Skip to content

Don't include footer row in example values during mapping - #305

Merged
rossjones merged 1 commit into
mainfrom
816-inaccurate-preview-filter
Sep 2, 2025
Merged

rossjones merged 1 commit into
mainfrom
816-inaccurate-preview-filter

Conversation

@rossjones

Copy link
Copy Markdown
Contributor

Currently during mapping we identify the data range to select some examples from, but the call that is made expects an inclusive range, and we want an exclusive one. Due to the lack of options wrt range types we instead just decrement the range we want to retrieve.

We may want to consider having different range types to avoid having to add/subtract one based on requirements, and perhaps make iteration and calculations easier

Currently during mapping we identify the data range to select some
examples from, but the call that is made expects an inclusive range, and
we want an exclusive one. Due to the lack of options wrt range types we
instead just decrement the range we want to retrieve.

We may want to consider having different range types to avoid having to
add/subtract one based on requirements, and perhaps make
iteration and calculations easier

@simonwo simonwo left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is there any milage in always using an inclusive/exclusive definition of ranges?

@rossjones

Copy link
Copy Markdown
Contributor Author

Is there any milage in always using an inclusive/exclusive definition of ranges?

I think generally we have inclusive ranges, but there are cases where we don't want that and it feels error prone to have to determine that on a case by case, and make sure we bound any loops with <= or < than the upper bound of the range. It feels a lot easier to have a default of inclusive range, and allow a exclusive range type for those other cases. We don't need to keep track of intent and if they both have a iteration trait we can traverse the ranges without fear of getting it wrong.

@rossjones
rossjones merged commit e7d59c6 into main Sep 2, 2025
8 of 9 checks passed
@rossjones
rossjones deleted the 816-inaccurate-preview-filter branch September 2, 2025 08:05
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Development

Successfully merging this pull request may close these issues.

2 participants