Network Working Group                                       R. Denenberg
Request for Comments: 2056                           Library of Congress
Category: Standards Track                                       J. Kunze
                                 University of California, San Francisco
                                                                D. Lynch
                                          SilverPlatter Information Ltd.
                                                                 Editors
                                                           November 1996


                  Uniform Resource Locators for Z39.50


Status of this Memo

   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

1. Introduction

   Z39.50 is an information retrieval protocol that does not fit neatly
   into a retrieval model designed primarily around the stateless fetch
   of data.  Instead, it models a general user inquiry as a session-
   oriented, multi-step task, any step of which may be suspended
   temporarily while the server requests additional parameters from the
   client before continuing.  Some, none, or all of these client/server
   interactions may require participation of the client user, depending
   only on the client software (the protocol itself makes no such
   requirements).

   On the other hand, retrieval of "well-known" data may be performed in
   a single step, that is, with a degenerate Z39.50 session consisting
   of exactly one protocol search request and response.  Besides the
   basic search sub-service, there are several ancillary sub-services
   (e.g., Scan, Result Set Delete).  Among the functions covered by
   combinations of the sub-services, two core functions emerge as
   appropriately handled by two separate URL schemes:  the Session URL
   and the Retrieval URL.

   Using two schemes instead of one makes a critical distinction between
   a Z39.50 Session URL, which opens a client session initialized for
   interactive use by the user, and a Z39.50 Retrieval URL, which opens
   and closes a client session to retrieve a specific information item.
   Making this distinction at the scheme level allows the user interface
   to reflect it on to the user, without requiring the user interface to



Denenberg, et. al.          Standards Track                     [Page 1]


RFC 2056                    URLs for Z39.50                November 1996


   parse otherwise opaque parts of the URL (consistent with current
   practice).

2. Some Basic Concepts

   This section briefly describes the usage of Z39.50-specific
   terminology within the URL definitions below: specifically, the terms
   database, elementset, recordsyntax, and docid.

   The Z39.50 protocol specifies various information retrieval
   operations, the two most basic of which are Search and Present. In a
   Search operation a client provides search criteria and indicates a
   database (or several databases) on the server to search.  The
   essential result of a Search operation is that a result set is
   created at the server, consisting of pointers to the selected
   database records.

   Z39.50 models database records, abstract database records, and
   retrieval records.  A database record is a unit of information in a
   database, represented in a data structure local to the server.  An
   abstract database record is an abstract representation of that
   information, where the client and server share a common understanding
   of the representation.  This allows logical elements to be addressed
   and selected for transfer, via an element set specification, or, as
   used below, an "elementset".  A retrieval record is the set of
   selected elements packaged in an exportable structure, by the
   application of a "recordsyntax".

   Thus a Search operation results in entries pointing to database
   records; via a Present operation, a client requests a retrieval
   record, corresponding to a database record, corresponding to an entry
   in the result set. The client indicates the composition and format of
   the retrieval record by specifying an elementset and recordsyntax,
   respectively.

   A special case of a Z39.50 search is a "known-item" search, when a
   client intends that a search identify a single, known database
   record, or "document" (for purposes of illustration, assume that a
   database record corresponds to a document), and further, the client
   knows an identifier for the document that can be used to effect this
   known-item search.  In this case, this identifier is often referred
   to as a document identifier, or "docid".









Denenberg, et. al.          Standards Track                     [Page 2]


RFC 2056                    URLs for Z39.50                November 1996


3. The Z39.50 Session URL

   The Z39.50 Session URL may be informally described as providing the
   mechanism to switch the user to a Z39.50 client application.

   -  Host is required.
   -  Port is optional, and defaults to 210.
   -  All other parameters are optional.
   -  The Z39.50 client will start a session to the specified host/port
      (alternatively, it need not explicitly start a session, but may
      instead utilize an already open session to the same host/port).
   -  A database must be included if docid is included.
   -  If docid is included, the client will perform the specified search
      (in the same manner as for the retrieval URL, specified below).
   -  If docid is not included, and other parameters (besides host/port)
      are specified, the client may use those parameters as "hints".
      Various clients may choose to treat them as requirements, or as
      preferences, or ignore them.
   -  In any case (whether a search is performed or not), the client
      will leave the Z39.50 session open for the user, to do
      retrievals, new searches, etc.  (This is the main distinction
      from the Retrieval URL which leaves it up to the client whether
      or not to keep the Z39.50 session open.)

4. The Z39.50 Retrieval URL

   The Z39.50 Retrieval URL is intended to allow a Z39.50 session to be
   used as a transparent transfer mechanism to retrieve a specific
   information object.  A Z39.50 client uses information in the URL to
   formulate a Search Request.  The server's Search Response indicates
   how many records match the Request.  If the number of matching
   records does not equal one, the retrieval is considered unsuccessful,
   and the client application's behavior is not defined.  If the number
   of matching records equals one, the server may have included the
   desired record in the Search Response.  If not, the client requests
   transmission of the record with a Present Request.  After the client
   has received the specified record it may close the Z39.50 session
   immediately, or keep it open for subsequent retrievals.

   -  Host is required.
   -  Port is optional, and defaults to 210.
   -  A database is required.
   -  The meaning of a retrieval URL with no docid is undefined.
   -  The docid is placed into a type-1 query, as the single term, in
      the general format (tag 45), using the Bib-1 attribute set, with
      a Use attribute value of docid, and a structure attribute of URx.
      The docid string is server-defined and completely opaque to the
      client.



Denenberg, et. al.          Standards Track                     [Page 3]


RFC 2056                    URLs for Z39.50                November 1996


   - If element set name (esn) is not specified, it is the client's
      choice.  If esn is specified, it should be used either in the
      Search request for the value of small- and/or medium-
      set-element-set-names or in a Present request following a
      Search.  These terms and their use are defined within the Z39.50
      Standard [2].
   -  If record syntax (rs) is not specified, it is the client's choice.
      If one or more record syntaxes are specified, the client should
      select one (preferably the first in the list that it supports)
      and use it in a Search or Present request as the value of
      PreferredRecordSyntax.

5. BNF for Z39.50 URLs

   The Z39.50 Session and Retrieval URLs follow the Common Internet
   Scheme Syntax as defined in RFC 1738, "Uniform Resource Locators
   (URL)" [1].  In the definition, literals are quoted with "", optional
   elements are enclosed in [brackets], "|" is used to designate
   alternatives, and elements may be preceded with <n>* to designate n
   or more repetitions of the following element; n defaults to 0.

z39.50url      = zscheme "://" host [":" port]
                      ["/" [database *["+" database]
                             ["?" docid]]
                           [";esn=" elementset]
                           [";rs=" recordsyntax *[ "+" recordsyntax]]]

zscheme        = "z39.50r" | "z39.50s"
database       = uchar
docid          = uchar
elementset     = uchar
recordsyntax   = uchar

   Future extensions to these URLs will be of the form of
   [;keyword=value].

   The following definitions are from RFC 1738. Between the Internet
   Draft version and RFC 1738 two relevant changes were made: '=' was
   moved from the <extra> character class to <reserved>, and <national>
   was removed from the alternatives in <unreserved>. Neither <national>
   nor <punctuation> is referred to in this document nor in RFC 1738.










Denenberg, et. al.          Standards Track                     [Page 4]


RFC 2056                    URLs for Z39.50                November 1996


lowalpha       = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" |
                 "i" | "j" | "k" | "l" | "m" | "n" | "o" | "p" |
                 "q" | "r" | "s" | "t" | "u" | "v" | "w" | "x" |
                 "y" | "z"
hialpha        = "A" | "B" | "C" | "D" | "E" | "F" | "G" | "H" | "I" |
                 "J" | "K" | "L" | "M" | "N" | "O" | "P" | "Q" | "R" |
                 "S" | "T" | "U" | "V" | "W" | "X" | "Y" | "Z"

alpha          = lowalpha | hialpha
digit          = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" |
                 "8" | "9"
safe           = "$" | "-" | "_" | "." | "+"
extra          = "!" | "*" | "'" | "(" | ")" | ","
national       = "{" | "}" | "|" | "\" | "^" | "~" | "[" | "]" | "`"
punctuation    = "<" | ">" | "#" | "%" | <">

reserved       = ";" | "/" | "?" | ":" | "@" | "&" | "="
hex            = digit | "A" | "B" | "C" | "D" | "E" | "F" |
                 "a" | "b" | "c" | "d" | "e" | "f"
escape         = "%" hex hex
unreserved     = alpha | digit | safe | extra
uchar          = unreserved | escape
xchar          = unreserved | reserved | escape
digits         = 1*digit

6. Security Considerations

   The two Z39.50 URL schemes are subject to the same security
   implications as the general URL scheme [1], so the usual precautions
   apply.  This means, for example, that a locator might no longer point
   to the object that was originally intended.  It also means that it
   may be possible to construct a URL so that an attempt to perform a
   harmless idempotent operation such as the retrieval of an object will
   in fact cause a possibly damaging remote operation to occur.

7. Acknowledgements

   The Z39.50 Implementors Group contributed the substance of this
   document.

8. References

   [1] Berners-Lee, T., Masinter, L., McCahill, M. (editors), "Uniform
       Resource Locators (URL)", RFC 1738, December 1994.
       ftp://ds.internic.net/rfc/rfc1738.txt

   [2] ANSI/NISO Z39.50-1995, "ANSI Z39.50: Information Retrieval
       Service and Protocol", 1995.  ftp://ftp.loc.gov/pub/z3950/



Denenberg, et. al.          Standards Track                     [Page 5]


RFC 2056                    URLs for Z39.50                November 1996


   [3] ANSI/NISO Z39.50-1992, "ANSI Z39.50: Information Retrieval
       Service and Protocol", 1992.
       ftp://ftp.cni.org/pub/NISO/docs/Z39.50-1992/www/Z39.50.toc.html
       (also available in hard copy from Omnicom Information Service,
       115 Park St., SE, Vienna, VA  22180).

9. Editors' Addresses

   Ray Denenberg
   Library of Congress
   Collections Services
   Network Development/MSO
   Washington DC 20540


   Phone: (202) 707-5795
   Fax:   (202) 707-0115
   EMail: ray@rden.loc.gov


   John A. Kunze
   Center for Knowledge Management
   University of California, San Francisco
   530 Parnassus Ave, Box 0840
   San Francisco, CA  94143-0840

   Phone: (415) 502-6660
   Fax:   (415) 476-4653
   EMail: jak@ckm.ucsf.edu


   Denis Lynch
   SilverPlatter Information Ltd.
   10 Barely Mow Passage
   Chiswick, London W4 4PH
   U.K.

   Voice: +44 (0)181-995-8242
   Fax:   +44 (0)181-995-5159
   EMail: DenisL@SilverPlatter.com











Denenberg, et. al.          Standards Track                     [Page 6]


RFC 2056                    URLs for Z39.50                November 1996


Appendix. Examples of Z39.50 URLs

   A basic Z39.50 session URL that a client might use to open a
   connection to the MELVYL union catalog "cat" at the University of
   California is

        z39.50s://melvyl.ucop.edu/cat

   A URL that would open the MELVYL magazine database just long enough
   to fetch an article from volume 30, number 19 of a hypothetical
   periodical might look like

        z39.50r://melvyl.ucop.edu/mags?elecworld.v30.n19

   As a final example, here is another retrieval URL that a client could
   use to request a full record (element set "f") in the MARC syntax
   from a hypothetical database called TMF at CNIDR:

        z39.50r://cnidr.org:2100/tmf?bkirch_rules__a1;esn=f;rs=marc

   As in the previous example, the part of the string after the `?' is
   determined by the server.  In this example, the server is running on
   non-standard port 2100.




























Denenberg, et. al.          Standards Track                     [Page 7]