Not only that - as parsing those files is the topic for what feels like a half year perhaps rethinking the global approach, timings needed and computational power available and tools to be used are in order. Importing in a database and then analyzing/reporting probably is highly possible, but analyzing non-relational ASCII files via parsing on a large scale pushing necessary performance is not really a vfp best-of-breed use case. The whole approach currently selected is probably subobtimal - an iterative import of all files with provisions to re-import classes of import files not totally correct in first run would probably be much better. That much of the offered code is clearly against DRY imperative is all to visible.

>Consider
>
>(1) Structuring the code - don't know whether you'll still understand the code in a month or so
> Create a class, add methods
>
>(2) Using regular expressions instead of wordnum/strextract/at and the like. They will make the work easier and easier to understand
>

>

>