Click above to get retro games delivered to your door ever month!

  Debugging:

  Use of HiPer-SEEK can generally be broken into three main parts: Building
  the index, searching the index, and verifying the index returns. If
  problems arise, the HiPer-SEEK related code should be examined in that
  order.

  To build a valid HiPer-SEEK index -- unless you're simply using
  hs_Index() -- you should make sure that the string passed to hs_Add() is
  in fact a string (no numeric types), trim the string right before passing
  it to hs_Add() or hs_Index(), set order to 0 (so the index is built in
  the natural order of the primary .DBF) and set deleted off (so deleted
  records are included).  The HiPer-SEEK index file should contain one
  record for each record in the primary .DBF. The number of records in a
  HiPer-SEEK index can be calculated like this:

    Number_of_Records == (File_Size - 512)/Index_Key_Size

    Example: 1000 == (32512 - 512)/32

  On the search side, first determine if hs_Next() is returning valid
  numbers.  If not, and you have determined that the index file is good,
  examine the code that gets the search string and passes it to hs_Set().
  Remember to trim it right before hs_Set().

  If hs_Next() is returning good values, then examine the verification
  procedure. Be sure the source and target values are constructed properly.
  Check again that the index was built in natural order and that it
  contains entries for all the .DBF records. When this is not the case, the
  number returned by hs_Next() will not correspond the correct record
  number, hs_Verify() will receive the source value from the wrong record
  and the it will only satisfy the verification process by coincidence.
  Most problems are related to simple logic errors and are easily
  corrected.