It's for reconciling customers' items such as point of sale, payroll, and other systems with bank transactions. Each client item has date, description, and amount. Each transaction similarly has date, description and amount. The goal is to match all items with corresponding bank transactions. However, it is more complicated because:
* bank dates and item dates may not be the same. A sale may take a few days to clear the bank.
* amounts may not be the same. A credit card sale may be for 100, but when it clears the bank, the CC provider has deducted fees, so it's 98.
* quantities may not be the same. Ten sales may be deposited as one deposit. In some cases, multiple items may match to multiple transactions.
* there are rules about which matches are best. A sale may match ten transactions, but it can be important to match the right one, typically the closest date.
As others have suggested, I can take shortcuts by ignoring some sets. But it is tricky and slow, especially for large sets.
