1 <!-- $Id: odr.xml,v 1.13 2004-08-11 12:47:35 adam Exp $ -->
2 <chapter id="odr"><title>The ODR Module</title>
4 <sect1 id="odr.introduction"><title>Introduction</title>
7 &odr; is the BER-encoding/decoding subsystem of &yaz;. Care as been taken
8 to isolate &odr; from the rest of the package - specifically from the
9 transport interface. &odr; may be used in any context where basic
10 ASN.1/BER representations are used.
14 If you are only interested in writing a Z39.50 implementation based on
15 the PDUs that are already provided with &yaz;, you only need to concern
16 yourself with the section on managing ODR streams
17 (<xref linkend="odr.use"/>). Only if you need to
18 implement ASN.1 beyond that which has been provided, should you
19 worry about the second half of the documentation
20 (<xref linkend="odr.programming"/>).
21 If you use one of the higher-level interfaces, you can skip this
26 This is important, so we'll repeat it for emphasis: <emphasis>You do
27 not need to read <xref linkend="odr.programming"/>
28 to implement Z39.50 with &yaz;.</emphasis>
32 If you need a part of the protocol that isn't already in &yaz;, you
33 should contact the authors before going to work on it yourself: We
34 might already be working on it. Conversely, if you implement a useful
35 part of the protocol before us, we'd be happy to include it in a
40 <sect1 id="odr.use"><title>Using ODR</title>
42 <sect2><title>ODR Streams</title>
45 Conceptually, the ODR stream is the source of encoded data in the
46 decoding mode; when encoding, it is the receptacle for the encoded
47 data. Before you can use an ODR stream it must be allocated. This is
48 done with the function
52 ODR odr_createmem(int direction);
56 The <function>odr_createmem()</function> function takes as argument one
57 of three manifest constants: <literal>ODR_ENCODE</literal>,
58 <literal>ODR_DECODE</literal>, or <literal>ODR_PRINT</literal>.
59 An &odr; stream can be in only one mode - it is not possible to change
60 its mode once it's selected. Typically, your program will allocate
61 at least two ODR streams - one for decoding, and one for encoding.
65 When you're done with the stream, you can use
69 void odr_destroy(ODR o);
73 to release the resources allocated for the stream.
77 <sect2><title id="memory">Memory Management</title>
80 Two forms of memory management take place in the &odr; system. The first
81 one, which has to do with allocating little bits of memory (sometimes
82 quite large bits of memory, actually) when a protocol package is
83 decoded, and turned into a complex of interlinked structures. This
84 section deals with this system, and how you can use it for your own
85 purposes. The next section deals with the memory management which is
86 required when encoding data - to make sure that a large enough buffer is
87 available to hold the fully encoded PDU.
91 The &odr; module has its own memory management system, which is
92 used whenever memory is required. Specifically, it is used to allocate
93 space for data when decoding incoming PDUs. You can use the memory
94 system for your own purposes, by using the function
98 void *odr_malloc(ODR o, int size);
102 You can't use the normal <function>free(2)</function> routine to free
103 memory allocated by this function, and &odr; doesn't provide a parallel
104 function. Instead, you can call
108 void odr_reset(ODR o, int size);
112 when you are done with the
113 memory: Everything allocated since the last call to
114 <function>odr_reset()</function> is released.
115 The <function>odr_reset()</function> call is also required to clear
116 up an error condition on a stream.
124 int odr_total(ODR o);
128 returns the number of bytes allocated on the stream since the last call to
129 <function>odr_reset()</function>.
133 The memory subsystem of &odr; is fairly efficient at allocating and
134 releasing little bits of memory. Rather than managing the individual,
135 small bits of space, the system maintains a free-list of larger chunks
136 of memory, which are handed out in small bits. This scheme is
137 generally known as a <emphasis>nibble memory</emphasis> system.
138 It is very useful for maintaining short-lived constructions such
143 If you want to retain a bit of memory beyond the next call to
144 <function>odr_reset()</function>, you can use the function
148 ODR_MEM odr_extract_mem(ODR o);
152 This function will give you control of the memory recently allocated
153 on the ODR stream. The memory will live (past calls to
154 <function>odr_reset()</function>), until you call the function
158 void odr_release_mem(ODR_MEM p);
162 The opaque <literal>ODR_MEM</literal> handle has no other purpose than
163 referencing the memory block for you until you want to release it.
167 You can use <function>odr_extract_mem()</function> repeatedly between
168 allocating data, to retain individual control of separate chunks of data.
172 <sect2><title>Encoding and Decoding Data</title>
175 When encoding data, the ODR stream will write the encoded octet string
176 in an internal buffer. To retrieve the data, use the function
180 char *odr_getbuf(ODR o, int *len, int *size);
184 The integer pointed to by len is set to the length of the encoded
185 data, and a pointer to that data is returned. <literal>*size</literal>
186 is set to the size of the buffer (unless <literal>size</literal> is null,
187 signaling that you are not interested in the size). The next call to
188 a primitive function using the same &odr; stream will overwrite the
189 data, unless a different buffer has been supplied using the call
193 void odr_setbuf(ODR o, char *buf, int len, int can_grow);
197 which sets the encoding (or decoding) buffer used by
198 <literal>o</literal> to <literal>buf</literal>, using the length
199 <literal>len</literal>.
200 Before a call to an encoding function, you can use
201 <function>odr_setbuf()</function> to provide the stream with an encoding
202 buffer of sufficient size (length). The <literal>can_grow</literal>
203 parameter tells the encoding &odr; stream whether it is allowed to use
204 <function>realloc(2)</function> to increase the size of the buffer when
205 necessary. The default condition of a new encoding stream is equivalent
206 to the results of calling
210 odr_setbuf(stream, 0, 0, 1);
214 In this case, the stream will allocate and reallocate memory as
215 necessary. The stream reallocates memory by repeatedly doubling the
216 size of the buffer - the result is that the buffer will typically
217 reach its maximum, working size with only a small number of reallocation
218 operations. The memory is freed by the stream when the latter is destroyed,
219 unless it was assigned by the user with the <literal>can_grow</literal>
220 parameter set to zero (in this case, you are expected to retain
221 control of the memory yourself).
225 To assume full control of an encoded buffer, you must first call
226 <function>odr_getbuf()</function> to fetch the buffer and its length.
227 Next, you should call <function>odr_setbuf()</function> to provide a
228 different buffer (or a null pointer) to the stream. In the simplest
229 case, you will reuse the same buffer over and over again, and you
230 will just need to call <function>odr_getbuf()</function> after each
231 encoding operation to get the length and address of the buffer.
232 Note that the stream may reallocate the buffer during an encoding
233 operation, so it is necessary to retrieve the correct address after
234 each encoding operation.
238 It is important to realize that the ODR stream will not release this
239 memory when you call <function>odr_reset()</function>: It will
240 merely update its internal pointers to prepare for the encoding of a
242 When the stream is released by the <function>odr_destroy()</function>
243 function, the memory given to it by <function>odr_setbuf</function> will
244 be released <emphasis>only</emphasis> if the <literal>can_grow</literal>
245 parameter to <function>odr_setbuf()</function> was nonzero. The
246 <literal>can_grow</literal> parameter, in other words, is a way of
247 signaling who is to own the buffer, you or the ODR stream. If you never call
248 <function>odr_setbuf()</function> on your encoding stream, which is
249 typically the case, the buffer allocated by the stream will belong to
250 the stream by default.
254 When you wish to decode data, you should first call
255 <function>odr_setbuf()</function>, to tell the decoding stream
256 where to find the encoded data, and how long the buffer is
257 (the <literal>can_grow</literal> parameter is ignored by a decoding
258 stream). After this, you can call the function corresponding to the
259 data you wish to decode (eg, <function>odr_integer()</function> odr
260 <function>z_APDU()</function>).
263 <example><title>Encoding and decoding functions</title>
265 int odr_integer(ODR o, int **p, int optional, const char *name);
267 int z_APDU(ODR o, Z_APDU **p, int optional, const char *name);
272 If the data is absent (or doesn't match the tag corresponding to
273 the type), the return value will be either 0 or 1 depending on the
274 <literal>optional</literal> flag. If <literal>optional</literal>
275 is 0 and the data is absent, an error flag will be raised in the
276 stream, and you'll need to call <function>odr_reset()</function> before
277 you can use the stream again. If <literal>optional</literal> is
278 nonzero, the pointer <emphasis>pointed</emphasis> to/ by
279 <literal>p</literal> will be set to the null value, and the function
281 The <literal>name</literal> argument is used to pretty-print the
282 tag in question. It may be set to <literal>NULL</literal> if
283 pretty-printing is not desired.
287 If the data value is found where it's expected, the pointer
288 <emphasis>pointed to</emphasis> by the <literal>p</literal> argument
289 will be set to point to the decoded type.
290 The space for the type will be allocated and owned by the &odr;
291 stream, and it will live until you call
292 <function>odr_reset()</function> on the stream. You cannot use
293 <function>free(2)</function> to release the memory.
294 You can decode several data elements (by repeated calls to
295 <function>odr_setbuf()</function> and your decoding function), and
296 new memory will be allocated each time. When you do call
297 <function>odr_reset()</function>, everything decoded since the
298 last call to <function>odr_reset()</function> will be released.
301 <example><title>Encoding and decoding of an integer</title>
303 The use of the double indirection can be a little confusing at first
304 (its purpose will become clear later on, hopefully),
305 so an example is in order. We'll encode an integer value, and
306 immediately decode it again using a different stream. A useless, but
307 informative operation.
309 <programlisting><![CDATA[
310 void do_nothing_useful(int value)
317 /* allocate streams */
318 if (!(encode = odr_createmem(ODR_ENCODE)))
320 if (!(decode = odr_createmem(ODR_DECODE)))
324 if (odr_integer(encode, &valp, 0, 0) == 0)
326 printf("encoding went bad\n");
329 bufferp = odr_getbuf(encode, &len);
330 printf("length of encoded data is %d\n", len);
332 /* now let's decode the thing again */
333 odr_setbuf(decode, bufferp, len);
334 if (odr_integer(decode, &resvalp, 0, 0) == 0)
336 printf("decoding went bad\n");
339 printf("the value is %d\n", *resvalp);
348 This looks like a lot of work, offhand. In practice, the &odr; streams
349 will typically be allocated once, in the beginning of your program
350 (or at the beginning of a new network session), and the encoding
351 and decoding will only take place in a few, isolated places in your
352 program, so the overhead is quite manageable.
358 <sect2><title>Printing</title>
360 When an ODR stream is created of type <literal>ODR_PRINT</literal>
361 the ODR module will print the contents of a PDU in a readable format.
362 By default output is written to the <literal>stderr</literal> stream.
363 This behavior can be changed, however, by calling the function
365 odr_setprint(ODR o, FILE *file);
367 before encoders or decoders are being invoked.
368 It is also possible to direct the output to a buffer (of indeed
369 another file), by using the more generic mechanism:
371 void odr_set_stream(ODR o, void *handle,
372 void (*stream_puts)(void *handle, const char *strz),
373 void (*stream_close)(void *handle));
375 Here the user provides an opaque handle and two handlers,
376 <replaceable>stream_puts</replaceable> for printing,
377 and <replaceable>stream_close</replaceable> which is supposed
378 to close/free resources associated with handle.
379 The <replaceable>stream_close</replaceable> handler is optional and
380 if NULL for the function is provided, it will not be invoked.
383 <sect2><title>Diagnostics</title>
386 The encoding/decoding functions all return 0 when an error occurs.
387 Until you call <function>odr_reset()</function>, you cannot use the
388 stream again, and any function called will immediately return 0.
392 To provide information to the programmer or administrator, the function
396 void odr_perror(ODR o, char *message);
400 is provided, which prints the <literal>message</literal> argument to
401 <literal>stderr</literal> along with an error message from the stream.
405 You can also use the function
409 int odr_geterror(ODR o);
413 to get the current error number from the screen. The number will be
414 one of these constants:
417 <table frame="top"><title>ODR Error codes</title>
422 <entry>Description</entry>
427 <entry>OMEMORY</entry><entry>Memory allocation failed.</entry>
431 <entry>OSYSERR</entry><entry>A system- or library call has failed.
432 The standard diagnostic variable <literal>errno</literal> should be
433 examined to determine the actual error.</entry>
437 <entry>OSPACE</entry><entry>No more space for encoding.
438 This will only occur when the user has explicitly provided a
439 buffer for an encoding stream without allowing the system to
440 allocate more space.</entry>
444 <entry>OREQUIRED</entry><entry>This is a common protocol error; A
445 required data element was missing during encoding or decoding.</entry>
449 <entry>OUNEXPECTED</entry><entry>An unexpected data element was
450 found during decoding.</entry>
453 <row><entry>OOTHER</entry><entry>Other error. This is typically an
454 indication of misuse of the &odr; system by the programmer, and also
455 that the diagnostic system isn't as good as it should be, yet.</entry>
462 The character string array
466 char *odr_errlist[]
470 can be indexed by the error code to obtain a human-readable
471 representation of the problem.
475 <sect2><title>Summary and Synopsis</title>
480 ODR odr_createmem(int direction);
482 void odr_destroy(ODR o);
484 void odr_reset(ODR o);
486 char *odr_getbuf(ODR o, int *len);
488 void odr_setbuf(ODR o, char *buf, int len);
490 void *odr_malloc(ODR o, int size);
492 ODR_MEM odr_extract_mem(ODR o);
494 void odr_release_mem(ODR_MEM r);
496 int odr_geterror(ODR o);
498 void odr_perror(char *message);
500 extern char *odr_errlist[];
506 <sect1 id="odr.programming"><title>Programming with ODR</title>
509 The API of &odr; is designed to reflect the structure of ASN.1, rather
510 than BER itself. Future releases may be able to represent data in
511 other external forms.
516 There is an ASN.1 tutorial available at
517 <ulink url="http://asn1.elibel.tm.fr/en/introduction/">this site</ulink>.
518 This site also has standards for ASN.1 (X.680) and BER (X.690)
519 <ulink url="http://asn1.elibel.tm.fr/en/standards/">online</ulink>.
524 The ODR interface is based loosely on that of the Sun Microsystems
526 Specifically, each function which corresponds to an ASN.1 primitive
527 type has a dual function. Depending on the settings of the ODR
528 stream which is supplied as a parameter, the function may be used
529 either to encode or decode data. The functions that can be built
530 using these primitive functions, to represent more complex data types,
531 share this quality. The result is that you only have to enter the
532 definition for a type once - and you have the functionality of encoding,
533 decoding (and pretty-printing) all in one unit.
534 The resulting C source code is quite compact, and is a pretty
535 straightforward representation of the source ASN.1 specification.
539 In many cases, the model of the XDR functions works quite well in this
541 In others, it is less elegant. Most of the hassle comes from the optional
542 SEQUENCE members which don't exist in XDR.
545 <sect2><title>The Primitive ASN.1 Types</title>
548 ASN.1 defines a number of primitive types (many of which correspond
549 roughly to primitive types in structured programming languages, such as C).
552 <sect3><title>INTEGER</title>
555 The &odr; function for encoding or decoding (or printing) the ASN.1
556 INTEGER type looks like this:
560 int odr_integer(ODR o, int **p, int optional, const char *name);
564 (we don't allow values that can't be contained in a C integer.)
568 This form is typical of the primitive &odr; functions. They are named
569 after the type of data that they encode or decode. They take an &odr;
570 stream, an indirect reference to the type in question, and an
571 <literal>optional</literal> flag (corresponding to the OPTIONAL keyword
572 of ASN.1) as parameters. They all return an integer value of either one
574 When you use the primitive functions to construct encoders for complex
575 types of your own, you should follow this model as well. This
576 ensures that your new types can be reused as elements in yet more
581 The <literal>o</literal> parameter should obviously refer to a properly
582 initialized &odr; stream of the right type (encoding/decoding/printing)
583 for the operation that you wish to perform.
587 When encoding or printing, the function first looks at
588 <literal>* p</literal>. If <literal>* p</literal> (the pointer pointed
589 to by <literal>p</literal>) is a null pointer, this is taken to mean that
590 the data element is absent. If the <literal>optional</literal> parameter
591 is nonzero, the function will return one (signifying success) without
592 any further processing. If the <literal>optional</literal> is zero, an
593 internal error flag is set in the &odr; stream, and the function will
594 return 0. No further operations can be carried out on the stream without
595 a call to the function <function>odr_reset()</function>.
599 If <literal>*p</literal> is not a null pointer, it is expected to
600 point to an instance of the data type. The data will be subjected to
601 the encoding rules, and the result will be placed in the buffer held
606 The other ASN.1 primitives have similar functions that operate in
610 <sect3><title>BOOLEAN</title>
613 int odr_bool(ODR o, bool_t **p, int optional, const char *name);
617 <sect3><title>REAL</title>
624 <sect3><title>NULL</title>
627 int odr_null(ODR o, bool_t **p, int optional, const char *name);
631 In this case, the value of **p is not important. If <literal>*p</literal>
632 is different from the null pointer, the null value is present, otherwise
637 <sect3><title>OCTET STRING</title>
640 typedef struct odr_oct
647 int odr_octetstring(ODR o, Odr_oct **p, int optional,
652 The <literal>buf</literal> field should point to the character array
653 that holds the octetstring. The <literal>len</literal> field holds the
654 actual length, while the <literal>size</literal> field gives the size
655 of the allocated array (not of interest to you, in most cases).
656 The character array need not be null terminated.
660 To make things a little easier, an alternative is given for string
661 types that are not expected to contain embedded NULL characters (eg.
666 int odr_cstring(ODR o, char **p, int optional, const char *name);
670 Which encoded or decodes between OCTETSTRING representations and
671 null-terminates C strings.
675 Functions are provided for the derived string types, eg:
679 int odr_visiblestring(ODR o, char **p, int optional,
684 <sect3><title>BIT STRING</title>
687 int odr_bitstring(ODR o, Odr_bitmask **p, int optional,
692 The opaque type <literal>Odr_bitmask</literal> is only suitable for
693 holding relatively brief bit strings, eg. for options fields, etc.
694 The constant <literal>ODR_BITMASK_SIZE</literal> multiplied by 8
695 gives the maximum possible number of bits.
699 A set of macros are provided for manipulating the
700 <literal>Odr_bitmask</literal> type:
704 void ODR_MASK_ZERO(Odr_bitmask *b);
706 void ODR_MASK_SET(Odr_bitmask *b, int bitno);
708 void ODR_MASK_CLEAR(Odr_bitmask *b, int bitno);
710 int ODR_MASK_GET(Odr_bitmask *b, int bitno);
714 The functions are modeled after the manipulation functions that
715 accompany the <literal>fd_set</literal> type used by the
716 <function>select(2)</function> call.
717 <literal>ODR_MASK_ZERO</literal> should always be called first on a
718 new bitmask, to initialize the bits to zero.
722 <sect3><title>OBJECT IDENTIFIER</title>
725 int odr_oid(ODR o, Odr_oid **p, int optional, const char *name);
729 The C OID representation is simply an array of integers, terminated by
730 the value -1 (the <literal>Odr_oid</literal> type is synonymous with
731 the <literal>int</literal> type).
732 We suggest that you use the OID database module (see
733 <xref linkend="asn.oid"/>) to handle object identifiers
739 <sect2 id="tag.prim"><title>Tagging Primitive Types</title>
742 The simplest way of tagging a type is to use the
743 <function>odr_implicit_tag()</function> or
744 <function>odr_explicit_tag()</function> macros:
748 int odr_implicit_tag(ODR o, Odr_fun fun, int class, int tag,
749 int optional, const char *name);
751 int odr_explicit_tag(ODR o, Odr_fun fun, int class, int tag,
752 int optional, const char *name);
756 To create a type derived from the integer type by implicit tagging, you
761 MyInt ::= [210] IMPLICIT INTEGER
765 In the &odr; system, this would be written like:
769 int myInt(ODR o, int **p, int optional, const char *name)
771 return odr_implicit_tag(o, odr_integer, p,
772 ODR_CONTEXT, 210, optional, name);
777 The function <function>myInt()</function> can then be used like any of
778 the primitive functions provided by &odr;. Note that the behavior of
779 <function>odr_explicit_tag()</function>
780 and <function>odr_implicit_tag()</function> macros
781 act exactly the same as the functions they are applied to - they
782 respond to error conditions, etc, in the same manner - they
783 simply have three extra parameters. The class parameter may
784 take one of the values: <literal>ODR_CONTEXT</literal>,
785 <literal>ODR_PRIVATE</literal>, <literal>ODR_UNIVERSAL</literal>, or
786 <literal>/ODR_APPLICATION</literal>.
790 <sect2><title>Constructed Types</title>
793 Constructed types are created by combining primitive types. The
794 &odr; system only implements the SEQUENCE and SEQUENCE OF constructions
795 (although adding the rest of the container types should be simple
796 enough, if the need arises).
800 For implementing SEQUENCEs, the functions
804 int odr_sequence_begin(ODR o, void *p, int size, const char *name);
805 int odr_sequence_end(ODR o);
813 The <function>odr_sequence_begin()</function> function should be
814 called in the beginning of a function that implements a SEQUENCE type.
815 Its parameters are the &odr; stream, a pointer (to a pointer to the type
816 you're implementing), and the <literal>size</literal> of the type
817 (typically a C structure). On encoding, it returns 1 if
818 <literal>* p</literal> is a null pointer. The <literal>size</literal>
819 parameter is ignored. On decoding, it returns 1 if the type is found in
820 the data stream. <literal>size</literal> bytes of memory are allocated,
821 and <literal>*p</literal> is set to point to this space.
822 <function>odr_sequence_end()</function> is called at the end of the
823 complex function. Assume that a type is defined like this:
827 MySequence ::= SEQUENCE {
829 boolval BOOLEAN OPTIONAL
834 The corresponding &odr; encoder/decoder function and the associated data
835 structures could be written like this:
839 typedef struct MySequence
845 int mySequence(ODR o, MySequence **p, int optional, const char *name)
847 if (odr_sequence_begin(o, p, sizeof(**p), name) == 0)
848 return optional && odr_ok(o);
850 odr_integer(o, &(*p)->intval, 0, "intval") &&
851 odr_bool(o, &(*p)->boolval, 1, "boolval") &&
858 Note the 1 in the call to <function>odr_bool()</function>, to mark
859 that the sequence member is optional.
860 If either of the member types had been tagged, the macros
861 <function>odr_implicit_tag()</function> or
862 <function>odr_explicit_tag()</function>
863 could have been used.
864 The new function can be used exactly like the standard functions provided
865 with &odr;. It will encode, decode or pretty-print a data value of the
866 <literal>MySequence</literal> type. We like to name types with an
867 initial capital, as done in ASN.1 definitions, and to name the
868 corresponding function with the first character of the name in lower case.
869 You could, of course, name your structures, types, and functions any way
870 you please - as long as you're consistent, and your code is easily readable.
871 <literal>odr_ok</literal> is just that - a predicate that returns the
872 state of the stream. It is used to ensure that the behavior of the new
873 type is compatible with the interface of the primitive types.
877 <sect2><title>Tagging Constructed Types</title>
881 See <xref linkend="tag.prim"/> for information on how to tag
882 the primitive types, as well as types that are already defined.
886 <sect3><title>Implicit Tagging</title>
889 Assume the type above had been defined as
893 MySequence ::= [10] IMPLICIT SEQUENCE {
895 boolval BOOLEAN OPTIONAL
900 You would implement this in &odr; by calling the function
904 int odr_implicit_settag(ODR o, int class, int tag);
908 which overrides the tag of the type immediately following it. The
909 macro <function>odr_implicit_tag()</function> works by calling
910 <function>odr_implicit_settag()</function> immediately
911 before calling the function pointer argument.
912 Your type function could look like this:
916 int mySequence(ODR o, MySequence **p, int optional, const char *name)
918 if (odr_implicit_settag(o, ODR_CONTEXT, 10) == 0 ||
919 odr_sequence_begin(o, p, sizeof(**p), name) == 0)
920 return optional && odr_ok(o);
922 odr_integer(o, &(*p)->intval, 0, "intval") &&
923 odr_bool(o, &(*p)->boolval, 1, "boolval") &&
929 The definition of the structure <literal>MySequence</literal> would be
934 <sect3><title>Explicit Tagging</title>
937 Explicit tagging of constructed types is a little more complicated,
938 since you are in effect adding a level of construction to the data.
942 Assume the definition:
946 MySequence ::= [10] IMPLICIT SEQUENCE {
948 boolval BOOLEAN OPTIONAL
953 Since the new type has an extra level of construction, two new functions
954 are needed to encapsulate the base type:
958 int odr_constructed_begin(ODR o, void *p, int class, int tag,
961 int odr_constructed_end(ODR o);
965 Assume that the IMPLICIT in the type definition above were replaced
966 with EXPLICIT (or that the IMPLICIT keyword were simply deleted, which
967 would be equivalent). The structure definition would look the same,
968 but the function would look like this:
972 int mySequence(ODR o, MySequence **p, int optional, const char *name)
974 if (odr_constructed_begin(o, p, ODR_CONTEXT, 10, name) == 0)
975 return optional && odr_ok(o);
976 if (o->direction == ODR_DECODE)
977 *p = odr_malloc(o, sizeof(**p));
978 if (odr_sequence_begin(o, p, sizeof(**p), 0) == 0)
980 *p = 0; /* this is almost certainly a protocol error */
984 odr_integer(o, &(*p)->intval, 0, "intval") &&
985 odr_bool(o, &(*p)->boolval, 1, "boolval") &&
986 odr_sequence_end(o) &&
987 odr_constructed_end(o);
992 Notice that the interface here gets kind of nasty. The reason is
993 simple: Explicitly tagged, constructed types are fairly rare in
994 the protocols that we care about, so the
995 esthetic annoyance (not to mention the dangers of a cluttered
996 interface) is less than the time that would be required to develop a
997 better interface. Nevertheless, it is far from satisfying, and it's a
998 point that will be worked on in the future. One option for you would
999 be to simply apply the <function>odr_explicit_tag()</function> macro to
1000 the first function, and not
1001 have to worry about <function>odr_constructed_*</function> yourself.
1002 Incidentally, as you might have guessed, the
1003 <function>odr_sequence_</function> functions are themselves
1004 implemented using the <function>/odr_constructed_</function> functions.
1009 <sect2><title>SEQUENCE OF</title>
1012 To handle sequences (arrays) of a specific type, the function
1016 int odr_sequence_of(ODR o, int (*fun)(ODR o, void *p, int optional),
1017 void *p, int *num, const char *name);
1021 The <literal>fun</literal> parameter is a pointer to the decoder/encoder
1022 function of the type. <literal>p</literal> is a pointer to an array of
1023 pointers to your type. <literal>num</literal> is the number of elements
1032 MyArray ::= SEQUENCE OF INTEGER
1036 The C representation might be
1040 typedef struct MyArray
1048 And the function might look like
1052 int myArray(ODR o, MyArray **p, int optional, const char *name)
1054 if (o->direction == ODR_DECODE)
1055 *p = odr_malloc(o, sizeof(**p));
1056 if (odr_sequence_of(o, odr_integer, &(*p)->elements,
1057 &(*p)->num_elements, name))
1060 return optional && odr_ok(o);
1065 <sect2><title>CHOICE Types</title>
1068 The choice type is used fairly often in some ASN.1 definitions, so
1069 some work has gone into streamlining its interface.
1073 CHOICE types are handled by the function:
1077 int odr_choice(ODR o, Odr_arm arm[], void *p, void *whichp,
1082 The <literal>arm</literal> array is used to describe each of the possible
1083 types that the CHOICE type may assume. Internally in your application,
1084 the CHOICE type is represented as a discriminated union. That is, a
1085 C union accompanied by an integer (or enum) identifying the active
1087 <literal>whichp</literal> is a pointer to the union discriminator.
1088 When encoding, it is examined to determine the current type.
1089 When decoding, it is set to reference the type that was found in
1094 The Odr_arm type is defined thus:
1098 typedef struct odr_arm
1110 The interpretation of the fields are:
1114 <varlistentry><term>tagmode</term>
1115 <listitem><para>Either <literal>ODR_IMPLICIT</literal>,
1116 <literal>ODR_EXPLICIT</literal>, or <literal>ODR_NONE</literal> (-1)
1117 to mark no tagging.</para></listitem>
1120 <varlistentry><term>which</term>
1121 <listitem><para>The value of the discriminator that corresponds to
1122 this CHOICE element. Typically, it will be a #defined constant, or
1123 an enum member.</para></listitem>
1126 <varlistentry><term>fun</term>
1127 <listitem><para>A pointer to a function that implements the type of
1128 the CHOICE member. It may be either a standard &odr; type or a type
1129 defined by yourself.</para></listitem>
1132 <varlistentry><term>name</term>
1133 <listitem><para>Name of tag.</para></listitem>
1138 A handy way to prepare the array for use by the
1139 <function>odr_choice()</function> function is to
1140 define it as a static, initialized array in the beginning of your
1141 decoding/encoding function. Assume the type definition:
1145 MyChoice ::= CHOICE {
1147 tagged [99] IMPLICIT INTEGER,
1153 Your C type might look like
1157 typedef struct MyChoice
1175 And your function could look like this:
1179 int myChoice(ODR o, MyChoice **p, int optional, const char *name)
1181 static Odr_arm arm[] =
1183 {-1, -1, -1, MyChoice_untagged, odr_integer, "untagged"},
1184 {ODR_IMPLICIT, ODR_CONTEXT, 99, MyChoice_tagged, odr_integer,
1186 {-1, -1, -1, MyChoice_other, odr_boolean, "other"},
1190 if (o->direction == ODR_DECODE)
1191 *p = odr_malloc(o, sizeof(**p);
1193 return optional && odr_ok(o);
1195 if (odr_choice(o, arm, &(*p)->u, &(*p)->which), name)
1198 return optional && odr_ok(o);
1203 In some cases (say, a non-optional choice which is a member of a
1204 sequence), you can "embed" the union and its discriminator in the
1205 structure belonging to the enclosing type, and you won't need to
1206 fiddle with memory allocation to create a separate structure to
1207 wrap the discriminator and union.
1211 The corresponding function is somewhat nicer in the Sun XDR interface.
1212 Most of the complexity of this interface comes from the possibility of
1213 declaring sequence elements (including CHOICEs) optional.
1217 The ASN.1 specifications naturally requires that each member of a
1218 CHOICE have a distinct tag, so they can be told apart on decoding.
1219 Sometimes it can be useful to define a CHOICE that has multiple types
1220 that share the same tag. You'll need some other mechanism, perhaps
1221 keyed to the context of the CHOICE type. In effect, we would like to
1222 introduce a level of context-sensitiveness to our ASN.1 specification.
1223 When encoding an internal representation, we have no problem, as long
1224 as each CHOICE member has a distinct discriminator value. For
1225 decoding, we need a way to tell the choice function to look for a
1226 specific arm of the table. The function
1230 void odr_choice_bias(ODR o, int what);
1234 provides this functionality. When called, it leaves a notice for the next
1235 call to <function>odr_choice()</function> to be called on the decoding
1236 stream <literal>o</literal> that only the <literal>arm</literal> entry with
1237 a <literal>which</literal> field equal to <literal>what</literal>
1242 The most important application (perhaps the only one, really) is in
1243 the definition of application-specific EXTERNAL encoders/decoders
1244 which will automatically decode an ANY member given the direct or
1251 <sect1 id="odr.debugging"><title>Debugging</title>
1254 The protocol modules are suffering somewhat from a lack of diagnostic
1255 tools at the moment. Specifically ways to pretty-print PDUs that
1256 aren't recognized by the system. We'll include something to this end
1257 in a not-too-distant release. In the meantime, what we do when we get
1258 packages we don't understand is to compile the ODR module with
1259 <literal>ODR_DEBUG</literal> defined. This causes the module to dump tracing
1260 information as it processes data units. With this output and the
1261 protocol specification (Z39.50), it is generally fairly easy to see
1266 <!-- Keep this comment at the end of the file
1271 sgml-minimize-attributes:nil
1272 sgml-always-quote-attributes:t
1275 sgml-parent-document: "yaz.xml"
1276 sgml-local-catalogs: nil
1277 sgml-namecase-general:t