Z39.50 server profile

Users may decide whether or not to download analytic records found in some monographs in the catalogue général de la BnF.

In order not to interfere with login parameters set up at a time when such records could not be downloaded, the default option is for analytic records not to be downloaded: if you do not wish to download analytic records, there is nothing to specify when logging in.

If you wish to download analytic records, you need to explicitly request this when logging in, by adding the string “-ANA1” to the value of the “Database name” parameter, giving TOUT-ANA1.

Character encoding for responses to Z39.50 queries may be any of the following, at the user’s choice:

• ISO 5426-compliant: “Extension of the Latin alphabet coded character set for bibliographic information interchange”. This is the default standard – there is nothing to specify when logging in.
• ISO 8859-1-compliant (better known as LATIN1): “8-bit single-byte coded graphic character sets – Part 1: Latin alphabet”. To select this option, you need to add the string “-LATIN1” to the value of the “Database name” parameter when you log in, giving TOUT-LATIN1.
• ISO/CEI 10646-compliant (better known as UNICODE): “Universal multiple-octet coded character set”. The available option is for 8-bit coding, i.e. UTF8. To select this option, you need to add the string “-UTF8” to the value of the “Database name” parameter when you log in, giving TOUT-UTF8.

Example YAZ client login scripts:

• Records returned in UNIMARC format, ISO5426, without analytic records auth Z3950 Z3950_BNF
  elements F format 1.2.840.10003.5.1
  open tcp:z3950.bnf.fr:2211/TOUT
• Records returned in UNIMARC format, with analytic records and UTF8 character encoding auth Z3950 Z3950_BNF
  elements F format 1.2.840.10003.5.1
  open tcp:z3950.bnf.fr:2211/TOUT-ANA1-UTF8
• Records returned in INTERMARC format, LATIN1, without analytic records auth Z3950 Z3950_BNF
  elements F format 1.2.840.10003.5.2
  open tcp:z3950.bnf.fr:2211/TOUT-LATIN1