UNIX htable and slashes

Wed 16 Apr 86 14:20:33-CST


Regarding your message (brought to my attention by <ai.clive@mcc.arpa>):

        Return-Path: <tcp-ip-RELAY@SRI-NIC.ARPA>
        Received: from SRI-NIC.ARPA by MCC.ARPA with TCP; Wed 16 Apr 86 06:07:28-CST
        Received: from BRL-SEM.ARPA by SRI-NIC.ARPA with TCP; Tue 15 Apr 86 04:52:09-PST
        Date: Tue, 15 Apr 86 6:39:04 EST
        From: Mike Muuss <mike@BRL.ARPA>
        To: Hostmaster@sri-nic.arpa
        cc: tcp-ip@sri-nic.arpa
        Subject: Bad host entries

        Turns out that there is more fun.
        The Berkeley HTABLE program will not accept CPUs of the form:
        SUN-2/50,they must instead be written SUN2-50. No slashes.

        There are 5 such errors in the new host table:

        This problem exists with both 4.2 and 4.3 BSD version of HTABLE,
        so I suggest that the table gets fixed.

Bull pucky.

0. htable is perfectly capable of handling the slash. There are slashed
   CpuTypes all through hosts.txt. Why should just these five bomb out?

   htable's parser expects OpSys, CpuType, and so forth to be a NAME
   token, which the lexical analyzer returns on encountering:

        [A-Z] or [A-Z][0-9A-Z./-]*[0-9A-Z]

   The embedded slash is quite kosher.

   There ARE problems with hosts.txt.530, but this isn't one of them.

1. htable does not like lower case in NAMEs; get those entries
   containing "Berkeley" out of there! ("BERKELEY", perhaps?)

2. The lexer throws away the ;-commented lines at the beginning;
   the parser is reporting syntax errors on line numbers other than
   physical line numbers in the hosts source file.

   The five offending hosts are in fact

HOST : : TIE1.BU.EDU,TIE1 : 3B2 : UNIX : UDP :
HOST : : TIE2.BU.EDU,TIE2 : 3B2 : UNIX : UDP :
HOST : : TIE4.BU.EDU,TIE4 : 3B2 : UNIX : UDP :

   As noted above, htable expects a NAME to begin with an ALPHA, i.e. [A-Z].
   3Bx doesn't quite cut it.

3. I edited the "Berkeley" (->"BERKELEY") and "3Bx" (-> "ATT3Bx") entries
   in hosts.txt.530 and ran it through htable on a 4.2 and a 4.3 beta
   machine. Worked fine on both.

4. I agree completely with your comment "I suggest that the table gets fixed."



This archive was generated by hypermail 2.0b3 on Thu Mar 09 2000 - 14:36:06 GMT