Conference ilbbak::ibi_focus

158.0. "Problems with MATCH and ON TABLE HOLD / ALIAS" by EFGV01::LEE () Wed Nov 23 1988 09:33

  I had some problems using MATCH. There were no problems when I used 
  MATCH with data files whose master files were made by myself. But 
  when I used MATCH after ON TABLE HOLD it didn't work correctly.
  I'll make it clear with a example.

  There are two data files:
   1) TEL: contains a name and a telephone number.
   2) JOB: contains a name and a jobname.
  There master files and data files are:

======================================================================
FILENAME = TEL, SUFFIX = FIX ,$
SEGNAME  = TEL               ,$
FIELDNAME = A1NAME ,ALIAS =    ,USAGE = A10   ,ACTUAL = A10   ,$
FIELDNAME = A1TEL  ,ALIAS =    ,USAGE = A03   ,ACTUAL = A03   ,$
======================================================================
Data file: TEL.DAT : 

PERSON    508
DOG       555
CAT       999
WORM      444
======================================================================
======================================================================
FILENAME = JOB, SUFFIX = FIX ,$
SEGNAME  = JOB               ,$
FIELDNAME = B1NAME ,ALIAS =       ,USAGE = A10   ,ACTUAL = A10   ,$
FIELDNAME = B1JOB  ,ALIAS =       ,USAGE = A16   ,ACTUAL = A16   ,$
======================================================================
Data file JOB.DAT :

ZORRO     BANDIT
PERSON    TRAINEE DEV SPEC
======================================================================

I combine the two files using MATCH:

MATCH FILE TEL
PRINT A1TEL
BY A1NAME 
RUN
FILE JOB
PRINT B1JOB
BY B1NAME
AFTER MATCH HOLD AS TELJOB OLD-OR-NEW
END

  The result is exactly what one expects; records with matching A1NAME
  and B1NAME are matched, the others remain separate and TELJOB has only 
  one field containing the name.

FILE=TELJOB         ,SUFFIX=FIX                                                 
SEGNAME=TELJOB                                                                  
FIELDNAME   =A1NAME       ,E01          ,A10      ,A10      ,$                  
FIELDNAME   =A1TEL        ,E02          ,A03      ,A03      ,$                  
FIELDNAME   =B1JOB        ,E03          ,A16      ,A16      ,$                  
======================================================================
Data file TELJOB.FTM:

CAT       999.
DOG       555.
PERSON    508TRAINEE DEV SPEC
WORM      444.
ZORRO     .  BANDIT
======================================================================

  However, when the file JOB is created by an ON TABLE HOLD, its master file
  looks like this:

FILENAME = JOB, SUFFIX = FIX ,$
SEGNAME  = JOB               ,$
FIELDNAME = B1NAME ,ALIAS = E01       ,USAGE = A10   ,ACTUAL = A10   ,$
FIELDNAME = B1JOB  ,ALIAS = E02       ,USAGE = A16   ,ACTUAL = A16   ,$

  Now the program gives a different result; TELJOB now has
  two fields containing the name, and the values in these fields don't
  match! FOCUS sorts the two files and then just combines the first 
  record of (sorted) TEL with the first record of (sorted) JOB, the 
  second record of TEL with the second record of JOB, etcetera. It 
  no longer considers them to be the 'same' field.

FILE=TELJOB         ,SUFFIX=FIX                                                 
SEGNAME=TELJOB                                                                  
FIELDNAME   =A1NAME       ,E01          ,A10      ,A10      ,$                  
FIELDNAME   =A1TEL        ,E02          ,A03      ,A03      ,$                  
FIELDNAME   =B1NAME       ,E03          ,A10      ,A10      ,$                  
FIELDNAME   =B1JOB        ,E04          ,A16      ,A16      ,$                  
=======================================================================
Data file TELJOB.FTM:

CAT       999PERSON    TRAINEE DEV SPEC
DOG       555ZORRO     BANDIT
PERSON    508.         .
WORM      444.         .
=======================================================================

  After some testing we discovered that the alias is giving this
  problem. When the alias of the sort-fields of the first and the
  second file are different then this problems occurs; when they
  are exactly the same the MATCH works perfectly. 
  But ON TABLE HOLD and AFTER MATCH HOLD create master files with 
  alias's in the format E%% in which % is a figure. So this means 
  that we can't use MATCH after an ON TABLE HOLD or an AFTER MATCH
  HOLD!! Now I am using ON TABLE SAVE and self-made master files but
  this can't be done with AFTER MATCH HOLD.

  After this long explanation my questions are:
   1) shouldn't the way that MATCH works be independent of the
      use of alias's?
   2) was MATCH already working like this in the old version or is
      this a difference between the old and new version (we are now 
      using version 5.2)? In the old version I never tried this.