This is probably possible in theory but exceeds my knowledge level at this time. If I imported smaller pieces of the file I would have to reassemble them after the import which is something I don't know how to do. Also I would have to programmatically figure out when a file needed to be split and treated this way. That also leaves me a bit cold. It might be easier to just process all the tables one record at a time but so far I haven't found the code to inserting the data into the tables short of a rather complicated technique of iterating all the fields via a structure analysis which would get messy in a hurry. Thoughts?

My task here is to improve on the DBC2MDB utility by John Koziol. His code fails when the data fields are empty because the tables created in the MDB are created with required data inputs. At least that's my experience with Access 2003. I can't seem to find any settings in the Access configuration that would overcome this.

If I continue with the XML route I am also going to need a way to validate the data before the import to eliminate any illegal characters in the XML. So far I have found an ASCII 2 and and ASCII 19 that have caused problems. I fixed these with STRTRAN when I found them but short of testing every character in every field I'm not quite sure how to deal with this one.

>Can you split import of one table into multiple parts?
>
>>OK I solved the first problem by converting the fields from text to memo. (That's not really a solution but it took care of the immediate issue.) The second problem turned out to be invalid ASCII characters in the memo fields. Apparently the data entry folks are sometimes clumsy on the keys. But now I have a third new problem. While importing a very large XML (39 meg) consisting of an integer field and a memo field, Access gets to around 1300 records and then hangs with no error messages at all. My various data checkers don't find any problems with the data. More ideas?
>>