1 <chapter id="querymodel">
2 <!-- $Id: querymodel.xml,v 1.27 2006-11-30 10:33:19 adam Exp $ -->
3 <title>Query Model</title>
5 <section id="querymodel-overview">
6 <title>Query Model Overview</title>
8 <section id="querymodel-query-languages">
9 <title>Query Languages</title>
12 Zebra is born as a networking Information Retrieval engine adhering
13 to the international standards
14 <ulink url="&url.z39.50;">Z39.50</ulink> and
15 <ulink url="&url.sru;">SRU</ulink>,
17 type-1 Reverse Polish Notation (RPN) query
19 Unfortunately, this model has only defined a binary
20 encoded representation, which is used as transport packaging in
21 the Z39.50 protocol layer. This representation is not human
22 readable, nor defines any convenient way to specify queries.
25 Since the type-1 (RPN)
26 query structure has no direct, useful string
27 representation, every client application needs to provide some
28 form of mapping from a local query notation or representation to it.
32 <section id="querymodel-query-languages-pqf">
33 <title>Prefix Query Format (PQF)</title>
35 Index Data has defined a textual representation in the
36 <ulink url="&url.yaz.pqf;">Prefix Query Format</ulink>, short
37 <emphasis>PQF</emphasis>, which maps
38 one-to-one to binary encoded
39 <emphasis>type-1 RPN</emphasis> queries.
40 PQF has been adopted by other
41 parties developing Z39.50 software, and is often referred to as
42 <emphasis>Prefix Query Notation</emphasis>, or in short
44 <xref linkend="querymodel-rpn"/> for further explanations and
45 descriptions of Zebra's capabilities.
49 <section id="querymodel-query-languages-cql">
50 <title>Common Query Language (CQL)</title>
52 The query model of the type-1 RPN,
53 expressed in PQF/PQN is natively supported.
54 On the other hand, the default SRU
55 web services <emphasis>Common Query Language</emphasis>
56 <ulink url="&url.cql;">CQL</ulink> is not natively supported.
59 Zebra can be configured to understand and map CQL to PQF. See
60 <xref linkend="querymodel-cql-to-pqf"/>.
66 <section id="querymodel-operation-types">
67 <title>Operation types</title>
69 Zebra supports all of the three different
70 Z39.50/SRU operations defined in the
71 standards: explain, search,
72 and scan. A short description of the
73 functionality and purpose of each is quite in order here.
76 <section id="querymodel-operation-type-explain">
77 <title>Explain Operation</title>
79 The <emphasis>syntax</emphasis> of Z39.50/SRU queries is
80 well known to any client, but the specific
81 <emphasis>semantics</emphasis> - taking into account a
82 particular servers functionalities and abilities - must be
83 discovered from case to case. Enters the
84 explain operation, which provides the means for learning which
85 <emphasis>fields</emphasis> (also called
86 <emphasis>indexes</emphasis> or <emphasis>access points</emphasis>)
87 are provided, which default parameter the server uses, which
88 retrieve document formats are defined, and which specific parts
89 of the general query model are supported.
92 The Z39.50 embeds the explain operation
95 <literal>IR-Explain-1</literal> database;
96 see <xref linkend="querymodel-exp1"/>.
99 In SRU, explain is an entirely separate
100 operation, which returns an ZeeRex XML record according to the
101 structure defined by the protocol.
104 In both cases, the information gathered through
105 explain operations can be used to
106 auto-configure a client user interface to the servers
111 <section id="querymodel-operation-type-search">
112 <title>Search Operation</title>
114 Search and retrieve interactions are the raison d'ĂȘtre.
115 They are used to query the remote database and
116 return search result documents. Search queries span from
117 simple free text searches to nested complex boolean queries,
118 targeting specific indexes, and possibly enhanced with many
119 query semantic specifications. Search interactions are the heart
120 and soul of Z39.50/SRU servers.
124 <section id="querymodel-operation-type-scan">
125 <title>Scan Operation</title>
127 The scan operation is a helper functionality,
128 which operates on one index or access point a time.
132 the means to investigate the content of specific indexes.
133 Scanning an index returns a handful of terms actually found in
134 the indexes, and in addition the scan
135 operation returns the number of documents indexed by each term.
136 A search client can use this information to propose proper
137 spelling of search terms, to auto-fill search boxes, or to
138 display controlled vocabularies.
147 <section id="querymodel-rpn">
148 <title>RPN queries and semantics</title>
150 The <ulink url="&url.yaz.pqf;">PQF grammar</ulink>
151 is documented in the YAZ manual, and shall not be
152 repeated here. This textual PQF representation
153 is not transmistted to Zebra during search, but it is in the
154 client mapped to the equivalent Z39.50 binary
158 <section id="querymodel-rpn-tree">
159 <title>RPN tree structure</title>
161 The RPN parse tree - or the equivalent textual representation in PQF -
162 may start with one specification of the
163 <emphasis>attribute set</emphasis> used. Following is a query
165 consists of <emphasis>atomic query parts (APT)</emphasis> or
166 <emphasis>named result sets</emphasis>, eventually
167 paired by <emphasis>boolean binary operators</emphasis>, and
168 finally <emphasis>recursively combined </emphasis> into
172 <section id="querymodel-attribute-sets">
173 <title>Attribute sets</title>
175 Attribute sets define the exact meaning and semantics of queries
176 issued. Zebra comes with some predefined attribute set
177 definitions, others can easily be defined and added to the
181 <table id="querymodel-attribute-sets-table" frame="top">
182 <title>Attribute sets predefined in Zebra</title>
186 <entry>Attribute set</entry>
187 <entry>PQF notation (Short hand)</entry>
188 <entry>Status</entry>
195 <entry>Explain</entry>
196 <entry><literal>exp-1</literal></entry>
197 <entry>Special attribute set used on the special automagic
198 <literal>IR-Explain-1</literal> database to gain information on
199 server capabilities, database names, and database
200 and semantics.</entry>
201 <entry>predefined</entry>
205 <entry><literal>bib-1</literal></entry>
206 <entry>Standard PQF query language attribute set which defines the
207 semantics of Z39.50 searching. In addition, all of the
208 non-use attributes (types 2-12) define the hard-wired
211 <entry>default</entry>
215 <entry><literal>gils</literal></entry>
216 <entry>Extension to the Bib-1 attribute set.</entry>
217 <entry>predefined</entry>
221 <entry>IDXPATH</entry>
222 <entry><literal>idxpath</literal></entry>
223 <entry>Hardwired XPATH like attribute set, only available for
224 indexing with the GRS record model</entry>
225 <entry>deprecated</entry>
233 The use attributes (type 1) mappings the
234 predefined attribute sets are found in the
235 attribute set configuration files <filename>tab/*.att</filename>.
240 The Zebra internal query processing is modeled after
241 the Bib-1 attribute set, and the non-use
242 attributes type 2-6 are hard-wired in. It is therefore essential
243 to be familiar with <xref linkend="querymodel-bib1-nonuse"/>.
249 <section id="querymodel-boolean-operators">
250 <title>Boolean operators</title>
252 A pair of sub query trees, or of atomic queries, is combined
253 using the standard boolean operators into new query trees.
254 Thus, boolean operators are always internal nodes in the query tree.
257 <table id="querymodel-boolean-operators-table" frame="top">
258 <title>Boolean operators</title>
262 <entry>Keyword</entry>
263 <entry>Operator</entry>
264 <entry>Description</entry>
268 <row><entry><literal>@and</literal></entry>
269 <entry>binary AND operator</entry>
270 <entry>Set intersection of two atomic queries hit sets</entry>
272 <row><entry><literal>@or</literal></entry>
273 <entry>binary OR operator</entry>
274 <entry>Set union of two atomic queries hit sets</entry>
276 <row><entry><literal>@not</literal></entry>
277 <entry>binary AND NOT operator</entry>
278 <entry>Set complement of two atomic queries hit sets</entry>
280 <row><entry><literal>@prox</literal></entry>
281 <entry>binary PROXIMITY operator</entry>
282 <entry>Set intersection of two atomic queries hit sets. In
283 addition, the intersection set is purged for all
284 documents which do not satisfy the requested query
285 term proximity. Usually a proper subset of the AND
293 For example, we can combine the terms
294 <emphasis>information</emphasis> and <emphasis>retrieval</emphasis>
295 into different searches in the default index of the default
296 attribute set as follows.
297 Querying for the union of all documents containing the
298 terms <emphasis>information</emphasis> OR
299 <emphasis>retrieval</emphasis>:
301 Z> find @or information retrieval
305 Querying for the intersection of all documents containing the
306 terms <emphasis>information</emphasis> AND
307 <emphasis>retrieval</emphasis>:
308 The hit set is a subset of the corresponding
311 Z> find @and information retrieval
315 Querying for the intersection of all documents containing the
316 terms <emphasis>information</emphasis> AND
317 <emphasis>retrieval</emphasis>, taking proximity into account:
318 The hit set is a subset of the corresponding
320 (see the <ulink url="&url.yaz.pqf;">PQF grammar</ulink> for
321 details on the proximity operator):
323 Z> find @prox 0 3 0 2 k 2 information retrieval
327 Querying for the intersection of all documents containing the
328 terms <emphasis>information</emphasis> AND
329 <emphasis>retrieval</emphasis>, in the same order and near each
330 other as described in the term list.
331 The hit set is a subset of the corresponding
334 Z> find "information retrieval"
340 <section id="querymodel-atomic-queries">
341 <title>Atomic queries (APT)</title>
343 Atomic queries are the query parts which work on one access point
344 only. These consist of <emphasis>an attribute list</emphasis>
345 followed by a <emphasis>single term</emphasis> or a
346 <emphasis>quoted term list</emphasis>, and are often called
347 <emphasis>Attributes-Plus-Terms (APT)</emphasis> queries.
350 Atomic (APT) queries are always leaf nodes in the PQF query tree.
351 UN-supplied non-use attributes types 2-12 are either inherited from
352 higher nodes in the query tree, or are set to Zebra's default values.
353 See <xref linkend="querymodel-bib1"/> for details.
356 <table id="querymodel-atomic-queries-table" frame="top">
357 <title>Atomic queries (APT)</title>
368 <entry><emphasis>attribute list</emphasis></entry>
369 <entry>List of <emphasis>orthogonal</emphasis> attributes</entry>
370 <entry>Any of the orthogonal attribute types may be omitted,
371 these are inherited from higher query tree nodes, or if not
372 inherited, are set to the default Zebra configuration values.
376 <entry><emphasis>term</emphasis></entry>
377 <entry>single <emphasis>term</emphasis>
378 or <emphasis>quoted term list</emphasis> </entry>
379 <entry>Here the search terms or list of search terms is added
386 Querying for the term <emphasis>information</emphasis> in the
387 default index using the default attribute set, the server choice
388 of access point/index, and the default non-use attributes.
394 Equivalent query fully specified including all default values:
396 Z> find @attrset bib-1 @attr 1=1017 @attr 2=3 @attr 3=3 @attr 4=1 @attr 5=100 @attr 6=1 information
401 Finding all documents which have the term
402 <emphasis>debussy</emphasis> in the title field.
404 Z> find @attr 1=4 debussy
409 The <emphasis>scan</emphasis> operation is only supported with
410 atomic APT queries, as it is bound to one access point at a
411 time. Boolean query trees are not allowed during
412 <emphasis>scan</emphasis>.
416 For example, we might want to scan the title index, starting with
418 <emphasis>debussy</emphasis>, and displaying this and the
419 following terms in lexicographic order:
421 Z> scan @attr 1=4 debussy
427 <section id="querymodel-resultset">
428 <title>Named Result Sets</title>
430 Named result sets are supported in Zebra, and result sets can be
431 used as operands without limitations. It follows that named
432 result sets are leaf nodes in the PQF query tree, exactly as
433 atomic APT queries are.
436 After the execution of a search, the result set is available at
437 the server, such that the client can use it for subsequent
438 searches or retrieval requests. The Z30.50 standard actually
439 stresses the fact that result sets are volatile. It may cease
440 to exist at any time point after search, and the server will
441 send a diagnostic to the effect that the requested
442 result set does not exist any more.
446 Defining a named result set and re-using it in the next query,
447 using <application>yaz-client</application>. Notice that the client, not
448 the server, assigns the string '1' to the
451 Z> f @attr 1=4 mozart
453 Number of hits: 43, setno 1
455 Z> f @and @set 1 @attr 1=4 amadeus
457 Number of hits: 14, setno 2
463 Named result sets are only supported by the Z39.50 protocol.
464 The SRU web service is stateless, and therefore the notion of
465 named result sets does not exist when accessing a Zebra server by
471 <section id="querymodel-use-string">
472 <title>Zebra's special access point of type 'string'</title>
474 The numeric <emphasis>use (type 1)</emphasis> attribute is usually
475 referred to from a given
476 attribute set. In addition, Zebra let you use
477 <emphasis>any internal index
478 name defined in your configuration</emphasis>
479 as use attribute value. This is a great feature for
480 debugging, and when you do
481 not need the complexity of defined use attribute values. It is
482 the preferred way of accessing Zebra indexes directly.
485 Finding all documents which have the term list "information
486 retrieval" in an Zebra index, using it's internal full string
487 name. Scanning the same index.
489 Z> find @attr 1=sometext "information retrieval"
490 Z> scan @attr 1=sometext aterm
494 Searching or scanning
495 the bib-1 use attribute 54 using it's string name:
497 Z> find @attr 1=Code-language eng
498 Z> scan @attr 1=Code-language ""
502 It is possible to search
503 in any silly string index - if it's defined in your
504 indexation rules and can be parsed by the PQF parser.
505 This is definitely not the recommended use of
506 this facility, as it might confuse your users with some very
509 Z> find @attr 1=silly/xpath/alike[@index]/name "information retrieval"
513 See also <xref linkend="querymodel-pqf-apt-mapping"/> for details, and
514 <xref linkend="zebrasrv-sru"/>
515 for the SRU PQF query extension using string names as a fast
520 <section id="querymodel-use-xpath">
521 <title>Zebra's special access point of type 'XPath'
522 for GRS filters</title>
524 As we have seen above, it is possible (albeit seldom a great
526 <ulink url="http://www.w3.org/TR/xpath">XPath 1.0</ulink> based
527 search by defining <emphasis>use (type 1)</emphasis>
528 <emphasis>string</emphasis> attributes which in appearance
529 <emphasis>resemble XPath queries</emphasis>. There are two
530 problems with this approach: first, the XPath-look-alike has to
531 be defined at indexation time, no new undefined
532 XPath queries can entered at search time, and second, it might
533 confuse users very much that an XPath-alike index name in fact
534 gets populated from a possible entirely different XML element
535 than it pretends to access.
538 When using the GRS Record Model
539 (see <xref linkend="grs"/>), we have the
540 possibility to embed <emphasis>life</emphasis>
542 in the PQF queries, which are here called
543 <emphasis>use (type 1)</emphasis> <emphasis>xpath</emphasis>
544 attributes. You must enable the
545 <literal>xpath enable</literal> directive in your
546 <literal>.abs</literal> configuration files.
550 Only a <emphasis>very</emphasis> restricted subset of the
551 <ulink url="http://www.w3.org/TR/xpath">XPath 1.0</ulink>
552 standard is supported as the GRS record model is simpler than
553 a full XML DOM structure. See the following examples for
558 Finding all documents which have the term "content"
559 inside a text node found in a specific XML DOM
560 <emphasis>subtree</emphasis>, whose starting element is
563 Z> find @attr 1=/root content
564 Z> find @attr 1=/root/first content
566 <emphasis>Notice that the
567 XPath must be absolute, i.e., must start with '/', and that the
568 XPath <literal>descendant-or-self</literal> axis followed by a
569 text node selection <literal>text()</literal> is implicitly
570 appended to the stated XPath.
572 It follows that the above searches are interpreted as:
574 Z> find @attr 1=/root//text() content
575 Z> find @attr 1=/root/first//text() content
580 Searching inside attribute strings is possible:
582 Z> find @attr 1=/link/@creator morten
587 Filter the addressing XPath by a predicate working on exact
589 attributes (in the XML sense) can be done: return all those docs which
590 have the term "english" contained in one of all text sub nodes of
591 the subtree defined by the XPath
592 <literal>/record/title[@lang='en']</literal>. And similar
595 Z> find @attr 1=/record/title[@lang='en'] english
596 Z> find @attr 1=/link[@creator='sisse'] sibelius
597 Z> find @attr 1=/link[@creator='sisse']/description[@xml:lang='da'] sibelius
602 Combining numeric indexes, boolean expressions,
603 and xpath based searches is possible:
605 Z> find @attr 1=/record/title @and foo bar
606 Z> find @and @attr 1=/record/title foo @attr 1=4 bar
610 Escaping PQF keywords and other non-parseable XPath constructs
611 with <literal>'{ }'</literal> to prevent client-side PQF parsing
614 Z> find @attr {1=/root/first[@attr='danish']} content
615 Z> find @attr {1=/record/@set} oai
620 It is worth mentioning that these dynamic performed XPath
621 queries are a performance bottleneck, as no optimized
622 specialized indexes can be used. Therefore, avoid the use of
623 this facility when speed is essential, and the database content
624 size is medium to large.
630 <section id="querymodel-exp1">
631 <title>Explain Attribute Set</title>
633 The Z39.50 standard defines the
634 <ulink url="&url.z39.50.explain;">Explain</ulink> attribute set
635 Exp-1, which is used to discover information
636 about a server's search semantics and functional capabilities
637 Zebra exposes a "classic"
638 Explain database by base name <literal>IR-Explain-1</literal>, which
639 is populated with system internal information.
642 The attribute-set <literal>exp-1</literal> consists of a single
643 use attribute (type 1).
646 In addition, the non-Use
647 Bib-1 attributes, that is, the types
648 <emphasis>Relation</emphasis>, <emphasis>Position</emphasis>,
649 <emphasis>Structure</emphasis>, <emphasis>Truncation</emphasis>,
650 and <emphasis>Completeness</emphasis> are imported from
651 the Bib-1 attribute set, and may be used
652 within any explain query.
655 <section id="querymodel-exp1-use">
656 <title>Use Attributes (type = 1)</title>
658 The following Explain search attributes are supported:
659 <literal>ExplainCategory</literal> (@attr 1=1),
660 <literal>DatabaseName</literal> (@attr 1=3),
661 <literal>DateAdded</literal> (@attr 1=9),
662 <literal>DateChanged</literal>(@attr 1=10).
665 A search in the use attribute <literal>ExplainCategory</literal>
666 supports only these predefined values:
667 <literal>CategoryList</literal>, <literal>TargetInfo</literal>,
668 <literal>DatabaseInfo</literal>, <literal>AttributeDetails</literal>.
671 See <filename>tab/explain.att</filename> and the
672 <ulink url="&url.z39.50;">Z39.50</ulink> standard
673 for more information.
677 <section id="querymodel-examples">
678 <title>Explain searches with yaz-client</title>
680 Classic Explain only defines retrieval of Explain information
681 via ASN.1. Practically no Z39.50 clients supports this. Fortunately
682 they don't have to - Zebra allows retrieval of this information
684 <literal>SUTRS</literal>, <literal>XML</literal>,
685 <literal>GRS-1</literal> and <literal>ASN.1</literal> Explain.
689 List supported categories to find out which explain commands are
693 Z> find @attr exp1 1=1 categorylist
700 Get target info, that is, investigate which databases exist at
701 this server endpoint:
704 Z> find @attr exp1 1=1 targetinfo
715 List all supported databases, the number of hits
716 is the number of databases found, which most commonly are the
718 the <literal>Default</literal> and the
719 <literal>IR-Explain-1</literal> databases.
722 Z> find @attr exp1 1=1 databaseinfo
729 Get database info record for database <literal>Default</literal>.
732 Z> find @and @attr exp1 1=1 databaseinfo @attr exp1 1=3 Default
734 Identical query with explicitly specified attribute set:
737 Z> find @attrset exp1 @and @attr 1=1 databaseinfo @attr 1=3 Default
742 Get attribute details record for database
743 <literal>Default</literal>.
744 This query is very useful to study the internal Zebra indexes.
745 If records have been indexed using the <literal>alvis</literal>
746 XSLT filter, the string representation names of the known indexes can be
750 Z> find @and @attr exp1 1=1 attributedetails @attr exp1 1=3 Default
752 Identical query with explicitly specified attribute set:
755 Z> find @attrset exp1 @and @attr 1=1 attributedetails @attr 1=3 Default
762 <section id="querymodel-bib1">
763 <title>Bib-1 Attribute Set</title>
765 Most of the information contained in this section is an excerpt of
766 the ATTRIBUTE SET BIB-1 (Z39.50-1995) SEMANTICS
767 found at <ulink url="&url.z39.50.attset.bib1.1995;">. The Bib-1
768 Attribute Set Semantics</ulink> from 1995, also in an updated
769 <ulink url="&url.z39.50.attset.bib1;">Bib-1
770 Attribute Set</ulink>
771 version from 2003. Index Data is not the copyright holder of this
772 information, except for the configuration details, the listing of
773 Zebra's capabilities, and the example queries.
777 <section id="querymodel-bib1-use">
778 <title>Use Attributes (type 1)</title>
781 A use attribute specifies an access point for any atomic query.
782 These access points are highly dependent on the attribute set used
783 in the query, and are user configurable using the following
784 default configuration files:
785 <filename>tab/bib1.att</filename>,
786 <filename>tab/dan1.att</filename>,
787 <filename>tab/explain.att</filename>, and
788 <filename>tab/gils.att</filename>.
791 For example, some few Bib-1 use
792 attributes from the <filename>tab/bib1.att</filename> are:
796 att 3 Conference-name
799 att 1009 Subject-name-personal
800 att 1010 Body-of-text
801 att 1011 Date/time-added-to-db
804 att 1017 Server-choice
808 att 1036 Author-Title-Subject
812 New attribute sets can be added by adding new
813 <filename>tab/*.att</filename> configuration files, which need to
814 be sourced in the main configuration <filename>zebra.cfg</filename>.
817 In addition, Zebra allows the access of
818 <emphasis>internal index names</emphasis> and <emphasis>dynamic
819 XPath</emphasis> as use attributes; see
820 <xref linkend="querymodel-use-string"/> and
821 <xref linkend="querymodel-use-xpath"/>.
825 Phrase search for <emphasis>information retrieval</emphasis> in
826 the title-register, scanning the same register afterwards:
828 Z> find @attr 1=4 "information retrieval"
829 Z> scan @attr 1=4 information
837 <section id="querymodel-bib1-nonuse">
838 <title>Zebra general Bib1 Non-Use Attributes (type 2-6)</title>
840 <section id="querymodel-bib1-relation">
841 <title>Relation Attributes (type 2)</title>
844 Relation attributes describe the relationship of the access
846 of the relation) to the search term as qualified by the attributes (right
847 side of the relation), e.g., Date-publication <= 1975.
850 <table id="querymodel-bib1-relation-table" frame="top">
851 <title>Relation Attributes (type 2)</title>
855 <entry>Relation</entry>
862 <entry>Less than</entry>
864 <entry>supported</entry>
867 <entry>Less than or equal</entry>
869 <entry>supported</entry>
874 <entry>default</entry>
877 <entry>Greater or equal</entry>
879 <entry>supported</entry>
882 <entry>Greater than</entry>
884 <entry>supported</entry>
887 <entry>Not equal</entry>
889 <entry>unsupported</entry>
892 <entry>Phonetic</entry>
894 <entry>unsupported</entry>
899 <entry>unsupported</entry>
902 <entry>Relevance</entry>
904 <entry>supported</entry>
907 <entry>AlwaysMatches</entry>
909 <entry>supported *</entry>
916 AlwaysMatches searches are only supported if alwaysmatches indexing
917 has been enabled. See <xref linkend="default-idx-file"/>
922 The relation attributes 1-5 are supported and work exactly as
924 All ordering operations are based on a lexicographical ordering,
925 <emphasis>expect</emphasis> when the
926 structure attribute numeric (109) is used. In
927 this case, ordering is numerical. See
928 <xref linkend="querymodel-bib1-structure"/>.
930 Z> find @attr 1=Title @attr 2=1 music
932 Number of hits: 11745, setno 1
934 Z> find @attr 1=Title @attr 2=2 music
936 Number of hits: 11771, setno 2
938 Z> find @attr 1=Title @attr 2=3 music
940 Number of hits: 532, setno 3
942 Z> find @attr 1=Title @attr 2=4 music
944 Number of hits: 11463, setno 4
946 Z> find @attr 1=Title @attr 2=5 music
948 Number of hits: 11419, setno 5
953 The relation attribute
954 <emphasis>Relevance (102)</emphasis> is supported, see
955 <xref linkend="administration-ranking"/> for full information.
959 Ranked search for <emphasis>information retrieval</emphasis> in
962 Z> find @attr 1=4 @attr 2=102 "information retrieval"
967 The relation attribute
968 <emphasis>AlwaysMatches (103)</emphasis> is in the default
970 supported in conjecture with structure attribute
971 <emphasis>Phrase (1)</emphasis> (which may be omitted by
973 It can be configured to work with other structure attributes,
974 see the configuration file
975 <filename>tab/default.idx</filename> and
976 <xref linkend="querymodel-pqf-apt-mapping"/>.
979 <emphasis>AlwaysMatches (103)</emphasis> is a
980 great way to discover how many documents have been indexed in a
981 given field. The search term is ignored, but needed for correct
982 PQF syntax. An empty search term may be supplied.
984 Z> find @attr 1=Title @attr 2=103 ""
985 Z> find @attr 1=Title @attr 2=103 @attr 4=1 ""
992 <section id="querymodel-bib1-position">
993 <title>Position Attributes (type 3)</title>
996 The position attribute specifies the location of the search term
997 within the field or subfield in which it appears.
1000 <table id="querymodel-bib1-position-table" frame="top">
1001 <title>Position Attributes (type 3)</title>
1005 <entry>Position</entry>
1006 <entry>Value</entry>
1007 <entry>Notes</entry>
1012 <entry>First in field </entry>
1014 <entry>supported *</entry>
1017 <entry>First in subfield</entry>
1019 <entry>supported *</entry>
1022 <entry>Any position in field</entry>
1024 <entry>default</entry>
1032 Zebra only supports first-in-field seaches if the
1033 <literal>firstinfield</literal> is enabled for the index
1034 Refer to <xref linkend="default-idx-file"/>.
1035 Zebra does not distinguish between first in field and
1036 first in subfield. They result in the same hit count.
1037 Searching for first position in (sub)field in only supported in Zebra
1043 <section id="querymodel-bib1-structure">
1044 <title>Structure Attributes (type 4)</title>
1047 The structure attribute specifies the type of search
1048 term. This causes the search to be mapped on
1049 different Zebra internal indexes, which must have been defined
1054 The possible values of the
1055 <literal>structure attribute (type 4)</literal> can be defined
1056 using the configuration file <filename>
1057 tab/default.idx</filename>.
1058 The default configuration is summarized in this table.
1061 <table id="querymodel-bib1-structure-table" frame="top">
1062 <title>Structure Attributes (type 4)</title>
1066 <entry>Structure</entry>
1067 <entry>Value</entry>
1068 <entry>Notes</entry>
1073 <entry>Phrase </entry>
1075 <entry>default</entry>
1080 <entry>supported</entry>
1085 <entry>supported</entry>
1090 <entry>supported</entry>
1093 <entry>Date (normalized)</entry>
1095 <entry>supported</entry>
1098 <entry>Word list</entry>
1100 <entry>supported</entry>
1103 <entry>Date (un-normalized)</entry>
1105 <entry>unsupported</entry>
1108 <entry>Name (normalized) </entry>
1110 <entry>unsupported</entry>
1113 <entry>Name (un-normalized) </entry>
1115 <entry>unsupported</entry>
1118 <entry>Structure</entry>
1120 <entry>unsupported</entry>
1125 <entry>supported</entry>
1128 <entry>Free-form-text</entry>
1130 <entry>supported</entry>
1133 <entry>Document-text</entry>
1135 <entry>supported</entry>
1138 <entry>Local-number</entry>
1140 <entry>supported</entry>
1143 <entry>String</entry>
1145 <entry>unsupported</entry>
1148 <entry>Numeric string</entry>
1150 <entry>supported</entry>
1157 The structure attribute values
1158 <literal>Word list (6)</literal>
1159 is supported, and maps to the boolean <literal>AND</literal>
1160 combination of words supplied. The word list is useful when
1161 google-like bag-of-word queries need to be translated from a GUI
1162 query language to PQF. For example, the following queries
1165 Z> find @attr 1=Title @attr 4=6 "mozart amadeus"
1166 Z> find @attr 1=Title @and mozart amadeus
1171 The structure attribute value
1172 <literal>Free-form-text (105)</literal> and
1173 <literal>Document-text (106)</literal>
1174 are supported, and map both to the boolean <literal>OR</literal>
1175 combination of words supplied. The following queries
1178 Z> find @attr 1=Body-of-text @attr 4=105 "bach salieri teleman"
1179 Z> find @attr 1=Body-of-text @attr 4=106 "bach salieri teleman"
1180 Z> find @attr 1=Body-of-text @or bach @or salieri teleman
1182 This <literal>OR</literal> list of terms is very useful in
1183 combination with relevance ranking:
1185 Z> find @attr 1=Body-of-text @attr 2=102 @attr 4=105 "bach salieri teleman"
1190 The structure attribute value
1191 <literal>Local number (107)</literal>
1192 is supported, and maps always to the Zebra internal document ID,
1193 irrespectively which use attribute is specified. The following queries
1194 have exactly the same unique record in the hit set:
1196 Z> find @attr 4=107 10
1197 Z> find @attr 1=4 @attr 4=107 10
1198 Z> find @attr 1=1010 @attr 4=107 10
1204 the GILS schema (<literal>gils.abs</literal>), the
1205 west-bounding-coordinate is indexed as type <literal>n</literal>,
1206 and is therefore searched by specifying
1207 <emphasis>structure</emphasis>=<emphasis>Numeric String</emphasis>.
1208 To match all those records with west-bounding-coordinate greater
1209 than -114 we use the following query:
1211 Z> find @attr 4=109 @attr 2=5 @attr gils 1=2038 -114
1216 The exact mapping between PQF queries and Zebra internal indexes
1217 and index types is explained in
1218 <xref linkend="querymodel-pqf-apt-mapping"/>.
1223 <section id="querymodel-bib1-truncation">
1224 <title>Truncation Attributes (type = 5)</title>
1227 The truncation attribute specifies whether variations of one or
1228 more characters are allowed between search term and hit terms, or
1229 not. Using non-default truncation attributes will broaden the
1230 document hit set of a search query.
1233 <table id="querymodel-bib1-truncation-table" frame="top">
1234 <title>Truncation Attributes (type 5)</title>
1238 <entry>Truncation</entry>
1239 <entry>Value</entry>
1240 <entry>Notes</entry>
1245 <entry>Right truncation </entry>
1247 <entry>supported</entry>
1250 <entry>Left truncation</entry>
1252 <entry>supported</entry>
1255 <entry>Left and right truncation</entry>
1257 <entry>supported</entry>
1260 <entry>Do not truncate</entry>
1262 <entry>default</entry>
1265 <entry>Process # in search term</entry>
1267 <entry>supported</entry>
1270 <entry>RegExpr-1 </entry>
1272 <entry>supported</entry>
1275 <entry>RegExpr-2</entry>
1277 <entry>supported</entry>
1284 The truncation attribute values 1-3 perform the obvious way:
1286 Z> scan @attr 1=Body-of-text schnittke
1292 Z> find @attr 1=Body-of-text @attr 5=1 schnittke
1294 Number of hits: 95, setno 7
1296 Z> find @attr 1=Body-of-text @attr 5=2 schnittke
1298 Number of hits: 81, setno 6
1300 Z> find @attr 1=Body-of-text @attr 5=3 schnittke
1302 Number of hits: 95, setno 8
1307 The truncation attribute value
1308 <literal>Process # in search term (101)</literal> is a
1309 poor-man's regular expression search. It maps
1310 each <literal>#</literal> to <literal>.*</literal>, and
1311 performs then a <literal>Regexp-1 (102)</literal> regular
1312 expression search. The following two queries are equivalent:
1314 Z> find @attr 1=Body-of-text @attr 5=101 schnit#ke
1315 Z> find @attr 1=Body-of-text @attr 5=102 schnit.*ke
1317 Number of hits: 89, setno 10
1322 The truncation attribute value
1323 <literal>Regexp-1 (102)</literal> is a normal regular search,
1324 see <xref linkend="querymodel-regular"/> for details.
1326 Z> find @attr 1=Body-of-text @attr 5=102 schnit+ke
1327 Z> find @attr 1=Body-of-text @attr 5=102 schni[a-t]+ke
1332 The truncation attribute value
1333 <literal>Regexp-2 (103) </literal> is a Zebra specific extension
1334 which allows <emphasis>fuzzy</emphasis> matches. One single
1335 error in spelling of search terms is allowed, i.e., a document
1336 is hit if it includes a term which can be mapped to the used
1337 search term by one character substitution, addition, deletion or
1340 Z> find @attr 1=Body-of-text @attr 5=100 schnittke
1342 Number of hits: 81, setno 14
1344 Z> find @attr 1=Body-of-text @attr 5=103 schnittke
1346 Number of hits: 103, setno 15
1352 <section id="querymodel-bib1-completeness">
1353 <title>Completeness Attributes (type = 6)</title>
1357 The <literal>Completeness Attributes (type = 6)</literal>
1358 is used to specify that a given search term or term list is either
1359 part of the terms of a given index/field
1360 (<literal>Incomplete subfield (1)</literal>), or is
1361 what literally is found in the entire field's index
1362 (<literal>Complete field (3)</literal>).
1365 <table id="querymodel-bib1-completeness-table" frame="top">
1366 <title>Completeness Attributes (type = 6)</title>
1370 <entry>Completeness</entry>
1371 <entry>Value</entry>
1372 <entry>Notes</entry>
1377 <entry>Incomplete subfield</entry>
1379 <entry>default</entry>
1382 <entry>Complete subfield</entry>
1384 <entry>deprecated</entry>
1387 <entry>Complete field</entry>
1389 <entry>supported</entry>
1396 The <literal>Completeness Attributes (type = 6)</literal>
1397 is only partially and conditionally
1398 supported in the sense that it is ignored if the hit index is
1399 not of structure <literal>type="w"</literal> or
1400 <literal>type="p"</literal>.
1403 <literal>Incomplete subfield (1)</literal> is the default, and
1405 register <literal>type="w"</literal>, whereas
1406 <literal>Complete field (3)</literal> triggers
1407 search and scan in index <literal>type="p"</literal>.
1410 The <literal>Complete subfield (2)</literal> is a reminiscens
1411 from the happy <literal>MARC</literal>
1412 binary format days. Zebra does not support it, but maps silently
1413 to <literal>Complete field (3)</literal>.
1418 The exact mapping between PQF queries and Zebra internal indexes
1419 and index types is explained in
1420 <xref linkend="querymodel-pqf-apt-mapping"/>.
1429 <section id="querymodel-zebra">
1430 <title>Extended Zebra RPN Features</title>
1432 The Zebra internal query engine has been extended to specific needs
1433 not covered by the <literal>bib-1</literal> attribute set query
1434 model. These extensions are <emphasis>non-standard</emphasis>
1435 and <emphasis>non-portable</emphasis>: most functional extensions
1436 are modeled over the <literal>bib-1</literal> attribute set,
1437 defining type 7 and higher values.
1438 There are also the special
1439 <literal>string</literal> type index names for the
1440 <literal>idxpath</literal> attribute set.
1443 <section id="querymodel-zebra-attr-allrecords">
1444 <title>Zebra specific retrieval of all records</title>
1446 Zebra defines a hardwired <literal>string</literal> index name
1447 called <literal>_ALLRECORDS</literal>. It matches any record
1448 contained in the database, if used in conjunction with
1449 the relation attribute
1450 <literal>AlwaysMatches (103)</literal>.
1453 The <literal>_ALLRECORDS</literal> index name is used for total database
1454 export. The search term is ignored, it may be empty.
1456 Z> find @attr 1=_ALLRECORDS @attr 2=103 ""
1460 Combination with other index types can be made. For example, to
1461 find all records which are <emphasis>not</emphasis> indexed in
1462 the <literal>Title</literal> register, issue one of the two
1465 Z> find @not @attr 1=_ALLRECORDS @attr 2=103 "" @attr 1=Title @attr 2=103 ""
1466 Z> find @not @attr 1=_ALLRECORDS @attr 2=103 "" @attr 1=4 @attr 2=103 ""
1471 The special string index <literal>_ALLRECORDS</literal> is
1472 experimental, and the provided functionality and syntax may very
1473 well change in future releases of Zebra.
1478 <section id="querymodel-zebra-attr-search">
1479 <title>Zebra specific Search Extensions to all Attribute Sets</title>
1481 Zebra extends the Bib-1 attribute types, and these extensions are
1482 recognized regardless of attribute
1483 set used in a <literal>search</literal> operation query.
1486 <table id="querymodel-zebra-attr-search-table" frame="top">
1487 <title>Zebra Search Attribute Extensions</title>
1492 <entry>Value</entry>
1493 <entry>Operation</entry>
1494 <entry>Zebra version</entry>
1499 <entry>Embedded Sort</entry>
1501 <entry>search</entry>
1505 <entry>Term Set</entry>
1507 <entry>search</entry>
1511 <entry>Rank Weight</entry>
1513 <entry>search</entry>
1517 <entry>Term Reference</entry>
1519 <entry>search</entry>
1523 <entry>Local Approx Limit</entry>
1525 <entry>search</entry>
1529 <entry>Global Approx Limit</entry>
1531 <entry>search</entry>
1532 <entry>2.0.8</entry>
1538 <section id="querymodel-zebra-attr-sorting">
1539 <title>Zebra Extension Embedded Sort Attribute (type 7)</title>
1541 The embedded sort is a way to specify sort within a query - thus
1542 removing the need to send a Sort Request separately. It is both
1543 faster and does not require clients to deal with the Sort
1548 All ordering operations are based on a lexicographical ordering,
1549 <emphasis>expect</emphasis> when the
1550 <literal>structure attribute numeric (109)</literal> is used. In
1551 this case, ordering is numerical. See
1552 <xref linkend="querymodel-bib1-structure"/>.
1556 The possible values after attribute <literal>type 7</literal> are
1557 <literal>1</literal> ascending and
1558 <literal>2</literal> descending.
1559 The attributes+term (APT) node is separate from the
1560 rest and must be <literal>@or</literal>'ed.
1561 The term associated with APT is the sorting level in integers,
1562 where <literal>0</literal> means primary sort,
1563 <literal>1</literal> means secondary sort, and so forth.
1564 See also <xref linkend="administration-ranking"/>.
1567 For example, searching for water, sort by title (ascending)
1569 Z> find @or @attr 1=1016 water @attr 7=1 @attr 1=4 0
1573 Or, searching for water, sort by title ascending, then date descending
1575 Z> find @or @or @attr 1=1016 water @attr 7=1 @attr 1=4 0 @attr 7=2 @attr 1=30 1
1581 Zebra Extension Term Set Attribute
1582 From the manual text, I can not see what is the point with this feature.
1583 I think it makes more sense when there are multiple terms in a query, or
1586 We decided 2006-06-03 to disable this feature, as it is covered by
1587 scan within a resultset. Better use ressources to upgrade this
1588 feature for good performance.
1592 <section id="querymodel-zebra-attr-estimation">
1593 <title>Zebra Extension Term Set Attribute (type 8)</title>
1595 The Term Set feature is a facility that allows a search to store
1596 hitting terms in a "pseudo" resultset; thus a search (as usual) +
1597 a scan-like facility. Requires a client that can do named result
1598 sets since the search generates two result sets. The value for
1599 attribute 8 is the name of a result set (string). The terms in
1600 the named term set are returned as SUTRS records.
1603 For example, searching for u in title, right truncated, and
1604 storing the result in term set named 'aset'
1606 Z> find @attr 5=1 @attr 1=4 @attr 8=aset u
1610 The model has one serious flaw: we don't know the size of term
1611 set. Experimental. Do not use in production code.
1617 <section id="querymodel-zebra-attr-weight">
1618 <title>Zebra Extension Rank Weight Attribute (type 9)</title>
1620 Rank weight is a way to pass a value to a ranking algorithm - so
1621 that one APT has one value - while another as a different one.
1622 See also <xref linkend="administration-ranking"/>.
1625 For example, searching for utah in title with weight 30 as well
1626 as any with weight 20:
1628 Z> find @attr 2=102 @or @attr 9=30 @attr 1=4 utah @attr 9=20 utah
1633 <section id="querymodel-zebra-attr-termref">
1634 <title>Zebra Extension Term Reference Attribute (type 10)</title>
1636 Zebra supports the searchResult-1 facility.
1637 If the Term Reference Attribute (type 10) is
1638 given, that specifies a subqueryId value returned as part of the
1639 search result. It is a way for a client to name an APT part of a
1650 Experimental. Do not use in production code.
1658 <section id="querymodel-zebra-local-attr-limit">
1659 <title>Local Approximative Limit Attribute (type 11)</title>
1661 Zebra computes - unless otherwise configured -
1662 the exact hit count for every APT
1663 (leaf) in the query tree. These hit counts are returned as part of
1664 the searchResult-1 facility in the binary encoded Z39.50 search
1668 By setting an estimation limit size of the resultset of the APT
1669 leaves, Zebra stoppes processing the result set when the limit
1671 Hit counts under this limit are still precise, but hit counts over it
1672 are estimated using the statistics gathered from the chopped
1676 Specifying a limit of <literal>0</literal> resuts in exact hit counts.
1679 For example, we might be interested in exact hit count for a, but
1680 for b we allow hit count estimates for 1000 and higher.
1682 Z> find @and a @attr 11=1000 b
1687 The estimated hit count facility makes searches faster, as one
1688 only needs to process large hit lists partially.
1689 It is mostly used in huge databases, where you you want trade
1690 exactness of hit counts against speed of execution.
1695 Do not use approximative hit count limits
1696 in conjunction with relevance ranking, as re-sorting of the
1697 result set only works when the entire result set has
1703 <section id="querymodel-zebra-global-attr-limit">
1704 <title>Global Approximative Limit Attribute (type 12)</title>
1706 By default Zebra computes precise hit counts for a query as
1707 a whole. Setting attribute 12 makes it perform approximative
1708 hit counts instead. It has the same semantics as
1709 <literal>estimatehits</literal> for the <xref linkend="zebra-cfg"/>.
1712 The attribute (12) can occur anywhere in the query tree.
1713 Unlike regular attributes it does not relate to the leaf (APT)
1714 - but to the whole query.
1718 Do not use approximative hit count limits
1719 in conjunction with relevance ranking, as re-sorting of the
1720 result set only works when the entire result set has
1728 <section id="querymodel-zebra-attr-scan">
1729 <title>Zebra specific Scan Extensions to all Attribute Sets</title>
1731 Zebra extends the Bib1 attribute types, and these extensions are
1732 recognized regardless of attribute
1733 set used in a scan operation query.
1735 <table id="querymodel-zebra-attr-scan-table" frame="top">
1736 <title>Zebra Scan Attribute Extensions</title>
1742 <entry>Operation</entry>
1743 <entry>Zebra version</entry>
1748 <entry>Result Set Narrow</entry>
1754 <entry>Approximative Limit</entry>
1763 <section id="querymodel-zebra-attr-narrow">
1764 <title>Zebra Extension Result Set Narrow (type 8)</title>
1766 If attribute Result Set Narrow (type 8)
1767 is given for scan, the value is the name of a
1768 result set. Each hit count in scan is
1769 <literal>@and</literal>'ed with the result set given.
1772 Consider for example
1773 the case of scanning all title fields around the
1774 scanterm <emphasis>mozart</emphasis>, then refining the scan by
1775 issuing a filtering query for <emphasis>amadeus</emphasis> to
1776 restrict the scan to the result set of the query:
1778 Z> scan @attr 1=4 mozart
1781 mozartforskningen (1)
1785 Z> f @attr 1=4 amadeus
1787 Number of hits: 15, setno 2
1789 Z> scan @attr 1=4 @attr 8=2 mozart
1792 mozartforskningen (0)
1800 Zebra 2.0.2 and later is able to skip 0 hit counts. This, however,
1801 is known not to scale if the number of terms to skip is high.
1802 This most likely will happen if the result set is small (and
1803 result in many 0 hits).
1807 <section id="querymodel-zebra-attr-approx">
1808 <title>Zebra Extension Approximative Limit (type 11)</title>
1810 The Zebra Extension Approximative Limit (type 11) is a way to
1811 enable approximate hit counts for scan hit counts, in the same
1812 way as for search hit counts.
1817 <section id="querymodel-idxpath">
1818 <title>Zebra special IDXPATH Attribute Set for GRS indexing</title>
1820 The attribute-set <literal>idxpath</literal> consists of a single
1821 Use (type 1) attribute. All non-use attributes behave as normal.
1824 This feature is enabled when defining the
1825 <literal>xpath enable</literal> option in the GRS filter
1826 <filename>*.abs</filename> configuration files. If one wants to use
1827 the special <literal>idxpath</literal> numeric attribute set, the
1828 main Zebra configuration file <filename>zebra.cfg</filename>
1829 directive <literal>attset: idxpath.att</literal> must be enabled.
1833 The <literal>idxpath</literal> is deprecated, may not be
1834 supported in future Zebra versions, and should definitely
1835 not be used in production code.
1839 <section id="querymodel-idxpath-use">
1840 <title>IDXPATH Use Attributes (type = 1)</title>
1842 This attribute set allows one to search GRS filter indexed
1843 records by XPATH like structured index names.
1848 The <literal>idxpath</literal> option defines hard-coded
1849 index names, which might clash with your own index names.
1853 <table id="querymodel-idxpath-use-table" frame="top">
1854 <title>Zebra specific IDXPATH Use Attributes (type 1)</title>
1858 <entry>IDXPATH</entry>
1859 <entry>Value</entry>
1860 <entry>String Index</entry>
1861 <entry>Notes</entry>
1866 <entry>XPATH Begin</entry>
1868 <entry>_XPATH_BEGIN</entry>
1869 <entry>deprecated</entry>
1872 <entry>XPATH End</entry>
1874 <entry>_XPATH_END</entry>
1875 <entry>deprecated</entry>
1878 <entry>XPATH CData</entry>
1880 <entry>_XPATH_CDATA</entry>
1881 <entry>deprecated</entry>
1884 <entry>XPATH Attribute Name</entry>
1886 <entry>_XPATH_ATTR_NAME</entry>
1887 <entry>deprecated</entry>
1890 <entry>XPATH Attribute CData</entry>
1892 <entry>_XPATH_ATTR_CDATA</entry>
1893 <entry>deprecated</entry>
1900 See <filename>tab/idxpath.att</filename> for more information.
1903 Search for all documents starting with root element
1904 <literal>/root</literal> (either using the numeric or the string
1907 Z> find @attrset idxpath @attr 1=1 @attr 4=3 root/
1908 Z> find @attr idxpath 1=1 @attr 4=3 root/
1909 Z> find @attr 1=_XPATH_BEGIN @attr 4=3 root/
1913 Search for all documents where specific nested XPATH
1914 <literal>/c1/c2/../cn</literal> exists. Notice the very
1915 counter-intuitive <emphasis>reverse</emphasis> notation!
1917 Z> find @attrset idxpath @attr 1=1 @attr 4=3 cn/cn-1/../c1/
1918 Z> find @attr 1=_XPATH_BEGIN @attr 4=3 cn/cn-1/../c1/
1922 Search for CDATA string <emphasis>text</emphasis> in any element
1924 Z> find @attrset idxpath @attr 1=1016 text
1925 Z> find @attr 1=_XPATH_CDATA text
1929 Search for CDATA string <emphasis>anothertext</emphasis> in any
1932 Z> find @attrset idxpath @attr 1=1015 anothertext
1933 Z> find @attr 1=_XPATH_ATTR_CDATA anothertext
1937 Search for all documents with have an XML element node
1938 including an XML attribute named <emphasis>creator</emphasis>
1940 Z> find @attrset idxpath @attr 1=3 @attr 4=3 creator
1941 Z> find @attr 1=_XPATH_ATTR_NAME @attr 4=3 creator
1945 Combining usual <literal>bib-1</literal> attribute set searches
1946 with <literal>idxpath</literal> attribute set searches:
1948 Z> find @and @attr idxpath 1=1 @attr 4=3 link/ @attr 1=4 mozart
1949 Z> find @and @attr 1=_XPATH_BEGIN @attr 4=3 link/ @attr 1=_XPATH_CDATA mozart
1953 Scanning is supported on all <literal>idxpath</literal>
1954 indexes, both specified as numeric use attributes, or as string
1957 Z> scan @attrset idxpath @attr 1=1016 text
1958 Z> scan @attr 1=_XPATH_ATTR_CDATA anothertext
1959 Z> scan @attrset idxpath @attr 1=3 @attr 4=3 ''
1967 <section id="querymodel-pqf-apt-mapping">
1968 <title>Mapping from PQF atomic APT queries to Zebra internal
1969 register indexes</title>
1971 The rules for PQF APT mapping are rather tricky to grasp in the
1972 first place. We deal first with the rules for deciding which
1973 internal register or string index to use, according to the use
1974 attribute or access point specified in the query. Thereafter we
1975 deal with the rules for determining the correct structure type of
1979 <section id="querymodel-pqf-apt-mapping-accesspoint">
1980 <title>Mapping of PQF APT access points</title>
1982 Zebra understands four fundamental different types of access
1983 points, of which only the
1984 <emphasis>numeric use attribute</emphasis> type access points
1985 are defined by the <ulink url="&url.z39.50;">Z39.50</ulink>
1987 All other access point types are Zebra specific, and non-portable.
1990 <table id="querymodel-zebra-mapping-accesspoint-types" frame="top">
1991 <title>Access point name mapping</title>
1995 <entry>Access Point</entry>
1997 <entry>Grammar</entry>
1998 <entry>Notes</entry>
2003 <entry>Use attribute</entry>
2004 <entry>numeric</entry>
2005 <entry>[1-9][1-9]*</entry>
2006 <entry>directly mapped to string index name</entry>
2009 <entry>String index name</entry>
2010 <entry>string</entry>
2011 <entry>[a-zA-Z](\-?[a-zA-Z0-9])*</entry>
2012 <entry>normalized name is used as internal string index name</entry>
2015 <entry>Zebra internal index name</entry>
2016 <entry>zebra</entry>
2017 <entry>_[a-zA-Z](_?[a-zA-Z0-9])*</entry>
2018 <entry>hardwired internal string index name</entry>
2021 <entry>XPATH special index</entry>
2022 <entry>XPath</entry>
2024 <entry>special xpath search for GRS indexed records</entry>
2031 <literal>Attribute set names</literal> and
2032 <literal>string index names</literal> are normalizes
2033 according to the following rules: all <emphasis>single</emphasis>
2034 hyphens <literal>'-'</literal> are stripped, and all upper case
2035 letters are folded to lower case.
2039 <emphasis>Numeric use attributes</emphasis> are mapped
2040 to the Zebra internal
2041 string index according to the attribute set definition in use.
2042 The default attribute set is <literal>Bib-1</literal>, and may be
2043 omitted in the PQF query.
2047 According to normalization and numeric
2048 use attribute mapping, it follows that the following
2049 PQF queries are considered equivalent (assuming the default
2050 configuration has not been altered):
2052 Z> find @attr 1=Body-of-text serenade
2053 Z> find @attr 1=bodyoftext serenade
2054 Z> find @attr 1=BodyOfText serenade
2055 Z> find @attr 1=bO-d-Y-of-tE-x-t serenade
2056 Z> find @attr 1=1010 serenade
2057 Z> find @attrset Bib-1 @attr 1=1010 serenade
2058 Z> find @attrset bib1 @attr 1=1010 serenade
2059 Z> find @attrset Bib1 @attr 1=1010 serenade
2060 Z> find @attrset b-I-b-1 @attr 1=1010 serenade
2065 The <emphasis>numerical</emphasis>
2066 <literal>use attributes (type 1)</literal>
2067 are interpreted according to the
2068 attribute sets which have been loaded in the
2069 <literal>zebra.cfg</literal> file, and are matched against specific
2070 fields as specified in the <literal>.abs</literal> file which
2071 describes the profile of the records which have been loaded.
2072 If no use attribute is provided, a default of
2073 Bib-1 Use Any (1016) is assumed.
2074 The predefined use attribute sets
2075 can be reconfigured by tweaking the configuration files
2076 <filename>tab/*.att</filename>, and
2077 new attribute sets can be defined by adding similar files in the
2078 configuration path <literal>profilePath</literal> of the server.
2082 String indexes can be accessed directly,
2083 independently which attribute set is in use. These are just
2084 ignored. The above mentioned name normalization applies.
2085 String index names are defined in the
2086 used indexing filter configuration files, for example in the
2087 <literal>GRS</literal>
2088 <filename>*.abs</filename> configuration files, or in the
2089 <literal>alvis</literal> filter XSLT indexing stylesheets.
2093 Zebra internal indexes can be accessed directly,
2094 according to the same rules as the user defined
2095 string indexes. The only difference is that
2096 Zebra internal index names are hardwired,
2098 must start with the character <literal>'_'</literal>.
2102 Finally, <literal>XPATH</literal> access points are only
2103 available using the <literal>GRS</literal> filter for indexing.
2104 These access point names must start with the character
2105 <literal>'/'</literal>, they are <emphasis>not
2106 normalized</emphasis>, but passed unaltered to the Zebra internal
2107 XPATH engine. See <xref linkend="querymodel-use-xpath"/>.
2115 <section id="querymodel-pqf-apt-mapping-structuretype">
2116 <title>Mapping of PQF APT structure and completeness to
2117 register type</title>
2119 Internally Zebra has in it's default configuration several
2120 different types of registers or indexes, whose tokenization and
2121 character normalization rules differ. This reflects the fact that
2122 searching fundamental different tokens like dates, numbers,
2123 bitfields and string based text needs different rule sets.
2126 <table id="querymodel-zebra-mapping-structure-types" frame="top">
2127 <title>Structure and completeness mapping to register types</title>
2131 <entry>Structure</entry>
2132 <entry>Completeness</entry>
2133 <entry>Register type</entry>
2134 <entry>Notes</entry>
2140 phrase (@attr 4=1), word (@attr 4=2),
2141 word-list (@attr 4=6),
2142 free-form-text (@attr 4=105), or document-text (@attr 4=106)
2144 <entry>Incomplete field (@attr 6=1)</entry>
2145 <entry>Word ('w')</entry>
2146 <entry>Traditional tokenized and character normalized word index</entry>
2150 phrase (@attr 4=1), word (@attr 4=2),
2151 word-list (@attr 4=6),
2152 free-form-text (@attr 4=105), or document-text (@attr 4=106)
2154 <entry>complete field' (@attr 6=3)</entry>
2155 <entry>Phrase ('p')</entry>
2156 <entry>Character normalized, but not tokenized index for phrase
2161 <entry>urx (@attr 4=104)</entry>
2162 <entry>ignored</entry>
2163 <entry>URX/URL ('u')</entry>
2164 <entry>Special index for URL web addresses</entry>
2167 <entry>numeric (@attr 4=109)</entry>
2168 <entry>ignored</entry>
2169 <entry>Numeric ('u')</entry>
2170 <entry>Special index for digital numbers</entry>
2173 <entry>key (@attr 4=3)</entry>
2174 <entry>ignored</entry>
2175 <entry>Null bitmap ('0')</entry>
2176 <entry>Used for non-tokenizated and non-normalized bit sequences</entry>
2179 <entry>year (@attr 4=4)</entry>
2180 <entry>ignored</entry>
2181 <entry>Year ('y')</entry>
2182 <entry>Non-tokenizated and non-normalized 4 digit numbers</entry>
2185 <entry>date (@attr 4=5)</entry>
2186 <entry>ignored</entry>
2187 <entry>Date ('d')</entry>
2188 <entry>Non-tokenizated and non-normalized ISO date strings</entry>
2191 <entry>ignored</entry>
2192 <entry>ignored</entry>
2193 <entry>Sort ('s')</entry>
2194 <entry>Used with special sort attribute set (@attr 7=1, @attr 7=2)</entry>
2197 <entry>overruled</entry>
2198 <entry>overruled</entry>
2199 <entry>special</entry>
2200 <entry>Internal record ID register, used whenever
2201 Relation Always Matches (@attr 2=103) is specified</entry>
2207 <!-- see in util/zebramap.c -->
2210 If a <emphasis>Structure</emphasis> attribute of
2211 <emphasis>Phrase</emphasis> is used in conjunction with a
2212 <emphasis>Completeness</emphasis> attribute of
2213 <emphasis>Complete (Sub)field</emphasis>, the term is matched
2214 against the contents of the phrase (long word) register, if one
2215 exists for the given <emphasis>Use</emphasis> attribute.
2216 A phrase register is created for those fields in the
2217 GRS <filename>*.abs</filename> file that contains a
2218 <literal>p</literal>-specifier.
2220 Z> scan @attr 1=Title @attr 4=1 @attr 6=3 beethoven
2222 bayreuther festspiele (1)
2223 * beethoven bibliography database (1)
2226 Z> find @attr 1=Title @attr 4=1 @attr 6=3 "beethoven bibliography"
2228 Number of hits: 0, setno 5
2230 Z> find @attr 1=Title @attr 4=1 @attr 6=3 "beethoven bibliography database"
2232 Number of hits: 1, setno 6
2237 If <emphasis>Structure</emphasis>=<emphasis>Phrase</emphasis> is
2238 used in conjunction with <emphasis>Incomplete Field</emphasis> - the
2239 default value for <emphasis>Completeness</emphasis>, the
2240 search is directed against the normal word registers, but if the term
2241 contains multiple words, the term will only match if all of the words
2242 are found immediately adjacent, and in the given order.
2243 The word search is performed on those fields that are indexed as
2244 type <literal>w</literal> in the GRS <filename>*.abs</filename> file.
2246 Z> scan @attr 1=Title @attr 4=1 @attr 6=1 beethoven
2252 Z> find @attr 1=Title @attr 4=1 @attr 6=1 beethoven
2254 Number of hits: 18, setno 1
2256 Z> find @attr 1=Title @attr 4=1 @attr 6=1 "beethoven bibliography"
2258 Number of hits: 2, setno 2
2264 If the <emphasis>Structure</emphasis> attribute is
2265 <emphasis>Word List</emphasis>,
2266 <emphasis>Free-form Text</emphasis>, or
2267 <emphasis>Document Text</emphasis>, the term is treated as a
2268 natural-language, relevance-ranked query.
2269 This search type uses the word register, i.e. those fields
2270 that are indexed as type <literal>w</literal> in the
2271 GRS <filename>*.abs</filename> file.
2275 If the <emphasis>Structure</emphasis> attribute is
2276 <emphasis>Numeric String</emphasis> the term is treated as an integer.
2277 The search is performed on those fields that are indexed
2278 as type <literal>n</literal> in the GRS
2279 <filename>*.abs</filename> file.
2283 If the <emphasis>Structure</emphasis> attribute is
2284 <emphasis>URX</emphasis> the term is treated as a URX (URL) entity.
2285 The search is performed on those fields that are indexed as type
2286 <literal>u</literal> in the <filename>*.abs</filename> file.
2290 If the <emphasis>Structure</emphasis> attribute is
2291 <emphasis>Local Number</emphasis> the term is treated as
2292 native Zebra Record Identifier.
2296 If the <emphasis>Relation</emphasis> attribute is
2297 <emphasis>Equals</emphasis> (default), the term is matched
2298 in a normal fashion (modulo truncation and processing of
2299 individual words, if required).
2300 If <emphasis>Relation</emphasis> is <emphasis>Less Than</emphasis>,
2301 <emphasis>Less Than or Equal</emphasis>,
2302 <emphasis>Greater than</emphasis>, or <emphasis>Greater than or
2303 Equal</emphasis>, the term is assumed to be numerical, and a
2304 standard regular expression is constructed to match the given
2306 If <emphasis>Relation</emphasis> is <emphasis>Relevance</emphasis>,
2307 the standard natural-language query processor is invoked.
2311 For the <emphasis>Truncation</emphasis> attribute,
2312 <emphasis>No Truncation</emphasis> is the default.
2313 <emphasis>Left Truncation</emphasis> is not supported.
2314 <emphasis>Process # in search term</emphasis> is supported, as is
2315 <emphasis>Regxp-1</emphasis>.
2316 <emphasis>Regxp-2</emphasis> enables the fault-tolerant (fuzzy)
2317 search. As a default, a single error (deletion, insertion,
2318 replacement) is accepted when terms are matched against the register
2325 <section id="querymodel-regular">
2326 <title>Zebra Regular Expressions in Truncation Attribute (type = 5)</title>
2329 Each term in a query is interpreted as a regular expression if
2330 the truncation value is either <emphasis>Regxp-1 (@attr 5=102)</emphasis>
2331 or <emphasis>Regxp-2 (@attr 5=103)</emphasis>.
2332 Both query types follow the same syntax with the operands:
2335 <table id="querymodel-regular-operands-table" frame="top">
2336 <title>Regular Expression Operands</title>
2340 <entry><literal>x</literal></entry>
2341 <entry>Matches the character <literal>x</literal>.</entry>
2344 <entry><literal>.</literal></entry>
2345 <entry>Matches any character.</entry>
2348 <entry><literal>[ .. ]</literal></entry>
2349 <entry>Matches the set of characters specified;
2350 such as <literal>[abc]</literal> or <literal>[a-c]</literal>.</entry>
2357 The above operands can be combined with the following operators:
2360 <table id="querymodel-regular-operators-table" frame="top">
2361 <title>Regular Expression Operators</title>
2365 <entry><literal>x*</literal></entry>
2366 <entry>Matches <literal>x</literal> zero or more times.
2367 Priority: high.</entry>
2370 <entry><literal>x+</literal></entry>
2371 <entry>Matches <literal>x</literal> one or more times.
2372 Priority: high.</entry>
2375 <entry><literal>x?</literal></entry>
2376 <entry> Matches <literal>x</literal> zero or once.
2377 Priority: high.</entry>
2380 <entry><literal>xy</literal></entry>
2381 <entry> Matches <literal>x</literal>, then <literal>y</literal>.
2382 Priority: medium.</entry>
2385 <entry><literal>x|y</literal></entry>
2386 <entry> Matches either <literal>x</literal> or <literal>y</literal>.
2387 Priority: low.</entry>
2390 <entry><literal>( )</literal></entry>
2391 <entry>The order of evaluation may be changed by using parentheses.</entry>
2398 If the first character of the <literal>Regxp-2</literal> query
2399 is a plus character (<literal>+</literal>) it marks the
2400 beginning of a section with non-standard specifiers.
2401 The next plus character marks the end of the section.
2402 Currently Zebra only supports one specifier, the error tolerance,
2403 which consists one digit.
2404 <!-- TODO Nice thing, but what does
2405 that error tolerance digit *mean*? Maybe an example would be nice? -->
2409 Since the plus operator is normally a suffix operator the addition to
2410 the query syntax doesn't violate the syntax for standard regular
2415 For example, a phrase search with regular expressions in
2416 the title-register is performed like this:
2418 Z> find @attr 1=4 @attr 5=102 "informat.* retrieval"
2423 Combinations with other attributes are possible. For example, a
2424 ranked search with a regular expression:
2426 Z> find @attr 1=4 @attr 5=102 @attr 2=102 "informat.* retrieval"
2434 The RecordType parameter in the <literal>zebra.cfg</literal> file, or
2435 the <literal>-t</literal> option to the indexer tells Zebra how to
2436 process input records.
2437 Two basic types of processing are available - raw text and structured
2438 data. Raw text is just that, and it is selected by providing the
2439 argument <literal>text</literal> to Zebra. Structured records are
2440 all handled internally using the basic mechanisms described in the
2441 subsequent sections.
2442 Zebra can read structured records in many different formats.
2448 <section id="querymodel-cql-to-pqf">
2449 <title>Server Side CQL to PQF Query Translation</title>
2452 <literal><cql2rpn>l2rpn.txt</cql2rpn></literal>
2453 YAZ Frontend Virtual
2454 Hosts option, one can configure
2455 the YAZ Frontend CQL-to-PQF
2456 converter, specifying the interpretation of various
2457 <ulink url="&url.cql;">CQL</ulink>
2458 indexes, relations, etc. in terms of Type-1 query attributes.
2459 <!-- The yaz-client config file -->
2462 For example, using server-side CQL-to-PQF conversion, one might
2463 query a zebra server like this:
2466 yaz-client localhost:9999
2468 Z> find text=(plant and soil)
2471 and - if properly configured - even static relevance ranking can
2472 be performed using CQL query syntax:
2475 Z> find text = /relevant (plant and soil)
2481 By the way, the same configuration can be used to
2482 search using client-side CQL-to-PQF conversion:
2483 (the only difference is <literal>querytype cql2rpn</literal>
2485 <literal>querytype cql</literal>, and the call specifying a local
2489 yaz-client -q local/cql2pqf.txt localhost:9999
2490 Z> querytype cql2rpn
2491 Z> find text=(plant and soil)
2497 Exhaustive information can be found in the
2498 Section "Specification of CQL to RPN mappings" in the YAZ manual.
2499 <ulink url="&url.yaz.cql2pqf;"/>,
2500 and shall therefore not be repeated here.
2505 <ulink url="http://www.loc.gov/z3950/agency/zing/cql/dc-indexes.html"/>
2506 for the Maintenance Agency's work-in-progress mapping of Dublin Core
2507 indexes to Attribute Architecture (util, XD and BIB-2)
2515 <!-- Keep this comment at the end of the file
2520 sgml-minimize-attributes:nil
2521 sgml-always-quote-attributes:t
2524 sgml-parent-document: "zebra.xml"
2525 sgml-local-catalogs: nil
2526 sgml-namecase-general:t