1 <chapter id="tools"><title>Supporting Tools</title>
4 In support of the service API - primarily the ASN module, which
5 provides the pro-grammatic interface to the Z39.50 APDUs, &yaz; contains
6 a collection of tools that support the development of applications.
9 <sect1 id="tools.query"><title>Query Syntax Parsers</title>
12 Since the type-1 (RPN) query structure has no direct, useful string
13 representation, every origin application needs to provide some form of
14 mapping from a local query notation or representation to a
15 <token>Z_RPNQuery</token> structure. Some programmers will prefer to
16 construct the query manually, perhaps using
17 <function>odr_malloc()</function> to simplify memory management.
18 The &yaz; distribution includes three separate, query-generating tools
19 that may be of use to you.
22 <sect2 id="PQF"><title>Prefix Query Format</title>
25 Since RPN or reverse polish notation is really just a fancy way of
26 describing a suffix notation format (operator follows operands), it
27 would seem that the confusion is total when we now introduce a prefix
28 notation for RPN. The reason is one of simple laziness - it's somewhat
29 simpler to interpret a prefix format, and this utility was designed
30 for maximum simplicity, to provide a baseline representation for use
31 in simple test applications and scripting environments (like Tcl). The
32 demonstration client included with YAZ uses the PQF.
37 The PQF have been adopted by other parties developing Z39.50
38 software. It is often referred to as Prefix Query Notation
43 The PQF is defined by the pquery module in the YAZ library.
44 There are two sets of function that have similar behavior. First
45 set operates on a PQF parser handle, second set doesn't. First set
46 set of functions are more flexible than the second set. Second set
47 is obsolete and is only provided to ensure backwards compatibility.
50 First set of functions all operate on a PQF parser handle:
53 #include <yaz/pquery.h>
55 YAZ_PQF_Parser yaz_pqf_create (void);
57 void yaz_pqf_destroy (YAZ_PQF_Parser p);
59 Z_RPNQuery *yaz_pqf_parse (YAZ_PQF_Parser p, ODR o, const char *qbuf);
61 Z_AttributesPlusTerm *yaz_pqf_scan (YAZ_PQF_Parser p, ODR o,
62 Odr_oid **attributeSetId, const char *qbuf);
65 int yaz_pqf_error (YAZ_PQF_Parser p, const char **msg, size_t *off);
68 A PQF parser is created and destructed by functions
69 <function>yaz_pqf_create</function> and
70 <function>yaz_pqf_destroy</function> respectively.
71 Function <function>yaz_pqf_parse</function> parses query given
72 by string <literal>qbuf</literal>. If parsing was successful,
73 a Z39.50 RPN Query is returned which is created using ODR stream
74 <literal>o</literal>. If parsing failed, a NULL pointer is
76 Function <function>yaz_pqf_scan</function> takes a scan query in
77 <literal>qbuf</literal>. If parsing was successful, the function
78 returns attributes plus term pointer and modifies
79 <literal>attributeSetId</literal> to hold attribute set for the
80 scan request - both allocated using ODR stream <literal>o</literal>.
81 If parsing failed, yaz_pqf_scan returns a NULL pointer.
82 Error information for bad queries can be obtained by a call to
83 <function>yaz_pqf_error</function> which returns an error code and
84 modifies <literal>*msg</literal> to point to an error description,
85 and modifies <literal>*off</literal> to the offset within last
86 query were parsing failed.
89 The second set of functions are declared as follows:
92 #include <yaz/pquery.h>
94 Z_RPNQuery *p_query_rpn (ODR o, oid_proto proto, const char *qbuf);
96 Z_AttributesPlusTerm *p_query_scan (ODR o, oid_proto proto,
97 Odr_oid **attributeSetP, const char *qbuf);
99 int p_query_attset (const char *arg);
102 The function <function>p_query_rpn()</function> takes as arguments an
103 &odr; stream (see section <link linkend="odr">The ODR Module</link>)
104 to provide a memory source (the structure created is released on
105 the next call to <function>odr_reset()</function> on the stream), a
106 protocol identifier (one of the constants <token>PROTO_Z3950</token> and
107 <token>PROTO_SR</token>), an attribute set reference, and
108 finally a null-terminated string holding the query string.
111 If the parse went well, <function>p_query_rpn()</function> returns a
112 pointer to a <literal>Z_RPNQuery</literal> structure which can be
113 placed directly into a <literal>Z_SearchRequest</literal>.
114 If parsing failed, due to syntax error, a NULL pointer is returned.
117 The <literal>p_query_attset</literal> specifies which attribute set
118 to use if the query doesn't specify one by the
119 <literal>@attrset</literal> operator.
120 The <literal>p_query_attset</literal> returns 0 if the argument is a
121 valid attribute set specifier; otherwise the function returns -1.
125 The grammar of the PQF is as follows:
129 query ::= top-set query-struct.
131 top-set ::= [ '@attrset' string ]
133 query-struct ::= attr-spec | simple | complex | '@term' term-type query
135 attr-spec ::= '@attr' [ string ] string query-struct
137 complex ::= operator query-struct query-struct.
139 operator ::= '@and' | '@or' | '@not' | '@prox' proximity.
141 simple ::= result-set | term.
143 result-set ::= '@set' string.
147 proximity ::= exclusion distance ordered relation which-code unit-code.
149 exclusion ::= '1' | '0' | 'void'.
151 distance ::= integer.
153 ordered ::= '1' | '0'.
155 relation ::= integer.
157 which-code ::= 'known' | 'private' | integer.
159 unit-code ::= integer.
161 term-type ::= 'general' | 'numeric' | 'string' | 'oid' | 'datetime' | 'null'.
165 You will note that the syntax above is a fairly faithful
166 representation of RPN, except for the Attribute, which has been
167 moved a step away from the term, allowing you to associate one or more
168 attributes with an entire query structure. The parser will
169 automatically apply the given attributes to each term as required.
173 The @attr operator is followed by an attribute specification
174 (<literal>attr-spec</literal> above). The specification consists
175 of an optional attribute set, an attribute type-value pair and
176 a sub-query. The attribute type-value pair is packed in one string:
177 an attribute type, an equals sign, and an attribute value, like this:
178 <literal>@attr 1=1003</literal>.
179 The type is always an integer but the value may be either an
180 integer or a string (if it doesn't start with a digit character).
181 A string attribute-value is encoded as a Type-1 ``complex''
182 attribute with the list of values containing the single string
183 specified, and including no semantic indicators.
187 Version 3 of the Z39.50 specification defines various encoding of terms.
188 Use <literal>@term </literal> <replaceable>type</replaceable>
189 <replaceable>string</replaceable>,
190 where type is one of: <literal>general</literal>,
191 <literal>numeric</literal> or <literal>string</literal>
192 (for InternationalString).
193 If no term type has been given, the <literal>general</literal> form
194 is used. This is the only encoding allowed in both versions 2 and 3
195 of the Z39.50 standard.
198 <sect3 id="PQF-prox">
199 <title>Using Proximity Operators with PQF</title>
202 This is an advanced topic, describing how to construct
203 queries that make very specific requirements on the
204 relative location of their operands.
205 You may wish to skip this section and go straight to
206 <link linkend="pqf-examples">the example PQF queries</link>.
211 Most Z39.50 servers do not support proximity searching, or
212 support only a small subset of the full functionality that
213 can be expressed using the PQF proximity operator. Be
214 aware that the ability to <emphasis>express</emphasis> a
215 query in PQF is no guarantee that any given server will
216 be able to <emphasis>execute</emphasis> it.
222 The proximity operator <literal>@prox</literal> is a special
223 and more restrictive version of the conjunction operator
224 <literal>@and</literal>. Its semantics are described in
225 section 3.7.2 (Proximity) of Z39.50 the standard itself, which
226 can be read on-line at
227 <ulink url="&url.z39.50.proximity;"/>
230 In PQF, the proximity operation is represented by a sequence
233 @prox <replaceable>exclusion</replaceable> <replaceable>distance</replaceable> <replaceable>ordered</replaceable> <replaceable>relation</replaceable> <replaceable>which-code</replaceable> <replaceable>unit-code</replaceable>
235 in which the meanings of the parameters are as described in in
236 the standard, and they can take the following values:
238 <listitem><formalpara><title>exclusion</title><para>
239 0 = false (i.e. the proximity condition specified by the
240 remaining parameters must be satisfied) or
241 1 = true (the proximity condition specified by the
242 remaining parameters must <emphasis>not</emphasis> be
244 </para></formalpara></listitem>
245 <listitem><formalpara><title>distance</title><para>
246 An integer specifying the difference between the locations
247 of the operands: e.g. two adjacent words would have
248 distance=1 since their locations differ by one unit.
249 </para></formalpara></listitem>
250 <listitem><formalpara><title>ordered</title><para>
251 1 = ordered (the operands must occur in the order the
252 query specifies them) or
253 0 = unordered (they may appear in either order).
254 </para></formalpara></listitem>
255 <listitem><formalpara><title>relation</title><para>
256 Recognised values are
260 4 (greaterThanOrEqual),
263 </para></formalpara></listitem>
264 <listitem><formalpara><title>which-code</title><para>
265 <literal>known</literal>
268 (the unit-code parameter is taken from the well-known list
269 of alternatives described in below) or
270 <literal>private</literal>
273 (the unit-code paramater has semantics specific to an
274 out-of-band agreement such as a profile).
275 </para></formalpara></listitem>
276 <listitem><formalpara><title>unit-code</title><para>
277 If the which-code parameter is <literal>known</literal>
278 then the recognised values are
290 If which-code is <literal>private</literal> then the
291 acceptable values are determined by the profile.
292 </para></formalpara></listitem>
294 (The numeric values of the relation and well-known unit-code
295 parameters are taken straight from
296 <ulink url="&url.z39.50.proximity.asn1;"
297 >the ASN.1</ulink> of the proximity structure in the standard.)
301 <sect3 id="pqf-examples"><title>PQF queries</title>
303 <example id="example.pqf.simple.terms">
304 <title>PQF queries using simple terms</title>
313 <example id="pqf.example.pqf.boolean.operators">
314 <title>PQF boolean operators</title>
317 @or "dylan" "zimmerman"
319 @and @or dylan zimmerman when
321 @and when @or dylan zimmerman
325 <example id="example.pqf.result.sets">
326 <title>PQF references to result sets</title>
331 @and @set seta @set setb
335 <example id="example.pqf.attributes">
336 <title>Attributes for terms</title>
341 @attr 1=4 @attr 4=1 "self portrait"
343 @attrset exp1 @attr 1=1 CategoryList
345 @attr gils 1=2008 Copenhagen
347 @attr 1=/book/title computer
351 <example id="example.pqf.proximity">
352 <title>PQF Proximity queries</title>
355 @prox 0 3 1 2 k 2 dylan zimmerman
358 Here the parameters 0, 3, 1, 2, k and 2 represent exclusion,
359 distance, ordered, relation, which-code and unit-code, in that
363 exclusion = 0: the proximity condition must hold
366 distance = 3: the terms must be three units apart
369 ordered = 1: they must occur in the order they are specified
372 relation = 2: lessThanOrEqual (to the distance of 3 units)
375 which-code is ``known'', so the standard unit-codes are used
381 So the whole proximity query means that the words
382 <literal>dylan</literal> and <literal>zimmerman</literal> must
383 both occur in the record, in that order, differing in position
384 by three or fewer words (i.e. with two or fewer words between
385 them.) The query would find ``Bob Dylan, aka. Robert
386 Zimmerman'', but not ``Bob Dylan, born as Robert Zimmerman''
387 since the distance in this case is four.
391 <example id="example.pqf.search.term.type">
392 <title>PQF specification of search term type</title>
395 @term string "a UTF-8 string, maybe?"
399 <example id="example.pqf.mixed.queries">
400 <title>PQF mixed queries</title>
403 @or @and bob dylan @set Result-1
405 @attr 4=1 @and @attr 1=1 "bob dylan" @attr 1=4 "slow train coming"
407 @and @attr 2=4 @attr gils 1=2038 -114 @attr 2=2 @attr gils 1=2039 -109
411 The last of these examples is a spatial search: in
412 <ulink url="http://www.gils.net/prof_v2.html#sec_7_4"
413 >the GILS attribute set</ulink>,
415 2038 indicates West Bounding Coordinate and
416 2030 indicates East Bounding Coordinate,
417 so the query is for areas extending from -114 degrees
418 to no more than -109 degrees.
425 <sect2 id="CCL"><title>CCL</title>
428 Not all users enjoy typing in prefix query structures and numerical
429 attribute values, even in a minimalistic test client. In the library
430 world, the more intuitive Common Command Language - CCL (ISO 8777)
431 has enjoyed some popularity - especially before the widespread
432 availability of graphical interfaces. It is still useful in
433 applications where you for some reason or other need to provide a
434 symbolic language for expressing boolean query structures.
437 <sect3 id="ccl.syntax">
438 <title>CCL Syntax</title>
441 The CCL parser obeys the following grammar for the FIND argument.
442 The syntax is annotated by in the lines prefixed by
443 <literal>--</literal>.
447 CCL-Find ::= CCL-Find Op Elements
450 Op ::= "and" | "or" | "not"
451 -- The above means that Elements are separated by boolean operators.
453 Elements ::= '(' CCL-Find ')'
456 | Qualifiers Relation Terms
457 | Qualifiers Relation '(' CCL-Find ')'
458 | Qualifiers '=' string '-' string
459 -- Elements is either a recursive definition, a result set reference, a
460 -- list of terms, qualifiers followed by terms, qualifiers followed
461 -- by a recursive definition or qualifiers in a range (lower - upper).
463 Set ::= 'set' = string
464 -- Reference to a result set
466 Terms ::= Terms Prox Term
468 -- Proximity of terms.
472 -- This basically means that a term may include a blank
474 Qualifiers ::= Qualifiers ',' string
476 -- Qualifiers is a list of strings separated by comma
478 Relation ::= '=' | '>=' | '<=' | '<>' | '>' | '<'
479 -- Relational operators. This really doesn't follow the ISO8777
483 -- Proximity operator
487 <example id="example.ccl.queries">
488 <title>CCL queries</title>
490 The following queries are all valid:
502 (dylan and bob) or set=1
506 Assuming that the qualifiers <literal>ti</literal>,
507 <literal>au</literal>
508 and <literal>date</literal> are defined we may use:
514 au=(bob dylan and slow train coming)
516 date>1980 and (ti=((self portrait)))
522 <sect3 id="ccl.qualifiers">
523 <title>CCL Qualifiers</title>
526 Qualifiers are used to direct the search to a particular searchable
527 index, such as title (ti) and author indexes (au). The CCL standard
528 itself doesn't specify a particular set of qualifiers, but it does
529 suggest a few short-hand notations. You can customize the CCL parser
530 to support a particular set of qualifiers to reflect the current target
531 profile. Traditionally, a qualifier would map to a particular
532 use-attribute within the BIB-1 attribute set. It is also
533 possible to set other attributes, such as the structure
538 A CCL profile is a set of predefined CCL qualifiers that may be
539 read from a file or set in the CCL API.
540 The YAZ client reads its CCL qualifiers from a file named
541 <filename>default.bib</filename>. There are four types of
542 lines in a CCL profile: qualifier specification,
543 qualifier alias, comments and directives.
545 <sect4 id="ccl.qualifier.specification">
546 <title>Qualifier specification</title>
548 A qualifier specification is of the form:
552 <replaceable>qualifier-name</replaceable>
553 [<replaceable>attributeset</replaceable><literal>,</literal>]<replaceable>type</replaceable><literal>=</literal><replaceable>val</replaceable>
554 [<replaceable>attributeset</replaceable><literal>,</literal>]<replaceable>type</replaceable><literal>=</literal><replaceable>val</replaceable> ...
558 where <replaceable>qualifier-name</replaceable> is the name of the
559 qualifier to be used (eg. <literal>ti</literal>),
560 <replaceable>type</replaceable> is attribute type in the attribute
561 set (Bib-1 is used if no attribute set is given) and
562 <replaceable>val</replaceable> is attribute value.
563 The <replaceable>type</replaceable> can be specified as an
564 integer or as it be specified either as a single-letter:
565 <literal>u</literal> for use,
566 <literal>r</literal> for relation,<literal>p</literal> for position,
567 <literal>s</literal> for structure,<literal>t</literal> for truncation
568 or <literal>c</literal> for completeness.
569 The attributes for the special qualifier name <literal>term</literal>
570 are used when no CCL qualifier is given in a query.
571 <table id="ccl.common.bib1.attributes">
572 <title>Common Bib-1 attributes</title>
574 <colspec colwidth="2*" colname="type"></colspec>
575 <colspec colwidth="9*" colname="description"></colspec>
579 <entry>Description</entry>
584 <entry><literal>u=</literal><replaceable>value</replaceable></entry>
586 Use attribute (1). Common use attributes are
587 1 Personal-name, 4 Title, 7 ISBN, 8 ISSN, 30 Date,
588 62 Subject, 1003 Author), 1016 Any. Specify value
594 <entry><literal>r=</literal><replaceable>value</replaceable></entry>
596 Relation attribute (2). Common values are
597 1 <, 2 <=, 3 =, 4 >=, 5 >, 6 <>,
598 100 phonetic, 101 stem, 102 relevance, 103 always matches.
603 <entry><literal>p=</literal><replaceable>value</replaceable></entry>
605 Position attribute (3). Values: 1 first in field, 2
606 first in any subfield, 3 any position in field.
611 <entry><literal>s=</literal><replaceable>value</replaceable></entry>
613 Structure attribute (4). Values: 1 phrase, 2 word,
614 3 key, 4 year, 5 date, 6 word list, 100 date (un),
615 101 name (norm), 102 name (un), 103 structure, 104 urx,
616 105 free-form-text, 106 document-text, 107 local-number,
617 108 string, 109 numeric string.
622 <entry><literal>t=</literal><replaceable>value</replaceable></entry>
624 Truncation attribute (5). Values: 1 right, 2 left,
625 3 left& right, 100 none, 101 process #, 102 regular-1,
626 103 regular-2, 104 CCL.
631 <entry><literal>c=</literal><replaceable>value</replaceable></entry>
633 Completeness attribute (6). Values: 1 incomplete subfield,
634 2 complete subfield, 3 complete field.
643 Refer to <xref linkend="bib1"/> or the complete
644 <ulink url="&url.z39.50.attset.bib1;">list of Bib-1 attributes</ulink>
647 It is also possible to specify non-numeric attribute values,
648 which are used in combination with certain types.
649 The special combinations are:
651 <table id="ccl.special.attribute.combos">
652 <title>Special attribute combos</title>
654 <colspec colwidth="2*" colname="name"></colspec>
655 <colspec colwidth="9*" colname="description"></colspec>
659 <entry>Description</entry>
664 <entry><literal>s=pw</literal></entry><entry>
665 The structure is set to either word or phrase depending
666 on the number of tokens in a term (phrase-word).
670 <entry><literal>s=al</literal></entry><entry>
671 Each token in the term is ANDed. (and-list).
672 This does not set the structure at all.
676 <row><entry><literal>s=ol</literal></entry><entry>
677 Each token in the term is ORed. (or-list).
678 This does not set the structure at all.
682 <row><entry><literal>s=ag</literal></entry><entry>
683 Tokens that appears as phrases (with blank in them) gets
684 structure phrase attached. Tokens that appers as words
685 gets structure phrase attached. Phrases and words are
686 ANDed. This is a variant of s=al and s=pw, with the main
687 difference that words are not split (with operator AND)
688 but instead kept in one RPN token. This facility appeared
693 <row><entry><literal>r=o</literal></entry><entry>
694 Allows ranges and the operators greather-than, less-than, ...
696 This sets Bib-1 relation attribute accordingly (relation
697 ordered). A query construct is only treated as a range if
698 dash is used and that is surrounded by white-space. So
699 <literal>-1980</literal> is treated as term
700 <literal>"-1980"</literal> not <literal><= 1980</literal>.
701 If <literal>- 1980</literal> is used, however, that is
706 <row><entry><literal>r=r</literal></entry><entry>
707 Similar to <literal>r=o</literal> but assumes that terms
708 are non-negative (not prefixed with <literal>-</literal>).
709 Thus, a dash will always be treated as a range.
710 The construct <literal>1980-1990</literal> is
711 treated as a range with <literal>r=r</literal> but as a
712 single term <literal>"1980-1990"</literal> with
713 <literal>r=o</literal>. The special attribute
714 <literal>r=r</literal> is available in YAZ 2.0.24 or later.
718 <row><entry><literal>t=l</literal></entry><entry>
719 Allows term to be left-truncated.
720 If term is of the form <literal>?x</literal>, the resulting
721 Type-1 term is <literal>x</literal> and truncation is left.
725 <row><entry><literal>t=r</literal></entry><entry>
726 Allows term to be right-truncated.
727 If term is of the form <literal>x?</literal>, the resulting
728 Type-1 term is <literal>x</literal> and truncation is right.
732 <row><entry><literal>t=n</literal></entry><entry>
733 If term is does not include <literal>?</literal>, the
734 truncation attribute is set to none (100).
738 <row><entry><literal>t=b</literal></entry><entry>
739 Allows term to be both left&right truncated.
740 If term is of the form <literal>?x?</literal>, the
741 resulting term is <literal>x</literal> and trunctation is
742 set to both left&right.
746 <row><entry><literal>t=x</literal></entry><entry>
747 Allows masking anywhere in a term, thus fully supporting
748 # (mask one character) and ? (zero or more of any).
749 If masking is used, trunction is set to 102 (regexp-1 in term)
750 and the term is converted accordingly to a regular expression.
754 <row><entry><literal>t=z</literal></entry><entry>
755 Allows masking anywhere in a term, thus fully supporting
756 # (mask one character) and ? (zero or more of any).
757 If masking is used, trunction is set to 104 (Z39.58 in term)
758 and the term is converted accordingly to Z39.58 masking term -
759 actually the same truncation as CCL itself.
767 <example id="example.ccl.profile"><title>CCL profile</title>
769 Consider the following definition:
780 <literal>ti</literal> and <literal>au</literal> both set
781 structure attribute to phrase (s=1).
782 <literal>ti</literal>
783 sets the use-attribute to 4. <literal>au</literal> sets the
785 When no qualifiers are used in the query the structure-attribute is
786 set to free-form-text (105) (rule for <literal>term</literal>).
787 The <literal>date</literal> sets the relation attribute to
788 the relation used in the CCL query and sets the use attribute
792 You can combine attributes. To Search for "ranked title" you
795 ti,ranked=knuth computer
797 which will set relation=ranked, use=title, structure=phrase.
804 is a valid query. But
812 <sect4 id="ccl.qualifier.alias">
813 <title>Qualifier alias</title>
815 A qualifier alias is of the form:
818 <replaceable>q</replaceable>
819 <replaceable>q1</replaceable> <replaceable>q2</replaceable> ..
822 which declares <replaceable>q</replaceable> to
823 be an alias for <replaceable>q1</replaceable>,
824 <replaceable>q2</replaceable>... such that the CCL
825 query <replaceable>q=x</replaceable> is equivalent to
826 <replaceable>q1=x or q2=x or ...</replaceable>.
830 <sect4 id="ccl.comments">
831 <title>Comments</title>
833 Lines with white space or lines that begin with
834 character <literal>#</literal> are treated as comments.
838 <sect4 id="ccl.directives">
839 <title>Directives</title>
841 Directive specifications takes the form
843 <para><literal>@</literal><replaceable>directive</replaceable> <replaceable>value</replaceable>
845 <table id="ccl.directives.table">
846 <title>CCL directives</title>
848 <colspec colwidth="2*" colname="name"></colspec>
849 <colspec colwidth="8*" colname="description"></colspec>
850 <colspec colwidth="1*" colname="default"></colspec>
854 <entry>Description</entry>
855 <entry>Default</entry>
860 <entry>truncation</entry>
861 <entry>Truncation character</entry>
862 <entry><literal>?</literal></entry>
866 <entry>Specifies how multiple fields are to be
867 combined. There are two modes: <literal>or</literal>:
868 multiple qualifier fields are ORed,
869 <literal>merge</literal>: attributes for the qualifier
870 fields are merged and assigned to one term.
872 <entry><literal>merge</literal></entry>
876 <entry>Specificies if CCL operatores and qualifiers should be
877 compared with case sensitivity or not. Specify 0 for
878 case sensitive; 1 for case insensitive.</entry>
879 <entry><literal>0</literal></entry>
884 <entry>Specifies token for CCL operator AND.</entry>
885 <entry><literal>and</literal></entry>
890 <entry>Specifies token for CCL operator OR.</entry>
891 <entry><literal>or</literal></entry>
896 <entry>Specifies token for CCL operator NOT.</entry>
897 <entry><literal>not</literal></entry>
902 <entry>Specifies token for CCL operator SET.</entry>
903 <entry><literal>set</literal></entry>
911 <title>CCL API</title>
913 All public definitions can be found in the header file
914 <filename>ccl.h</filename>. A profile identifier is of type
915 <literal>CCL_bibset</literal>. A profile must be created with the call
916 to the function <function>ccl_qual_mk</function> which returns a profile
917 handle of type <literal>CCL_bibset</literal>.
921 To read a file containing qualifier definitions the function
922 <function>ccl_qual_file</function> may be convenient. This function
923 takes an already opened <literal>FILE</literal> handle pointer as
924 argument along with a <literal>CCL_bibset</literal> handle.
928 To parse a simple string with a FIND query use the function
931 struct ccl_rpn_node *ccl_find_str (CCL_bibset bibset, const char *str,
932 int *error, int *pos);
935 which takes the CCL profile (<literal>bibset</literal>) and query
936 (<literal>str</literal>) as input. Upon successful completion the RPN
937 tree is returned. If an error occur, such as a syntax error, the integer
938 pointed to by <literal>error</literal> holds the error code and
939 <literal>pos</literal> holds the offset inside query string in which
944 An English representation of the error may be obtained by calling
945 the <literal>ccl_err_msg</literal> function. The error codes are
946 listed in <filename>ccl.h</filename>.
950 To convert the CCL RPN tree (type
951 <literal>struct ccl_rpn_node *</literal>)
952 to the Z_RPNQuery of YAZ the function <function>ccl_rpn_query</function>
953 must be used. This function which is part of YAZ is implemented in
954 <filename>yaz-ccl.c</filename>.
955 After calling this function the CCL RPN tree is probably no longer
956 needed. The <literal>ccl_rpn_delete</literal> destroys the CCL RPN tree.
960 A CCL profile may be destroyed by calling the
961 <function>ccl_qual_rm</function> function.
965 The token names for the CCL operators may be changed by setting the
966 globals (all type <literal>char *</literal>)
967 <literal>ccl_token_and</literal>, <literal>ccl_token_or</literal>,
968 <literal>ccl_token_not</literal> and <literal>ccl_token_set</literal>.
969 An operator may have aliases, i.e. there may be more than one name for
970 the operator. To do this, separate each alias with a space character.
974 <sect2 id="cql"><title>CQL</title>
976 <ulink url="&url.cql;">CQL</ulink>
977 - Common Query Language - was defined for the
978 <ulink url="&url.sru;">SRU</ulink> protocol.
979 In many ways CQL has a similar syntax to CCL.
980 The objective of CQL is different. Where CCL aims to be
981 an end-user language, CQL is <emphasis>the</emphasis> protocol
982 query language for SRU.
986 If you are new to CQL, read the
987 <ulink url="&url.cql.intro;">Gentle Introduction</ulink>.
991 The CQL parser in &yaz; provides the following:
995 It parses and validates a CQL query.
1000 It generates a C structure that allows you to convert
1001 a CQL query to some other query language, such as SQL.
1006 The parser converts a valid CQL query to PQF, thus providing a
1007 way to use CQL for both SRU servers and Z39.50 targets at the
1013 The parser converts CQL to
1014 <ulink url="&url.xcql;">XCQL</ulink>.
1015 XCQL is an XML representation of CQL.
1016 XCQL is part of the SRU specification. However, since SRU
1017 supports CQL only, we don't expect XCQL to be widely used.
1018 Furthermore, CQL has the advantage over XCQL that it is
1024 <sect3 id="cql.parsing"><title>CQL parsing</title>
1026 A CQL parser is represented by the <literal>CQL_parser</literal>
1027 handle. Its contents should be considered &yaz; internal (private).
1029 #include <yaz/cql.h>
1031 typedef struct cql_parser *CQL_parser;
1033 CQL_parser cql_parser_create(void);
1034 void cql_parser_destroy(CQL_parser cp);
1036 A parser is created by <function>cql_parser_create</function> and
1037 is destroyed by <function>cql_parser_destroy</function>.
1040 To parse a CQL query string, the following function
1043 int cql_parser_string(CQL_parser cp, const char *str);
1045 A CQL query is parsed by the <function>cql_parser_string</function>
1046 which takes a query <parameter>str</parameter>.
1047 If the query was valid (no syntax errors), then zero is returned;
1048 otherwise -1 is returned to indicate a syntax error.
1052 int cql_parser_stream(CQL_parser cp,
1053 int (*getbyte)(void *client_data),
1054 void (*ungetbyte)(int b, void *client_data),
1057 int cql_parser_stdio(CQL_parser cp, FILE *f);
1059 The functions <function>cql_parser_stream</function> and
1060 <function>cql_parser_stdio</function> parses a CQL query
1061 - just like <function>cql_parser_string</function>.
1062 The only difference is that the CQL query can be
1063 fed to the parser in different ways.
1064 The <function>cql_parser_stream</function> uses a generic
1065 byte stream as input. The <function>cql_parser_stdio</function>
1066 uses a <literal>FILE</literal> handle which is opened for reading.
1070 <sect3 id="cql.tree"><title>CQL tree</title>
1072 The the query string is valid, the CQL parser
1073 generates a tree representing the structure of the
1078 struct cql_node *cql_parser_result(CQL_parser cp);
1080 <function>cql_parser_result</function> returns the
1081 a pointer to the root node of the resulting tree.
1084 Each node in a CQL tree is represented by a
1085 <literal>struct cql_node</literal>.
1086 It is defined as follows:
1088 #define CQL_NODE_ST 1
1089 #define CQL_NODE_BOOL 2
1099 struct cql_node *modifiers;
1103 struct cql_node *left;
1104 struct cql_node *right;
1105 struct cql_node *modifiers;
1110 There are two node types: search term (ST) and boolean (BOOL).
1111 A modifier is treated as a search term too.
1114 The search term node has five members:
1118 <literal>index</literal>: index for search term.
1119 If an index is unspecified for a search term,
1120 <literal>index</literal> will be NULL.
1125 <literal>index_uri</literal>: index URi for search term
1126 or NULL if none could be resolved for the index.
1131 <literal>term</literal>: the search term itself.
1136 <literal>relation</literal>: relation for search term.
1141 <literal>relation_uri</literal>: relation URI for search term.
1146 <literal>modifiers</literal>: relation modifiers for search
1147 term. The <literal>modifiers</literal> list itself of cql_nodes
1148 each of type <literal>ST</literal>.
1155 The boolean node represents both <literal>and</literal>,
1156 <literal>or</literal>, not as well as
1161 <literal>left</literal> and <literal>right</literal>: left
1162 - and right operand respectively.
1167 <literal>modifiers</literal>: proximity arguments.
1174 <sect3 id="cql.to.pqf"><title>CQL to PQF conversion</title>
1176 Conversion to PQF (and Z39.50 RPN) is tricky by the fact
1177 that the resulting RPN depends on the Z39.50 target
1178 capabilities (combinations of supported attributes).
1179 In addition, the CQL and SRU operates on index prefixes
1180 (URI or strings), whereas the RPN uses Object Identifiers
1184 The CQL library of &yaz; defines a <literal>cql_transform_t</literal>
1185 type. It represents a particular mapping between CQL and RPN.
1186 This handle is created and destroyed by the functions:
1188 cql_transform_t cql_transform_open_FILE (FILE *f);
1189 cql_transform_t cql_transform_open_fname(const char *fname);
1190 void cql_transform_close(cql_transform_t ct);
1192 The first two functions create a tranformation handle from
1193 either an already open FILE or from a filename respectively.
1196 The handle is destroyed by <function>cql_transform_close</function>
1197 in which case no further reference of the handle is allowed.
1200 When a <literal>cql_transform_t</literal> handle has been created
1201 you can convert to RPN.
1203 int cql_transform_buf(cql_transform_t ct,
1204 struct cql_node *cn, char *out, int max);
1206 This function converts the CQL tree <literal>cn</literal>
1207 using handle <literal>ct</literal>.
1208 For the resulting PQF, you supply a buffer <literal>out</literal>
1209 which must be able to hold at at least <literal>max</literal>
1213 If conversion failed, <function>cql_transform_buf</function>
1214 returns a non-zero SRU error code; otherwise zero is returned
1215 (conversion successful). The meanings of the numeric error
1216 codes are listed in the SRU specifications at
1217 <ulink url="&url.sru.diagnostics.list;"/>
1220 If conversion fails, more information can be obtained by calling
1222 int cql_transform_error(cql_transform_t ct, char **addinfop);
1224 This function returns the most recently returned numeric
1225 error-code and sets the string-pointer at
1226 <literal>*addinfop</literal> to point to a string containing
1227 additional information about the error that occurred: for
1228 example, if the error code is 15 (``Illegal or unsupported context
1229 set''), the additional information is the name of the requested
1230 context set that was not recognised.
1233 The SRU error-codes may be translated into brief human-readable
1234 error messages using
1236 const char *cql_strerror(int code);
1240 If you wish to be able to produce a PQF result in a different
1241 way, there are two alternatives.
1243 void cql_transform_pr(cql_transform_t ct,
1244 struct cql_node *cn,
1245 void (*pr)(const char *buf, void *client_data),
1248 int cql_transform_FILE(cql_transform_t ct,
1249 struct cql_node *cn, FILE *f);
1251 The former function produces output to a user-defined
1252 output stream. The latter writes the result to an already
1253 open <literal>FILE</literal>.
1256 <sect3 id="cql.to.rpn">
1257 <title>Specification of CQL to RPN mappings</title>
1259 The file supplied to functions
1260 <function>cql_transform_open_FILE</function>,
1261 <function>cql_transform_open_fname</function> follows
1262 a structure found in many Unix utilities.
1263 It consists of mapping specifications - one per line.
1264 Lines starting with <literal>#</literal> are ignored (comments).
1267 Each line is of the form
1269 <replaceable>CQL pattern</replaceable><literal> = </literal> <replaceable> RPN equivalent</replaceable>
1273 An RPN pattern is a simple attribute list. Each attribute pair
1276 [<replaceable>set</replaceable>] <replaceable>type</replaceable><literal>=</literal><replaceable>value</replaceable>
1278 The attribute <replaceable>set</replaceable> is optional.
1279 The <replaceable>type</replaceable> is the attribute type,
1280 <replaceable>value</replaceable> the attribute value.
1283 The character <literal>*</literal> (asterisk) has special meaning
1284 when used in the RPN pattern.
1285 Each occurrence of <literal>*</literal> is substituted with the
1286 CQL matching name (index, relation, qualifier etc).
1287 This facility can be used to copy a CQL name verbatim to the RPN result.
1290 The following CQL patterns are recognized:
1292 <varlistentry><term>
1293 <literal>index.</literal><replaceable>set</replaceable><literal>.</literal><replaceable>name</replaceable>
1297 This pattern is invoked when a CQL index, such as
1298 dc.title is converted. <replaceable>set</replaceable>
1299 and <replaceable>name</replaceable> are the context set and index
1301 Typically, the RPN specifies an equivalent use attribute.
1304 For terms not bound by an index the pattern
1305 <literal>index.cql.serverChoice</literal> is used.
1306 Here, the prefix <literal>cql</literal> is defined as
1307 <literal>http://www.loc.gov/zing/cql/cql-indexes/v1.0/</literal>.
1308 If this pattern is not defined, the mapping will fail.
1312 <literal>index.</literal><replaceable>set</replaceable><literal>.*</literal>
1313 is used when no other index pattern is matched.
1317 <varlistentry><term>
1318 <literal>qualifier.</literal><replaceable>set</replaceable><literal>.</literal><replaceable>name</replaceable>
1323 For backwards compatibility, this is recognised as a synonym of
1324 <literal>index.</literal><replaceable>set</replaceable><literal>.</literal><replaceable>name</replaceable>
1328 <varlistentry><term>
1329 <literal>relation.</literal><replaceable>relation</replaceable>
1333 This pattern specifies how a CQL relation is mapped to RPN.
1334 <replaceable>pattern</replaceable> is name of relation
1335 operator. Since <literal>=</literal> is used as
1336 separator between CQL pattern and RPN, CQL relations
1337 including <literal>=</literal> cannot be
1338 used directly. To avoid a conflict, the names
1339 <literal>ge</literal>,
1340 <literal>eq</literal>,
1341 <literal>le</literal>,
1342 must be used for CQL operators, greater-than-or-equal,
1343 equal, less-than-or-equal respectively.
1344 The RPN pattern is supposed to include a relation attribute.
1347 For terms not bound by a relation, the pattern
1348 <literal>relation.scr</literal> is used. If the pattern
1349 is not defined, the mapping will fail.
1352 The special pattern, <literal>relation.*</literal> is used
1353 when no other relation pattern is matched.
1358 <varlistentry><term>
1359 <literal>relationModifier.</literal><replaceable>mod</replaceable>
1363 This pattern specifies how a CQL relation modifier is mapped to RPN.
1364 The RPN pattern is usually a relation attribute.
1369 <varlistentry><term>
1370 <literal>structure.</literal><replaceable>type</replaceable>
1374 This pattern specifies how a CQL structure is mapped to RPN.
1375 Note that this CQL pattern is somewhat to similar to
1376 CQL pattern <literal>relation</literal>.
1377 The <replaceable>type</replaceable> is a CQL relation.
1380 The pattern, <literal>structure.*</literal> is used
1381 when no other structure pattern is matched.
1382 Usually, the RPN equivalent specifies a structure attribute.
1387 <varlistentry><term>
1388 <literal>position.</literal><replaceable>type</replaceable>
1392 This pattern specifies how the anchor (position) of
1393 CQL is mapped to RPN.
1394 The <replaceable>type</replaceable> is one
1395 of <literal>first</literal>, <literal>any</literal>,
1396 <literal>last</literal>, <literal>firstAndLast</literal>.
1399 The pattern, <literal>position.*</literal> is used
1400 when no other position pattern is matched.
1405 <varlistentry><term>
1406 <literal>set.</literal><replaceable>prefix</replaceable>
1410 This specification defines a CQL context set for a given prefix.
1411 The value on the right hand side is the URI for the set -
1412 <emphasis>not</emphasis> RPN. All prefixes used in
1413 index patterns must be defined this way.
1418 <varlistentry><term>
1419 <literal>set</literal>
1423 This specification defines a default CQL context set for index names.
1424 The value on the right hand side is the URI for the set.
1431 <example id="example.cql.to.rpn.mapping">
1432 <title>CQL to RPN mapping file</title>
1434 This simple file defines two context sets, three indexes and three
1435 relations, a position pattern and a default structure.
1437 <programlisting><![CDATA[
1438 set.cql = http://www.loc.gov/zing/cql/context-sets/cql/v1.1/
1439 set.dc = http://www.loc.gov/zing/cql/dc-indexes/v1.0/
1441 index.cql.serverChoice = 1=1016
1442 index.dc.title = 1=4
1443 index.dc.subject = 1=21
1449 position.any = 3=3 6=1
1455 With the mappings above, the CQL query
1459 is converted to the PQF:
1461 @attr 1=1016 @attr 2=3 @attr 4=1 @attr 3=3 @attr 6=1 "computer"
1463 by rules <literal>index.cql.serverChoice</literal>,
1464 <literal>relation.scr</literal>, <literal>structure.*</literal>,
1465 <literal>position.any</literal>.
1472 is rejected, since <literal>position.right</literal> is
1478 >my = "http://www.loc.gov/zing/cql/dc-indexes/v1.0/" my.title = x
1482 @attr 1=4 @attr 2=3 @attr 4=1 @attr 3=3 @attr 6=1 "x"
1486 <example id="example.cql.to.rpn.string">
1487 <title>CQL to RPN string attributes</title>
1489 In this example we allow any index to be passed to RPN as
1492 <programlisting><![CDATA[
1493 # Identifiers for prefixes used in this file. (index.*)
1494 set.cql = info:srw/cql-context-set/1/cql-v1.1
1495 set.rpn = http://bogus/rpn
1496 set = http://bogus/rpn
1498 # The default index when none is specified by the query
1499 index.cql.serverChoice = 1=any
1508 The <literal>http://bogus/rpn</literal> context set is also the default
1509 so we can make queries such as
1513 which is converted to
1515 @attr 2=3 @attr 4=1 @attr 3=3 @attr 1=title "a"
1519 <example id="example.cql.to.rpn.bathprofile">
1520 <title>CQL to RPN using Bath Profile</title>
1522 The file <filename>etc/pqf.properties</filename> has mappings from
1523 the Bath Profile and Dublin Core to RPN.
1524 If YAZ is installed as a package it's usually located
1525 in <filename>/usr/share/yaz/etc</filename> and part of the
1526 development package, such as <literal>libyaz-dev</literal>.
1530 <sect3 id="cql.xcql"><title>CQL to XCQL conversion</title>
1532 Conversion from CQL to XCQL is trivial and does not
1533 require a mapping to be defined.
1534 There three functions to choose from depending on the
1535 way you wish to store the resulting output (XML buffer
1538 int cql_to_xml_buf(struct cql_node *cn, char *out, int max);
1539 void cql_to_xml(struct cql_node *cn,
1540 void (*pr)(const char *buf, void *client_data),
1542 void cql_to_xml_stdio(struct cql_node *cn, FILE *f);
1544 Function <function>cql_to_xml_buf</function> converts
1545 to XCQL and stores result in a user supplied buffer of a given
1549 <function>cql_to_xml</function> writes the result in
1550 a user defined output stream.
1551 <function>cql_to_xml_stdio</function> writes to a
1557 <sect1 id="tools.oid"><title>Object Identifiers</title>
1560 The basic YAZ representation of an OID is an array of integers,
1561 terminated with the value -1. This integer is of type
1562 <literal>Odr_oid</literal>.
1565 Fundamental OID operations and the type <literal>Odr_oid</literal>
1566 are defined in <filename>yaz/oid_util.h</filename>.
1569 An OID can either be declared as a automatic variable or it can
1570 allocated using the memory utilities or ODR/NMEM. It's
1571 guaranteed that an OID can fit in <literal>OID_SIZE</literal> integers.
1573 <example id="tools.oid.bib1.1"><title>Create OID on stack</title>
1575 We can create an OID for the Bib-1 attribute set with:
1577 Odr_oid bib1[OID_SIZE];
1589 And OID may also be filled from a string-based representation using
1590 dots (.). This is achieved by function
1592 int oid_dotstring_to_oid(const char *name, Odr_oid *oid);
1594 This functions returns 0 if name could be converted; -1 otherwise.
1596 <example id="tools.oid.bib1.2"><title>Using oid_oiddotstring_to_oid</title>
1598 We can fill the Bib-1 attribute set OID easier with:
1600 Odr_oid bib1[OID_SIZE];
1601 oid_oiddotstring_to_oid("1.2.840.10003.3.1", bib1);
1606 We can also allocate an OID dynamically on a ODR stream with:
1608 Odr_oid *odr_getoidbystr(ODR o, const char *str);
1610 This creates an OID from string-based representation using dots.
1611 This function take an &odr; stream as parameter. This stream is used to
1612 allocate memory for the data elements, which is released on a
1613 subsequent call to <function>odr_reset()</function> on that stream.
1616 <example id="tools.oid.bib1.3"><title>Using odr_getoidbystr</title>
1618 We can create a OID for the Bib-1 attribute set with:
1620 Odr_oid *bib1 = odr_getoidbystr(odr, "1.2.840.10003.3.1");
1628 char *oid_oid_to_dotstring(const Odr_oid *oid, char *oidbuf)
1630 does the reverse of <function>oid_oiddotstring_to_oid</function>. It
1631 converts an OID to the string-based representation using dots.
1632 The supplied char buffer <literal>oidbuf</literal> holds the resulting
1633 string and must be at least <literal>OID_STR_MAX</literal> in size.
1637 OIDs can be copied with <function>oid_oidcpy</function> which takes
1638 two OID lists as arguments. Alternativly, an OID copy can be allocated
1639 on a ODR stream with:
1641 Odr_oid *odr_oiddup(ODR odr, const Odr_oid *o);
1646 OIDs can be compared with <function>oid_oidcmp</function> which returns
1647 zero if the two OIDs provided are identical; non-zero otherwise.
1650 <sect2 id="tools.oid.database"><title>OID database</title>
1652 From YAZ version 3 and later, the oident system has been replaced
1653 by an OID database. OID database is a misnomer .. the old odient
1654 system was also a database.
1657 The OID database is really just a map between named Object Identifiers
1658 (string) and their OID raw equivalents. Most operations either
1659 convert from string to OID or other way around.
1662 Unfortunately, whenever we supply a string we must also specify the
1663 <emphasis>OID class</emphasis>. The class is necessary because some
1664 strings correspond to multiple OIDs. An example of such a string is
1665 <literal>Bib-1</literal> which may either be an attribute-set
1666 or a diagnostic-set.
1669 Applications using the YAZ database should include
1670 <filename>yaz/oid_db.h</filename>.
1673 A YAZ database handle is of type <literal>yaz_oid_db_t</literal>.
1674 Actually that's a pointer. You need not think deal with that.
1675 YAZ has a built-in database which can be considered "constant" for
1677 We can get hold that by using function <function>yaz_oid_std</function>.
1680 All functions with prefix <function>yaz_string_to_oid</function>
1681 converts from class + string to OID. We have variants of this
1682 operation due to different memory allocation strategies.
1685 All functions with prefix
1686 <function>yaz_oid_to_string</function> converts from OID to string
1690 <example id="tools.oid.bib1.4"><title>Create OID with YAZ DB</title>
1692 We can create an OID for the Bib-1 attribute set on the ODR stream
1696 yaz_string_to_oid_odr(yaz_oid_std(), CLASS_ATTSET, "Bib-1", odr);
1698 This is more complex than using <function>odr_getoidbystr</function>.
1699 You would only use <function>yaz_string_to_oid_odr</function> when the
1700 string (here Bib-1) is supplied by a user or configuration.
1705 <sect2 id="tools.oid.std"><title>Standard OIDs</title>
1708 All the object identifers in the standard OID database as returned
1709 by <function>yaz_oid_std</function> can referenced directly in a
1710 program as a constant OID.
1711 Each constant OID is prefixed with <literal>yaz_oid_</literal> -
1712 followed by OID class (lowercase) - then by OID name (normalized and
1716 See <xref linkend="list-oids"/> for list of all object identifiers
1718 These are declared in <filename>yaz/oid_std.h</filename> but are
1719 included by <filename>yaz/oid_db.h</filename> as well.
1722 <example id="tools.oid.bib1.5"><title>Use a built-in OID</title>
1724 We can allocate our own OID filled with the constant OID for
1727 Odr_oid *bib1 = odr_oiddup(o, yaz_oid_attset_bib1);
1733 <sect1 id="tools.nmem"><title>Nibble Memory</title>
1736 Sometimes when you need to allocate and construct a large,
1737 interconnected complex of structures, it can be a bit of a pain to
1738 release the associated memory again. For the structures describing the
1739 Z39.50 PDUs and related structures, it is convenient to use the
1740 memory-management system of the &odr; subsystem (see
1741 <xref linkend="odr.use"/>). However, in some circumstances
1742 where you might otherwise benefit from using a simple nibble memory
1743 management system, it may be impractical to use
1744 <function>odr_malloc()</function> and <function>odr_reset()</function>.
1745 For this purpose, the memory manager which also supports the &odr;
1746 streams is made available in the NMEM module. The external interface
1747 to this module is given in the <filename>nmem.h</filename> file.
1751 The following prototypes are given:
1755 NMEM nmem_create(void);
1756 void nmem_destroy(NMEM n);
1757 void *nmem_malloc(NMEM n, size_t size);
1758 void nmem_reset(NMEM n);
1759 size_t nmem_total(NMEM n);
1760 void nmem_init(void);
1761 void nmem_exit(void);
1765 The <function>nmem_create()</function> function returns a pointer to a
1766 memory control handle, which can be released again by
1767 <function>nmem_destroy()</function> when no longer needed.
1768 The function <function>nmem_malloc()</function> allocates a block of
1769 memory of the requested size. A call to <function>nmem_reset()</function>
1770 or <function>nmem_destroy()</function> will release all memory allocated
1771 on the handle since it was created (or since the last call to
1772 <function>nmem_reset()</function>. The function
1773 <function>nmem_total()</function> returns the number of bytes currently
1774 allocated on the handle.
1778 The nibble memory pool is shared amongst threads. POSIX
1779 mutex'es and WIN32 Critical sections are introduced to keep the
1780 module thread safe. Function <function>nmem_init()</function>
1781 initializes the nibble memory library and it is called automatically
1782 the first time the <literal>YAZ.DLL</literal> is loaded. &yaz; uses
1783 function <function>DllMain</function> to achieve this. You should
1784 <emphasis>not</emphasis> call <function>nmem_init</function> or
1785 <function>nmem_exit</function> unless you're absolute sure what
1786 you're doing. Note that in previous &yaz; versions you'd have to call
1787 <function>nmem_init</function> yourself.
1792 <sect1 id="tools.log"><title>Log</title>
1794 &yaz; has evolved a fairly complex log system which should be useful both
1795 for debugging &yaz; itself, debugging applications that use &yaz;, and for
1796 production use of those applications.
1799 The log functions are declared in header <filename>yaz/log.h</filename>
1800 and implemented in <filename>src/log.c</filename>.
1801 Due to name clash with syslog and some math utilities the logging
1802 interface has been modified as of YAZ 2.0.29. The obsolete interface
1803 is still available if in header file <filename>yaz/log.h</filename>.
1804 The key points of the interface are:
1807 void yaz_log(int level, const char *fmt, ...)
1809 void yaz_log_init(int level, const char *prefix, const char *name);
1810 void yaz_log_init_file(const char *fname);
1811 void yaz_log_init_level(int level);
1812 void yaz_log_init_prefix(const char *prefix);
1813 void yaz_log_time_format(const char *fmt);
1814 void yaz_log_init_max_size(int mx);
1816 int yaz_log_mask_str(const char *str);
1817 int yaz_log_module_level(const char *name);
1821 The reason for the whole log module is the <function>yaz_log</function>
1822 function. It takes a bitmask indicating the log levels, a
1823 <literal>printf</literal>-like format string, and a variable number of
1828 The <literal>log level</literal> is a bit mask, that says on which level(s)
1829 the log entry should be made, and optionally set some behaviour of the
1830 logging. In the most simple cases, it can be one of <literal>YLOG_FATAL,
1831 YLOG_DEBUG, YLOG_WARN, YLOG_LOG</literal>. Those can be combined with bits
1832 that modify the way the log entry is written:<literal>YLOG_ERRNO,
1833 YLOG_NOTIME, YLOG_FLUSH</literal>.
1834 Most of the rest of the bits are deprecated, and should not be used. Use
1835 the dynamic log levels instead.
1839 Applications that use &yaz;, should not use the LOG_LOG for ordinary
1840 messages, but should make use of the dynamic loglevel system. This consists
1841 of two parts, defining the loglevel and checking it.
1845 To define the log levels, the (main) program should pass a string to
1846 <function>yaz_log_mask_str</function> to define which log levels are to be
1847 logged. This string should be a comma-separated list of log level names,
1848 and can contain both hard-coded names and dynamic ones. The log level
1849 calculation starts with <literal>YLOG_DEFAULT_LEVEL</literal> and adds a bit
1850 for each word it meets, unless the word starts with a '-', in which case it
1851 clears the bit. If the string <literal>'none'</literal> is found,
1852 all bits are cleared. Typically this string comes from the command-line,
1853 often identified by <literal>-v</literal>. The
1854 <function>yaz_log_mask_str</function> returns a log level that should be
1855 passed to <function>yaz_log_init_level</function> for it to take effect.
1859 Each module should check what log bits it should be used, by calling
1860 <function>yaz_log_module_level</function> with a suitable name for the
1861 module. The name is cleared from a preceding path and an extension, if any,
1862 so it is quite possible to use <literal>__FILE__</literal> for it. If the
1863 name has been passed to <function>yaz_log_mask_str</function>, the routine
1864 returns a non-zero bitmask, which should then be used in consequent calls
1865 to yaz_log. (It can also be tested, so as to avoid unnecessary calls to
1866 yaz_log, in time-critical places, or when the log entry would take time
1871 Yaz uses the following dynamic log levels:
1872 <literal>server, session, request, requestdetail</literal> for the server
1874 <literal>zoom</literal> for the zoom client api.
1875 <literal>ztest</literal> for the simple test server.
1876 <literal>malloc, nmem, odr, eventl</literal> for internal debugging of yaz itself.
1877 Of course, any program using yaz is welcome to define as many new ones, as
1882 By default the log is written to stderr, but this can be changed by a call
1883 to <function>yaz_log_init_file</function> or
1884 <function>yaz_log_init</function>. If the log is directed to a file, the
1885 file size is checked at every write, and if it exceeds the limit given in
1886 <function>yaz_log_init_max_size</function>, the log is rotated. The
1887 rotation keeps one old version (with a <literal>.1</literal> appended to
1888 the name). The size defaults to 1GB. Setting it to zero will disable the
1893 A typical yaz-log looks like this
1894 13:23:14-23/11 yaz-ztest(1) [session] Starting session from tcp:127.0.0.1 (pid=30968)
1895 13:23:14-23/11 yaz-ztest(1) [request] Init from 'YAZ' (81) (ver 2.0.28) OK
1896 13:23:17-23/11 yaz-ztest(1) [request] Search Z: @attrset Bib-1 foo OK:7 hits
1897 13:23:22-23/11 yaz-ztest(1) [request] Present: [1] 2+2 OK 2 records returned
1898 13:24:13-23/11 yaz-ztest(1) [request] Close OK
1902 The log entries start with a time stamp. This can be omitted by setting the
1903 <literal>YLOG_NOTIME</literal> bit in the loglevel. This way automatic tests
1904 can be hoped to produce identical log files, that are easy to diff. The
1905 format of the time stamp can be set with
1906 <function>yaz_log_time_format</function>, which takes a format string just
1907 like <function>strftime</function>.
1911 Next in a log line comes the prefix, often the name of the program. For
1912 yaz-based servers, it can also contain the session number. Then
1913 comes one or more logbits in square brackets, depending on the logging
1914 level set by <function>yaz_log_init_level</function> and the loglevel
1915 passed to <function>yaz_log_init_level</function>. Finally comes the format
1916 string and additional values passed to <function>yaz_log</function>
1920 The log level <literal>YLOG_LOGLVL</literal>, enabled by the string
1921 <literal>loglevel</literal>, will log all the log-level affecting
1922 operations. This can come in handy if you need to know what other log
1923 levels would be useful. Grep the logfile for <literal>[loglevel]</literal>.
1927 The log system is almost independent of the rest of &yaz;, the only
1928 important dependence is of <filename>nmem</filename>, and that only for
1929 using the semaphore definition there.
1933 The dynamic log levels and log rotation were introduced in &yaz; 2.0.28. At
1934 the same time, the log bit names were changed from
1935 <literal>LOG_something</literal> to <literal>YLOG_something</literal>,
1936 to avoid collision with <filename>syslog.h</filename>.
1941 <sect1 id="marc"><title>MARC</title>
1944 YAZ provides a fast utility for working with MARC records.
1945 Early versions of the MARC utility only allowed decoding of ISO2709.
1946 Today the utility may both encode - and decode to a varity of formats.
1949 #include <yaz/marcdisp.h>
1951 /* create handler */
1952 yaz_marc_t yaz_marc_create(void);
1954 void yaz_marc_destroy(yaz_marc_t mt);
1956 /* set XML mode YAZ_MARC_LINE, YAZ_MARC_SIMPLEXML, ... */
1957 void yaz_marc_xml(yaz_marc_t mt, int xmlmode);
1958 #define YAZ_MARC_LINE 0
1959 #define YAZ_MARC_SIMPLEXML 1
1960 #define YAZ_MARC_OAIMARC 2
1961 #define YAZ_MARC_MARCXML 3
1962 #define YAZ_MARC_ISO2709 4
1963 #define YAZ_MARC_XCHANGE 5
1964 #define YAZ_MARC_CHECK 6
1965 #define YAZ_MARC_TURBOMARC 7
1967 /* supply iconv handle for character set conversion .. */
1968 void yaz_marc_iconv(yaz_marc_t mt, yaz_iconv_t cd);
1970 /* set debug level, 0=none, 1=more, 2=even more, .. */
1971 void yaz_marc_debug(yaz_marc_t mt, int level);
1973 /* decode MARC in buf of size bsize. Returns >0 on success; <=0 on failure.
1974 On success, result in *result with size *rsize. */
1975 int yaz_marc_decode_buf(yaz_marc_t mt, const char *buf, int bsize,
1976 const char **result, size_t *rsize);
1978 /* decode MARC in buf of size bsize. Returns >0 on success; <=0 on failure.
1979 On success, result in WRBUF */
1980 int yaz_marc_decode_wrbuf(yaz_marc_t mt, const char *buf,
1981 int bsize, WRBUF wrbuf);
1986 The synopsis is just a basic subset of all functionality. Refer
1987 to the actual header file <filename>marcdisp.h</filename> for
1992 A MARC conversion handle must be created by using
1993 <function>yaz_marc_create</function> and destroyed
1994 by calling <function>yaz_marc_destroy</function>.
1997 All other function operate on a <literal>yaz_marc_t</literal> handle.
1998 The output is specified by a call to <function>yaz_marc_xml</function>.
1999 The <literal>xmlmode</literal> must be one of
2002 <term>YAZ_MARC_LINE</term>
2005 A simple line-by-line format suitable for display but not
2006 recommend for further (machine) processing.
2012 <term>YAZ_MARC_MARCXML</term>
2015 <ulink url="&url.marcxml;">MARCXML</ulink>.
2021 <term>YAZ_MARC_ISO2709</term>
2024 ISO2709 (sometimes just referred to as "MARC").
2030 <term>YAZ_MARC_XCHANGE</term>
2033 <ulink url="&url.marcxchange;">MarcXchange</ulink>.
2039 <term>YAZ_MARC_CHECK</term>
2042 Pseudo format for validation only. Does not generate
2043 any real output except diagnostics.
2049 <term>YAZ_MARC_TURBOMARC</term>
2052 XML format with same semantics as MARCXML but more compact
2053 and geared towards fast processing with XSLT. Refer to
2054 <xref linkend="tools.turbomarc"/> for more information.
2062 The actual conversion functions are
2063 <function>yaz_marc_decode_buf</function> and
2064 <function>yaz_marc_decode_wrbuf</function> which decodes and encodes
2065 a MARC record. The former function operates on simple buffers, the
2066 stores the resulting record in a WRBUF handle (WRBUF is a simple string
2069 <example id="example.marc.display">
2070 <title>Display of MARC record</title>
2072 The following program snippet illustrates how the MARC API may
2073 be used to convert a MARC record to the line-by-line format:
2074 <programlisting><![CDATA[
2075 void print_marc(const char *marc_buf, int marc_buf_size)
2077 char *result; /* for result buf */
2078 size_t result_len; /* for size of result */
2079 yaz_marc_t mt = yaz_marc_create();
2080 yaz_marc_xml(mt, YAZ_MARC_LINE);
2081 yaz_marc_decode_buf(mt, marc_buf, marc_buf_size,
2082 &result, &result_len);
2083 fwrite(result, result_len, 1, stdout);
2084 yaz_marc_destroy(mt); /* note that result is now freed... */
2090 <sect2 id="tools.turbomarc">
2091 <title>TurboMARC</title>
2093 TurboMARC is yet another XML encoding of a MARC record. The format
2094 was designed for fast processing with XSLT.
2098 Pazpar2 uses XSLT to convert an XML encoded MARC record to an internal
2099 representation. This conversion mostly check the tag of a MARC field
2100 to determine the basic rules in the conversion. This check is
2101 costly when that is tag is encoded as an attribute in MARCXML.
2102 By having the tag value as the element instead, makes processing
2103 many times faster (at least for Libxslt).
2106 TurboMARC is encoded as follows:
2109 Record elements is part of namespace
2110 "<literal>http://www.indexdata.com/turbomarc</literal>".
2113 A record is enclosed in element <literal>r</literal>.
2116 A collection of records is enclosed in element
2117 <literal>collection</literal>.
2120 The leader is encoded as element <literal>l</literal> with the
2121 leader content as its (text) value.
2124 A control field is encoded as element <literal>c</literal> concatenated
2125 with the tag value of the control field if the tag value
2126 matches the regular expression <literal>[a-zA-Z0-9]*</literal>.
2127 If the tag value do not match the regular expression
2128 <literal>[a-zA-Z0-9]*</literal> the control field is encoded
2129 as element <literal>c</literal> and attribute <literal>code</literal>
2130 will hold the tag value.
2131 This rule ensure that in the rare cases where a tag value might
2132 result in a non-wellformed XML YAZ encode it as a coded attribute
2136 The control field content is the the text value of this element.
2137 Indicators are encoded as attribute names
2138 <literal>i1</literal>, <literal>i2</literal>, etc.. and
2139 corresponding values for each indicator.
2142 A data field is encoded as element <literal>d</literal> concatenated
2143 with the tag value of the data field or using the attribute
2144 <literal>code</literal> as described in the rules for control fields.
2145 The children of the data field element is subfield elements.
2146 Each subfield element is encoded as <literal>s</literal>
2147 concatenated with the sub field code.
2148 The text of the subfield element is the contents of the subfield.
2149 Indicators are encoded as attributes for the data field element similar
2150 to the encoding for control fields.
2157 <sect1 id="tools.retrieval">
2158 <title>Retrieval Facility</title>
2160 YAZ version 2.1.20 or later includes a Retrieval facility tool
2161 which allows a SRU/Z39.50 to describe itself and perform record
2162 conversions. The idea is the following:
2167 An SRU/Z39.50 client sends a retrieval request which includes
2168 a combination of the following parameters: syntax (format),
2169 schema (or element set name).
2175 The retrieval facility is invoked with parameters in a
2176 server/proxy. The retrieval facility matches the parameters a set of
2177 "supported" retrieval types.
2178 If there is no match, the retrieval signals an error
2179 (syntax and / or schema not supported).
2185 For a successful match, the backend is invoked with the same
2186 or altered retrieval parameters (syntax, schema). If
2187 a record is received from the backend, it is converted to the
2188 frontend name / syntax.
2194 The resulting record is sent back the client and tagged with
2195 the frontend syntax / schema.
2202 The Retrieval facility is driven by an XML configuration. The
2203 configuration is neither Z39.50 ZeeRex or SRU ZeeRex. But it
2204 should be easy to generate both of them from the XML configuration.
2205 (unfortunately the two versions
2206 of ZeeRex differ substantially in this regard).
2208 <sect2 id="tools.retrieval.format">
2209 <title>Retrieval XML format</title>
2211 All elements should be covered by namespace
2212 <literal>http://indexdata.com/yaz</literal> .
2213 The root element node must be <literal>retrievalinfo</literal>.
2216 The <literal>retrievalinfo</literal> must include one or
2217 more <literal>retrieval</literal> elements. Each
2218 <literal>retrieval</literal> defines specific combination of
2219 syntax, name and identifier supported by this retrieval service.
2222 The <literal>retrieval</literal> element may include any of the
2223 following attributes:
2225 <varlistentry><term><literal>syntax</literal> (REQUIRED)</term>
2228 Defines the record syntax. Possible values is any
2229 of the names defined in YAZ' OID database or a raw
2234 <varlistentry><term><literal>name</literal> (OPTIONAL)</term>
2237 Defines the name of the retrieval format. This can be
2238 any string. For SRU, the value, is equivalent to schema (short-hand);
2239 for Z39.50 it's equivalent to simple element set name.
2240 For YAZ 3.0.24 and later this name may be specified as a glob
2241 expression with operators
2242 <literal>*</literal> and <literal>?</literal>.
2246 <varlistentry><term><literal>identifier</literal> (OPTIONAL)</term>
2249 Defines the URI schema name of the retrieval format. This can be
2250 any string. For SRU, the value, is equivalent to URI schema.
2251 For Z39.50, there is no equivalent.
2258 The <literal>retrieval</literal> may include one
2259 <literal>backend</literal> element. If a <literal>backend</literal>
2260 element is given, it specifies how the records are retrieved by
2261 some backend and how the records are converted from the backend to
2265 The attributes, <literal>name</literal> and <literal>syntax</literal>
2266 may be specified for the <literal>backend</literal> element. These
2267 semantics of these attributes is equivalent to those for the
2268 <literal>retrieval</literal>. However, these values are passed to
2272 The <literal>backend</literal> element may includes one or more
2273 conversion instructions (as children elements). The supported
2276 <varlistentry><term><literal>marc</literal></term>
2279 The <literal>marc</literal> element specifies a conversion
2280 to - and from ISO2709 encoded MARC and
2281 <ulink url="&url.marcxml;">&acro.marcxml;</ulink>/MarcXchange.
2282 The following attributes may be specified:
2285 <varlistentry><term><literal>inputformat</literal> (REQUIRED)</term>
2288 Format of input. Supported values are
2289 <literal>marc</literal> (for ISO2709); and <literal>xml</literal>
2290 for MARCXML/MarcXchange.
2295 <varlistentry><term><literal>outputformat</literal> (REQUIRED)</term>
2298 Format of output. Supported values are
2299 <literal>line</literal> (MARC line format);
2300 <literal>marcxml</literal> (for MARCXML),
2301 <literal>marc</literal> (ISO2709),
2302 <literal>marcxhcange</literal> (for MarcXchange).
2307 <varlistentry><term><literal>inputcharset</literal> (OPTIONAL)</term>
2310 Encoding of input. For XML input formats, this need not
2311 be given, but for ISO2709 based inputformats, this should
2312 be set to the encoding used. For MARC21 records, a common
2313 inputcharset value would be <literal>marc-8</literal>.
2318 <varlistentry><term><literal>outputcharset</literal> (OPTIONAL)</term>
2321 Encoding of output. If outputformat is XML based, it is
2322 strongly recommened to use <literal>utf-8</literal>.
2331 <varlistentry><term><literal>xslt</literal></term>
2334 The <literal>xslt</literal> element specifies a conversion
2335 via &acro.xslt;. The following attributes may be specified:
2338 <varlistentry><term><literal>stylesheet</literal> (REQUIRED)</term>
2353 <sect2 id="tools.retrieval.examples">
2354 <title>Retrieval Facility Examples</title>
2355 <example id="tools.retrieval.marc21">
2356 <title>MARC21 backend</title>
2358 A typical way to use the retrieval facility is to enable XML
2359 for servers that only supports ISO2709 encoded MARC21 records.
2361 <programlisting><![CDATA[
2363 <retrieval syntax="usmarc" name="F"/>
2364 <retrieval syntax="usmarc" name="B"/>
2365 <retrieval syntax="xml" name="marcxml"
2366 identifier="info:srw/schema/1/marcxml-v1.1">
2367 <backend syntax="usmarc" name="F">
2368 <marc inputformat="marc" outputformat="marcxml"
2369 inputcharset="marc-8"/>
2372 <retrieval syntax="xml" name="dc">
2373 <backend syntax="usmarc" name="F">
2374 <marc inputformat="marc" outputformat="marcxml"
2375 inputcharset="marc-8"/>
2376 <xslt stylesheet="MARC21slim2DC.xsl"/>
2383 This means that our frontend supports:
2387 MARC21 F(ull) records.
2392 MARC21 B(rief) records.
2404 Dublin core records.
2411 <sect2 id="tools.retrieval.api">
2414 It should be easy to use the retrieval systems from applications. Refer
2416 <filename>yaz/retrieval.h</filename> and
2417 <filename>yaz/record_conv.h</filename>.
2423 <!-- Keep this comment at the end of the file
2428 sgml-minimize-attributes:nil
2429 sgml-always-quote-attributes:t
2432 sgml-parent-document: "yaz.xml"
2433 sgml-local-catalogs: nil
2434 sgml-namecase-general:t