Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!rice!sun-spots-request
From: lwv27@cas.BITNET
Newsgroups: comp.sys.sun
Subject: Spellin resolution
Keywords: Software
Message-ID: <8578@brazos.Rice.edu>
Date: 5 Jun 90 19:16:00 GMT
Sender: root@rice.edu
Organization: Sun-Spots
Lines: 40
Approved: Sun-Spots@rice.edu
X-Refs:  Original: v9n187, Replies: v9n199
X-Sun-Spots-Digest: Volume 9, Issue 201, message 3

Finally!  After many false starts and stumblings I have figured out what
appears to be the necessary steps to add a set of plain ASCII text words
into a 'compressed spelling list' under SunOS 4.1 on SparcStations.  Here
is what I do:

1. First, get the list of ASCII words. One word per line.
2. Call /usr/lib/spell/hashmake with this list.  The output is a list of
   hashed numbers.

          hashmake < list1 > /tmp/t1$$

3. To merge into an existing hashed list (such as hlista), you have to
   call hashcheck - note that on my system, I had to call it differently than
   documented under section 1 of spell.

          hashcheck </usr/lib/spell/hista >> /tmp/t1$$

4. Now, sort the hash codes.  I use the -u option as well - no use in
   adding a word multiple times.

          sort -u /tmp/t1$$ -o /tmp/t1$$

5. Now it is time to call spellin.  Note that spellin wants as an argument
   to know how many codes to read.  Giving it too small a number results in
   only part of the list going in.  Giving it too large a number SEEMS to
   work out - but I didn't bother.

   Note that some statistics come out on stderr - you might want to redirect
   these to /dev/null.

          spellin `wc -l </tmp/t1$$` < /tmp/t1$$ > myhlista

6. And of course, now the hashed list of numbers are no longer needed.

          rm -f /tmp/t1$$

Larry W. Virden
Business: UUCP: osu-cis!chemabs!lwv27  INET: lwv27%cas.BITNET@CUNYVM.CUNY.Edu
Personal: 674 Falls Place,   Reynoldsburg,OH 43068-1614
Proline: lvirden@pro-tcc.cts.com   America Online: lvirden    CIS: [75046,606]