X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fextract.c;h=5be148815d0c6c1c593b9f652ef5a71f48211f21;hb=cd405b18db3526be1855d7e14a598616237a4b2e;hp=c70e6a1ee26b25b37476e21f23e0f53442ca11ba;hpb=e199777080c6fa0963d51b7df1763fd5286ca9a4;p=idzebra-moved-to-github.git diff --git a/index/extract.c b/index/extract.c index c70e6a1..5be1488 100644 --- a/index/extract.c +++ b/index/extract.c @@ -1,4 +1,4 @@ -/* $Id: extract.c,v 1.267 2007-10-31 16:56:14 adam Exp $ +/* $Id: extract.c,v 1.270 2007-11-30 12:19:08 adam Exp $ Copyright (C) 1995-2007 Index Data ApS @@ -71,7 +71,6 @@ static void extract_flush_sort_keys(ZebraHandle zh, zint sysno, int cmd, zebra_rec_keys_t skp); static void extract_schema_add(struct recExtractCtrl *p, Odr_oid *oid); static void extract_token_add(RecWord *p); -static void extract_token_add2(RecWord *p); static void check_log_limit(ZebraHandle zh) { @@ -521,13 +520,29 @@ struct recordLogInfo { struct recordGroup *rGroup; }; -static void all_matches_add(struct recExtractCtrl *ctrl) +/** \brief add the always-matches index entry and map to real record ID + \param ctrl record control + \param record_id custom record ID + \param sysno system record ID + + This function serves two purposes.. It adds the always matches + entry and makes a pointer from the custom record ID (if defined) + back to the system record ID (sysno) + See zebra_recid_to_sysno . + */ +static void all_matches_add(struct recExtractCtrl *ctrl, zint record_id, + zint sysno) { RecWord word; extract_init(ctrl, &word); + word.record_id = record_id; + /* we use the seqno as placeholder for a way to get back to + record database from _ALLRECORDS.. This is used if a custom + RECORD was defined */ + word.seqno = sysno; word.index_name = "_ALLRECORDS"; word.index_type = "w"; - word.seqno = 1; + extract_add_index_string(&word, zinfo_index_category_alwaysmatches, "", 0); } @@ -821,14 +836,7 @@ ZEBRA_RES zebra_extract_record_stream(ZebraHandle zh, stream->endf(stream, &null_offset);; extractCtrl.init = extract_init; - if (zh->reg->index_types) - { - extractCtrl.tokenAdd = extract_token_add2; - } - else - { - extractCtrl.tokenAdd = extract_token_add; - } + extractCtrl.tokenAdd = extract_token_add; extractCtrl.schemaAdd = extract_schema_add; extractCtrl.dh = zh->reg->dh; extractCtrl.handle = zh; @@ -882,8 +890,6 @@ ZEBRA_RES zebra_extract_record_stream(ZebraHandle zh, else end_offset = stream->tellf(stream); - all_matches_add(&extractCtrl); - if (extractCtrl.match_criteria[0]) match_criteria = extractCtrl.match_criteria; } @@ -927,6 +933,7 @@ ZEBRA_RES zebra_extract_record_stream(ZebraHandle zh, } } } + if (zebra_rec_keys_empty(zh->reg->keys)) { /* the extraction process returned no information - the record @@ -959,6 +966,15 @@ ZEBRA_RES zebra_extract_record_stream(ZebraHandle zh, *sysno = rec->sysno; + + if (stream) + { + all_matches_add(&extractCtrl, + zebra_rec_keys_get_custom_record_id(zh->reg->keys), + *sysno); + } + + recordAttr = rec_init_attr(zh->reg->zei, rec); if (extractCtrl.staticrank < 0) { @@ -1000,6 +1016,13 @@ ZEBRA_RES zebra_extract_record_stream(ZebraHandle zh, rec = rec_get(zh->reg->records, *sysno); assert(rec); + + if (stream) + { + all_matches_add(&extractCtrl, + zebra_rec_keys_get_custom_record_id(zh->reg->keys), + *sysno); + } recordAttr = rec_init_attr(zh->reg->zei, rec); @@ -1700,14 +1723,14 @@ static void extract_add_complete_field(RecWord *p, zebra_map_t zm) extract_add_string(p, zm, buf, i); } -static void extract_token_add2_index(ZebraHandle zh, zebra_index_type_t type, - RecWord *p) +static void extract_add_icu(RecWord *p, zebra_map_t zm) { struct it_key key; const char *res_buf = 0; size_t res_len = 0; - int r = zebra_index_type_tokenize(type, p->term_buf, p->term_len, - &res_buf, &res_len); + ZebraHandle zh = p->extractCtrl->handle; + int r = zebra_map_tokenize(zm, p->term_buf, p->term_len, + &res_buf, &res_len); int cat = zinfo_index_category_index; int ch = zebraExplain_lookup_attr_str(zh->reg->zei, cat, p->index_type, p->index_name); if (ch < 0) @@ -1724,31 +1747,13 @@ static void extract_token_add2_index(ZebraHandle zh, zebra_index_type_t type, key.mem[i++] = p->seqno; key.len = i; - yaz_log(YLOG_LOG, "keys_write %.*s", (int) res_len, res_buf); zebra_rec_keys_write(zh->reg->keys, res_buf, res_len, &key); p->seqno++; - r = zebra_index_type_tokenize(type, 0, 0, &res_buf, &res_len); + r = zebra_map_tokenize(zm, 0, 0, &res_buf, &res_len); } } -static void extract_token_add2(RecWord *p) -{ - ZebraHandle zh = p->extractCtrl->handle; - zebra_index_type_t type = zebra_index_type_get(zh->reg->index_types, p->index_type); - if (type) - { - if (zebra_index_type_is_index(type)) - { - extract_token_add2_index(zh, type, p); - } - else if (zebra_index_type_is_sort(type)) - { - ; - - } - } -} /** \brief top-level indexing handler for recctrl system \param p token data to be indexed @@ -1780,13 +1785,20 @@ static void extract_token_add(RecWord *p) } if ((wrbuf = zebra_replace(zm, 0, p->term_buf, p->term_len))) { - p->term_buf = wrbuf_buf(wrbuf); - p->term_len = wrbuf_len(wrbuf); + p->term_buf = wrbuf_buf(wrbuf); + p->term_len = wrbuf_len(wrbuf); + } + if (zebra_maps_is_icu(zm)) + { + extract_add_icu(p, zm); } - if (zebra_maps_is_complete(zm)) - extract_add_complete_field(p, zm); else - extract_add_incomplete_field(p, zm); + { + if (zebra_maps_is_complete(zm)) + extract_add_complete_field(p, zm); + else + extract_add_incomplete_field(p, zm); + } } static void extract_set_store_data_cb(struct recExtractCtrl *p,