Conference vaxuum::document_ft

615.0. "No INXtoBRN. HELP PLEASE" by OPUS::PREISIG () Wed Jul 08 1987 16:45

	This problem is explained in note 587 but has not been solved yet.
I'm thinking that maybe it has been inadvertently forgotten so I'm adding
another note.  This is the same problem explained in a different manner.
Please see note 587.2 for a complete log listing.  Not being able to
resolve this problem is beginning to impact us.

	We are experienceing a problem with version 1.0 of document
that I hope someone can solve.  We have DOCUMENT on two systems.  They
are supposed to be identical, but the same files run on one system experience
an error on the other.  The error occurs during index generation.  On one
system, the program INXTOBRN is run, but in the other, it is not.  The
commands I run are the smae on both.  I use the following COM file and submit
the file to batch:

$ Document/nodev/contents/index chappro general ln03
$ document/notag/nodev chappro_contents general ln03
$ document/notag/nodev chappro_index general ln03
$ Document/notag/notext chappro general ln03

The error occurs in the first command string and has the following information
in the log file:

Good run -
.
.
.
%TEX-I-PAGESOUT, 36 pages written.
%TEX-I-OUTFILENAME, 'disk$user1:[preisig]CHAPPRO.DVI_LN3'
[ C o n t e n t s   G e n e r a t i o n ]...
[ I n d e x   G e n e r a t i o n ]...
INXtoBRN Version T1.0
IDX-W-TAGERR Error processing the PREFIX tag on line         91 of the INX file.
IDX-W-CUTPRFX The page prefix was truncated to: GLOSSA . Processing continues.
IDX-W-TAGERR Error processing the PREFIX tag on line         92 of the INX file.
IDX-W-CUTPRFX The page prefix was truncated to: GLOSSA . Processing continues.
IDX-W-TAGERR Error processing the PREFIX tag on line         93 of the INX file.
IDX-W-CUTPRFX The page prefix was truncated to: GLOSSA . Processing continues.
%DOC-I-IDENT, DOCINDEX version X1.00
%DOC-I-PROCFILE, processing file 'DISK$USER1:[PREISIG]CHAPPRO.BRN;1'
%DOC-S-COMPLETE, processing complete 'DISK$USER1:[PREISIG]CHAPPRO.BRN;1'
%DOC-S-CREATED, 'DISK$USER1:[PREISIG]CHAPPRO_INDEX.TEX;5' created
.
.
.


Bad run -
.
.
.
%TEX-I-PAGESOUT, 36 pages written.
%TEX-I-OUTFILENAME, 'opus$dua1:[preisig]CHAPPRO.DVI_LN3'
[ C o n t e n t s   G e n e r a t i o n ]...
[ I n d e x   G e n e r a t i o n ]...
%DOC-I-IDENT, DOCINDEX version X1.00
%DOC-F-OPENIN, error opening 'OPUS$DUA1:[PREISIG]CHAPPRO.BRN;' for input
-XPO-E-NO_FILE, file does not exist
  PREISIG      job terminated at  1-JUL-1987 15:05:35.17


	I'm not concerned about the 3 truncate errors in the good run, as I
realize that that is a known bug.

	We had another problem that took 2 weeks to track down.  It turned out
to be Open File parameter on one system that only allowed 20 files to be open
at once.  This caused an ACP file error in the log that said it couldn't find
the file.  So consider system parameters when responding to this problem.

	We'd appreciate any help you can give.  If the info is already in this
notes file, please point me to the appropriate note.  Thanks.

Ernie

    	Thanks to help from Marty Friedman, I've been able to narrow
    this problem down to DOC$INDEXER.  The program DOC$INDEXER on
    DOC$LOCAL_TOOLS is the same on both systems, but when run on one
    system it runs INXtoBRN (the "good" system), while when run on the
    other (the "bad" system) it does not.  The results are the same
    as shown in the log file for the index generation portion of the
    bookbuild.
    
    	The question then is: How does DOC$INDEXER.EXE run INXtoBRN?
     Since both programs are the same, it seems that it must pass control
    to some other program.  If we knew the mechanism it uses, we could
    determine (hopefully) why it works on one system but not on the
    other.
    
    	Thanks for any help in this area.
    
    Ernie

    
    still looking at this. meanwhile, though, you should be able
    to generate an index file using:
    
    $ docum/keep=inx filename/contents/index etc.
    $ doc$index := $doc$root:[tools]doc$indexer
    $ doc$index filename.inx
    
    

    	There is no DOC$INDEXER at DOC$ROOT:[TOOLS] on our system. 
    There is a DOC$INDEXSHR.EXE there and there is a TDXtoBRN there.
     I tried both of them on an .INX file but they both blew up.
    
    DOC$INDEXER is on DOC$LOCAL_TOOLS.  Thats the one I used in .1 that
    still doesn't run INXtoBRN.
    
    Ernie

    
    We are working on this off-line. At present, it appears that the
    indexing generation is successful when run interactively, but
    fails in a batch job. Has anyone else experienced this?
    
    -patti