Skip to content

refactor(expr-ir): Generate dispatch methods (wip) #3073

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 4 commits into from
Sep 1, 2025
Merged

refactor(expr-ir): Generate dispatch methods (wip) #3073

merged 4 commits into from
Sep 1, 2025

Conversation

@dangotbanned
Copy link
Member

@dangotbanned dangotbanned commented Sep 1, 2025 •
edited
Loading

What type of PR is this? (check all applicable)

  • 💾 Refactor
  • ✨ Feature
  • 🐛 Bug Fix
  • 🔧 Optimization
  • 📝 Documentation
  • ✅ Test
  • 🐳 Other

Related issues

Checklist

  • Code follows style guide (ruff)
  • Tests added
  • Documented the changes

If you have comments or can explain your changes, please do so below

Will eventually be replacing most of this (so far only cast 😄):

narwhals/narwhals/_plan/protocols.py

Lines 149 to 200 in c72c3e4

class ExprDispatch(StoresVersion, Protocol[FrameT_contra, R_co, NamespaceT_co]):
_DISPATCH: ClassVar[Mapping[type[ExprIR], Callable[[Any, ExprIR, Any, str], Any]]] = {
expr.Column: lambda self, node, frame, name: namespace(self).col(
node, frame, name
),
expr.Literal: lambda self, node, frame, name: namespace(self).lit(
node, frame, name
),
expr.Len: lambda self, node, frame, name: namespace(self).len(node, frame, name),
# expr.Cast: lambda self, node, frame, name: self.cast(node, frame, name),
expr.Sort: lambda self, node, frame, name: self.sort(node, frame, name),
expr.SortBy: lambda self, node, frame, name: self.sort_by(node, frame, name),
expr.Filter: lambda self, node, frame, name: self.filter(node, frame, name),
agg.First: lambda self, node, frame, name: self.first(node, frame, name),
agg.Last: lambda self, node, frame, name: self.last(node, frame, name),
agg.ArgMin: lambda self, node, frame, name: self.arg_min(node, frame, name),
agg.ArgMax: lambda self, node, frame, name: self.arg_max(node, frame, name),
agg.Sum: lambda self, node, frame, name: self.sum(node, frame, name),
agg.NUnique: lambda self, node, frame, name: self.n_unique(node, frame, name),
agg.Std: lambda self, node, frame, name: self.std(node, frame, name),
agg.Var: lambda self, node, frame, name: self.var(node, frame, name),
agg.Quantile: lambda self, node, frame, name: self.quantile(node, frame, name),
agg.Count: lambda self, node, frame, name: self.count(node, frame, name),
agg.Max: lambda self, node, frame, name: self.max(node, frame, name),
agg.Mean: lambda self, node, frame, name: self.mean(node, frame, name),
agg.Median: lambda self, node, frame, name: self.median(node, frame, name),
agg.Min: lambda self, node, frame, name: self.min(node, frame, name),
expr.BinaryExpr: lambda self, node, frame, name: self.binary_expr(
node, frame, name
),
expr.RollingExpr: lambda self, node, frame, name: self.rolling_expr(
node, frame, name
),
expr.AnonymousExpr: lambda self, node, frame, name: self.map_batches(
node, frame, name
),
expr.FunctionExpr: lambda self, node, frame, name: self._dispatch_function(
node, frame, name
),
# NOTE: Keeping it simple for now
# When adding other `*_range` functions, this should instead map to `range_expr`
expr.RangeExpr: lambda self, node, frame, name: namespace(self).int_range(
node, frame, name
),
expr.OrderedWindowExpr: lambda self, node, frame, name: self.over_ordered(
node, frame, name
),
expr.WindowExpr: lambda self, node, frame, name: self.over(node, frame, name),
expr.Ternary: lambda self, node, frame, name: self.ternary_expr(
node, frame, name
),
}

And then something similar for this group as well:

narwhals/narwhals/_plan/protocols.py

Lines 201 to 222 in c72c3e4

_DISPATCH_FUNCTION: ClassVar[
Mapping[type[Function], Callable[[Any, FunctionExpr, Any, str], Any]]
] = {
boolean.AnyHorizontal: lambda self, node, frame, name: namespace(
self
).any_horizontal(node, frame, name),
boolean.AllHorizontal: lambda self, node, frame, name: namespace(
self
).all_horizontal(node, frame, name),
F.SumHorizontal: lambda self, node, frame, name: namespace(self).sum_horizontal(
node, frame, name
),
F.MinHorizontal: lambda self, node, frame, name: namespace(self).min_horizontal(
node, frame, name
),
F.MaxHorizontal: lambda self, node, frame, name: namespace(self).max_horizontal(
node, frame, name
),
F.MeanHorizontal: lambda self, node, frame, name: namespace(self).mean_horizontal(
node, frame, name
),
strings.ConcatHorizontal: lambda self, node, frame, name: namespace(

@dangotbanned dangotbanned changed the title refactor(expr-ir): Generate dispatch methods refactor(expr-ir): Generate dispatch methods (wip) Sep 1, 2025
@dangotbanned dangotbanned marked this pull request as ready for review September 1, 2025 16:19
@dangotbanned dangotbanned merged commit 92666dd into expr-ir/nw-namespace Sep 1, 2025
32 of 35 checks passed
@dangotbanned dangotbanned deleted the expr-ir/dispatch-gen branch September 1, 2025 16:20
@dangotbanned dangotbanned mentioned this pull request Sep 1, 2025
Merged
10 tasks
dangotbanned added a commit that referenced this pull request Sep 1, 2025
@dangotbanned
refactor(expr-ir): Remove all default redefs
b9611ab 
Handled by #3073 **without** any other changes 🥳
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

internal

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@dangotbanned