Generating Test Cases - but why is one is consistently wrong?

Share your app/project
1

Using Davinci to generate simple test cases from simple requirements but it keeps generating “1000.01” as an entry to be rejected. According to the requirement that entry should be accepted. Am I missing something? Two attempts pasted below…

Write test cases for requirement.
Requirement: “Dollar Amount” field must be numeric, greater than zero, up to 9 positions including 1 or to 2 decimal places for cents.
Example:
Test case: Enter “0”, Expected Result: entry rejected.

Test case: Enter “0.1”, Expected Result: entry accepted.
Test case: Enter “999.9”, Expected Result: entry accepted.
Test case: Enter “999.99”, Expected Result: entry accepted.
Test case: Enter “1000”, Expected Result: entry accepted.
Test case: Enter “1000.1”, Expected Result: entry rejected.
Test case: Enter “-1”, Expected Result: entry rejected.
Test case: Enter “a”, Expected Result: entry rejected.
Test case: Enter “9a”, Expected Result: entry rejected.

Write test cases for requirement.
Requirement: “Dollar Amount” field must be numeric, greater than zero, up to 9 positions including 1 or to 2 decimal places for cents.
Example:
Test case: Enter “10.1”, Expected Result: entry accepted.

Test case: Enter “999.99”, Expected Result: entry accepted.
Test case: Enter “1000”, Expected Result: entry accepted.
Test case: Enter “1000.01”, Expected Result: entry rejected.
Test case: Enter “-0.1”, Expected Result: entry rejected.
Test case: Enter “a”, Expected Result: entry rejected.
Test case: Enter “9a”, Expected Result: entry rejected.

2

well the wording is a little weird

3

GPT3 does not do well with numeric reasoning. You might try a fine-tuned model

4

The wording is odd, you are right. Haste makes waste! Thank you for pointing that out.

5

Thank you for the tip about GPT3. Interesting.