", 0);
}
+ZEBRA_RES zebra_extract_records_stream(ZebraHandle zh,
+ struct ZebraRecStream *stream,
+ enum zebra_recctrl_action_t action,
+ int test_mode,
+ const char *recordType,
+ zint *sysno,
+ const char *match_criteria,
+ const char *fname,
+ RecType recType,
+ void *recTypeClientData);
+
+
ZEBRA_RES zebra_extract_file(ZebraHandle zh, zint *sysno, const char *fname,
int deleteFlag)
{
}
if (!zh->m_record_type)
{
- if (zh->records_processed < zh->m_file_verbose_limit)
+ check_log_limit(zh);
+ if (zh->records_processed + zh->records_skipped
+ < zh->m_file_verbose_limit)
yaz_log (YLOG_LOG, "? %s", fname);
+ zh->records_skipped++;
return 0;
}
/* determine match criteria */
streamp = &stream;
zebra_create_stream_fd(streamp, fd, 0);
}
- while(1)
- {
- int more = 0;
- r = zebra_extract_record_stream(zh, streamp,
- deleteFlag,
- 0, /* tst_mode */
- zh->m_record_type,
- sysno,
- 0, /*match_criteria */
- fname,
- 1, /* force_update */
- 1, /* allow_update */
- recType, recTypeClientData, &more);
- if (!more)
- break;
- if (sysno)
- {
- break;
- }
- }
+ r = zebra_extract_records_stream(zh, streamp,
+ deleteFlag ?
+ action_delete : action_update,
+ 0, /* tst_mode */
+ zh->m_record_type,
+ sysno,
+ 0, /*match_criteria */
+ fname,
+ recType, recTypeClientData);
if (streamp)
stream.destroy(streamp);
zh->m_record_type = original_record_type;
ZEBRA_RES zebra_buffer_extract_record(ZebraHandle zh,
const char *buf, size_t buf_size,
- int delete_flag,
+ enum zebra_recctrl_action_t action,
int test_mode,
const char *recordType,
zint *sysno,
const char *match_criteria,
- const char *fname,
- int force_update,
- int allow_update)
+ const char *fname)
{
struct ZebraRecStream stream;
ZEBRA_RES res;
void *clientData;
RecType recType = 0;
- int more = 0;
if (recordType && *recordType)
{
zebra_create_stream_mem(&stream, buf, buf_size);
- res = zebra_extract_record_stream(zh, &stream,
- delete_flag,
- test_mode,
- recordType,
- sysno,
- match_criteria,
- fname,
- force_update,
- allow_update,
- recType, clientData, &more);
+ res = zebra_extract_records_stream(zh, &stream,
+ action,
+ test_mode,
+ recordType,
+ sysno,
+ match_criteria,
+ fname,
+ recType, clientData);
stream.destroy(&stream);
return res;
}
+ZEBRA_RES zebra_extract_records_stream(ZebraHandle zh,
+ struct ZebraRecStream *stream,
+ enum zebra_recctrl_action_t action,
+ int test_mode,
+ const char *recordType,
+ zint *sysno,
+ const char *match_criteria,
+ const char *fname,
+ RecType recType,
+ void *recTypeClientData)
+{
+ ZEBRA_RES res = ZEBRA_OK;
+ while (1)
+ {
+ int more = 0;
+ res = zebra_extract_record_stream(zh, stream,
+ action,
+ test_mode,
+ recordType,
+ sysno,
+ match_criteria,
+ fname,
+ recType, recTypeClientData, &more);
+ if (!more)
+ {
+ res = ZEBRA_OK;
+ break;
+ }
+ if (res != ZEBRA_OK)
+ break;
+ if (sysno)
+ break;
+ }
+ return res;
+}
+
ZEBRA_RES zebra_extract_record_stream(ZebraHandle zh,
struct ZebraRecStream *stream,
- int delete_flag,
+ enum zebra_recctrl_action_t action,
int test_mode,
const char *recordType,
zint *sysno,
const char *match_criteria,
const char *fname,
- int force_update,
- int allow_update,
RecType recType,
void *recTypeClientData,
int *more)
Record rec;
off_t start_offset = 0, end_offset = 0;
const char *pr_fname = fname; /* filename to print .. */
- int show_progress = zh->records_processed < zh->m_file_verbose_limit ? 1:0;
+ int show_progress = zh->records_processed + zh->records_skipped
+ < zh->m_file_verbose_limit ? 1:0;
zebra_init_log_level();
extract_set_store_data_prepare(&extractCtrl);
r = (*recType->extract)(recTypeClientData, &extractCtrl);
-
- if (r == RECCTRL_EXTRACT_EOF)
- return ZEBRA_FAIL;
- else if (r == RECCTRL_EXTRACT_ERROR_GENERIC)
+
+ switch (r)
{
+ case RECCTRL_EXTRACT_EOF:
+ return ZEBRA_FAIL;
+ case RECCTRL_EXTRACT_ERROR_GENERIC:
/* error occured during extraction ... */
yaz_log (YLOG_WARN, "extract error: generic");
return ZEBRA_FAIL;
- }
- else if (r == RECCTRL_EXTRACT_ERROR_NO_SUCH_FILTER)
- {
+ case RECCTRL_EXTRACT_ERROR_NO_SUCH_FILTER:
/* error occured during extraction ... */
yaz_log (YLOG_WARN, "extract error: no such filter");
return ZEBRA_FAIL;
+ case RECCTRL_EXTRACT_SKIP:
+ if (show_progress)
+ yaz_log (YLOG_LOG, "skip %s %s " ZINT_FORMAT,
+ recordType, pr_fname, (zint) start_offset);
+ *more = 1;
+
+ end_offset = stream->endf(stream, 0);
+ if (end_offset)
+ stream->seekf(stream, end_offset);
+
+ return ZEBRA_OK;
+ case RECCTRL_EXTRACT_OK:
+ break;
+ default:
+ yaz_log (YLOG_WARN, "extract error: unknown error: %d", r);
+ return ZEBRA_FAIL;
}
-
+ end_offset = stream->endf(stream, 0);
+ if (end_offset)
+ stream->seekf(stream, end_offset);
+ else
+ end_offset = stream->tellf(stream);
+
all_matches_add(&extractCtrl);
if (extractCtrl.match_criteria[0])
match_criteria = extractCtrl.match_criteria;
-
-
- end_offset = stream->endf(stream, 0);
-
- if (!end_offset)
- end_offset = stream->tellf(stream);
- else
- stream->seekf(stream, end_offset);
-
}
-
*more = 1;
if (!sysno)
{
if (! *sysno)
{
/* new record */
- if (delete_flag)
+ if (action == action_delete)
{
yaz_log (YLOG_LOG, "delete %s %s " ZINT_FORMAT, recordType,
pr_fname, (zint) start_offset);
yaz_log (YLOG_WARN, "cannot delete record above (seems new)");
return ZEBRA_FAIL;
}
+ else if (action == action_replace)
+ {
+ yaz_log (YLOG_LOG, "update %s %s " ZINT_FORMAT, recordType,
+ pr_fname, (zint) start_offset);
+ yaz_log (YLOG_WARN, "cannot update record above (seems new)");
+ return ZEBRA_FAIL;
+ }
if (show_progress)
yaz_log (YLOG_LOG, "add %s %s " ZINT_FORMAT, recordType, pr_fname,
(zint) start_offset);
/* record already exists */
zebra_rec_keys_t delkeys = zebra_rec_keys_open();
zebra_rec_keys_t sortKeys = zebra_rec_keys_open();
- if (!allow_update)
+ if (action == action_insert)
{
yaz_log (YLOG_LOG, "skipped %s %s " ZINT_FORMAT,
recordType, pr_fname, (zint) start_offset);
extract_flush_sort_keys(zh, *sysno, 0, sortKeys);
extract_flush_record_keys(zh, *sysno, 0, delkeys,
recordAttr->staticrank);
- if (delete_flag)
+ if (action == action_delete)
{
/* record going to be deleted */
if (zebra_rec_keys_empty(delkeys))
return ZEBRA_OK;
}
else
- {
+ { /* update or special_update */
if (show_progress)
yaz_log(YLOG_LOG, "update %s %s " ZINT_FORMAT, recordType,
pr_fname, (zint) start_offset);
struct it_key key;
while (zebra_rec_keys_read(reckeys, &str, &slen, &key))
{
- char dst_buf[IT_MAX_WORD];
- char *dst_term = dst_buf;
+ char *dst_term = 0;
int ord;
zint seqno;
int index_type;
zebra_rec_keys_write(zh->reg->sortKeys, str, length, &key);
}
+static void extract_add_staticrank_string(RecWord *p,
+ const char *str, int length)
+{
+ char valz[40];
+ struct recExtractCtrl *ctrl = p->extractCtrl;
+
+ if (length > sizeof(valz)-1)
+ length = sizeof(valz)-1;
+
+ memcpy(valz, str, length);
+ valz[length] = '\0';
+ ctrl->staticrank = atozint(valz);
+}
+
static void extract_add_string(RecWord *p, const char *string, int length)
{
ZebraHandle zh = p->extractCtrl->handle;
if (!p->index_name)
return;
- if (zebra_maps_is_sort(zh->reg->zebra_maps, p->index_type))
- extract_add_sort_string(p, string, length);
- else
+ if (zebra_maps_is_index(zh->reg->zebra_maps, p->index_type))
{
extract_add_index_string(p, zinfo_index_category_index,
string, length);
&word, zinfo_index_category_alwaysmatches, "", 0);
}
}
+ else if (zebra_maps_is_sort(zh->reg->zebra_maps, p->index_type))
+ {
+ extract_add_sort_string(p, string, length);
+ }
+ else if (zebra_maps_is_staticrank(zh->reg->zebra_maps, p->index_type))
+ {
+ extract_add_staticrank_string(p, string, length);
+ }
}
static void extract_add_incomplete_field(RecWord *p)
void extract_flush_sort_keys(ZebraHandle zh, zint sysno,
int cmd, zebra_rec_keys_t reckeys)
{
+#if 0
+ yaz_log(YLOG_LOG, "extract_flush_sort_keys cmd=%d sysno=" ZINT_FORMAT,
+ cmd, sysno);
+ extract_rec_keys_log(zh, cmd, reckeys, YLOG_LOG);
+#endif
+
if (zebra_rec_keys_rewind(reckeys))
{
- SortIdx sortIdx = zh->reg->sortIdx;
+ zebra_sort_index_t si = zh->reg->sort_index;
size_t slen;
const char *str;
struct it_key key_in;
- sortIdx_sysno (sortIdx, sysno);
+ zebra_sort_sysno(si, sysno);
while (zebra_rec_keys_read(reckeys, &str, &slen, &key_in))
{
int ord = CAST_ZINT_TO_INT(key_in.mem[0]);
- sortIdx_type(sortIdx, ord);
+ zebra_sort_type(si, ord);
if (cmd == 1)
- sortIdx_add(sortIdx, str, slen);
+ zebra_sort_add(si, str, slen);
else
- sortIdx_add(sortIdx, "", 1);
+ zebra_sort_delete(si);
}
}
}
projects / idzebra-moved-to-github.git / blobdiff