Solution or Workaround
The EXPORT command returns this message:
Mismatched or corrupted INFO template
EXPORT executes and leaves behind a corrupt export file. This occurs with coverages originally created with RECEIVE, IMPORT, INFODB, or ARC33-34.
The INFO template is no longer valid. An item in the attribute file has been modified or added since it was brought into the 3.4x environment.
OPTION 1: Delete the corresponding INFO template (AAT.INF, PAT.INF) from the coverage directory and then run EXPORT again. This will produce the default conversion of the items.
OPTION 2: Add or delete the mistmatched items to make the PAT or AAT match the INFO structure specified in the .INF files.
OPTION 3: Modify the PAT.INF or AAT.INF files to reflect the current status of the PAT.DBF or AAT.DBF. This is a good way to add/delete redefined items and convert items before you run EXPORT. Simply modify the .INF file using the following description of INFO template files as a guide.
INFO template files:
IMPORT,EXPORT,SEND,RECEIVE and INFODB create/read INFO template (filter) files for all attribute files associated with a coverage. They are not created for attribute files that are not associated with a coverage.
The template file describes what the INFO data file looked like before it was converted to DBASE. If you don't change the structure of the dBASE file, you can use EXPORT or SEND to reconstruct the INFO data file perfectly. You can resize items and it will still work. It won't work if you change the number of items, the order of items, the name of items or the type of items (numeric to char or char to numeric). If there is a problem with the template file, EXPORT or SEND will not continue and you will get an error message saying the template has changed. If no template file exists, the normal rules for dBASE to INFO translation will be used. If necessary, delete the template file in order to bring the file across.
A template (xxx.INF) is created for each dBASE file (xxx.DBF) in a coverage. This is an ASCII file that you can manipulate (with extreme care). Its format is shown below. There is one line for each item on the file (normal or redefined). Normal items come first followed by redefined items. The order of the normal items must match the order of the items on the dBASE file. Each field is separated by a space.
dbase_name info_name item_pos item_width item_type item_dec item_owidth
10 c 16 c 4 I 3 I 1 I 2 I 3 I
dbase_name 10 character field that holds the name of the dBASE item.
If this is blank, this is a redefined item.
info_name 16 character field that holds the name of the INFO item.
This is the name to use for dbase_name on the INFO
datafile. If this field is blank, then this item is combined
with the previous non-blank item to make one big item.
This is how split character items are handled (character
items >254 c that got split into two or more items during
the DBASE->INFO conversion). Note that dbase_name and
cannot both be blank.
item_pos 4 wide numeric field that holds the position of this
item in the INFO data file. Mainly used by redefined items.
item_width 3 wide numeric field that holds the width of this item
in the INFO data file.
item_type 1 width numeric field that holds the type of this item
in the INFO data file.
1 = date D
2 = character C
3 = Integer I
4 = Numeric N
5 = Binary integer B
6 = Binary real F
dBase only supports date, character numeric item types.
item_dec 2 wide numeric field indicating number of decimal places for
types 4 and 6. All others have either -1 or 0.
item_owidth 3 wide numeric field that holds the output width of the item.