Improve using .target_number in Story Table

[plocket]

• Allow .target_number of 0 to set .there_are_any to False.
• Allow .target_number to set .there_are_any in general so the author doesn't need to set .there_are_any separately.
• Fix .target_number using one fewer loop than it should.
• Create feature tests for those items.

[plocket]

This is a wilder ride than I anticipated. Adding some record of the digging so far. This is how the story table flow currently works for the relevant code:

Notes: Each 'row' is an object representing a row.

• for each page

  • for each target_number row

    • keep the target_number row (don't do anything to it)
    • create the there_are_any row with a reference to the original target_number as its source property
      • get the there_are_any value
    • create there_is_another row with a reference to the original target_number as its source property
      • get the there_is_another value depending on the accumulated times_used value in the original target_number row

  • for each field

    • find which table row matches the field's variable name
    • set the answer of the field using the table row value
    • increment times_used of the row and also of the source object in the row. Mutating the source is what allows target_number to persistently accumulate the times it's used and later tell there_is_another which value it should for that particular page.

Problem: Take the following row

| children.target_number | 2 |  |

If there_are_any needs to be set by the user, it adds to the number of times target_number is used (times_used). That counts 1 use for there_are_any and 1 use for there_is_another, coming out to the correct 2 uses. But if the author sets there_are_any themselves, that first accumulation doesn't happen. The test only counts 1 use for there_is_another since there_are_any never gets asked. In that case, the test thinks it hasn't added enough children and tries to add another child.

My solution: We need to track there_are_any separately from there_is_another for target_number rows. Then we use only the there_is_another accumulator to get a consistent number. We can't use that number to determine (for the report) which rows were used through the whole test. Example:

| children.target_number | 0 |  |

That should correctly set .there_are_any to False and will be a legitimate use of the target_number row, but will not show up in the there_is_another accumulator.

In order to write a correct report of what rows were used, we can either keep times_used as it is or add up the two separate accumulators and get rid of times_used.

[plocket]

Closed by #709