GHC.Core.Make

Bind a binding group over an expression, using a let or case as appropriate (see GHC.Core#let_can_float_invariant)

Bind a list of binding groups over an expression. The leftmost binding group becomes the outermost group in the resulting expression

Construct an expression which represents the application of one expression to the other

Construct an expression which represents the application of a number of expressions to another. The leftmost expression in the list is applied first

Construct an expression which represents the application of a number of expressions to that of a data constructor expression. The leftmost expression in the list is applied first

Create a lambda where the given expression has a number of variables bound over it. The leftmost binder is that bound by the outermost lambda in the result

Make a case expression whose case binder is unused The alts and res_ty should not have any occurrences of WildId

Make a wildcard binder. This is typically used when you need a binder that you expect to use only at a *binding* site. Do not use it at occurrence sites because it has a single, fixed unique, and it's very easy to get into difficulties with shadowing. That's why it is used so little.

See Note [WildCard binders] in GHC.Core.Opt.Simplify.Env

Sort the variables, putting type and covars first, in scoped order, and then other Ids

It is a deterministic sort, meaning it doesn't look at the values of Uniques. For explanation why it's important See Note [Unique Determinism] in GHC.Types.Unique.

Create a CoreExpr which will evaluate to a Word with the given value

Create a CoreExpr which will evaluate to the given Int

Create a CoreExpr which will evaluate to the given Int

Create a CoreExpr which will evaluate to the given Int. Don't check that the number is in the range of the target platform Int

Create a CoreExpr which will evaluate to the given Integer

Create a CoreExpr which will evaluate to the given Natural

Create a CoreExpr which will evaluate to the given Float

Create a CoreExpr which will evaluate to the given Double

Create a CoreExpr which will evaluate to the given Char

Create a CoreExpr which will evaluate to the given String

Create a CoreExpr which will evaluate to a string morally equivalent to the given FastString

Applies the floats from right to left. That is wrapFloats [b1, b2, …, bn] u = let b1 in let b2 in … in let bn in u

Build the type of a small tuple that holds the specified variables One-tuples are flattened; see Note [Flattening one-tuples]

Build a small tuple holding the specified expressions One-tuples are flattened; see Note [Flattening one-tuples]

Build a small unboxed tuple holding the specified expressions. Do not include the RuntimeRep specifiers; this function calculates them for you. Does not flatten one-tuples; see Note [Flattening one-tuples]

Build an unboxed sum.

Alternative number ("alt") starts from 1.

Make a core tuple of the given boxity; don't flatten 1-tuples

The unit expression

Lifts a "small" constructor into a "big" constructor by recursive decomposition

Split a list into lists that are small enough to have a corresponding tuple arity. The sub-lists of the result all have length <= mAX_TUPLE_SIZE But there may be more than mAX_TUPLE_SIZE sub-lists

Build a big tuple holding the specified variables One-tuples are flattened; see Note [Flattening one-tuples] Arguments don't have to have kind Type

Build the type of a big tuple that holds the specified variables One-tuples are flattened; see Note [Flattening one-tuples]

Build the type of a big tuple that holds the specified type of thing One-tuples are flattened; see Note [Flattening one-tuples]

Build a "big" tuple holding the specified expressions One-tuples are flattened; see Note [Flattening one-tuples] Arguments don't have to have kind Type; ones that do not are boxed This function crashes (in wrapBox) if given a non-Type argument that it doesn't know how to box.

Builds a selector which scrutinises the given expression and extracts the one name from the list given. If you want the no-shadowing rule to apply, the caller is responsible for making sure that none of these names are in scope.

If there is just one Id in the tuple, then the selector is just the identity.

If necessary, we pattern match on a "big" tuple.

A tuple selector is not linear in its argument. Consequently, the case expression built by mkBigTupleSelector must consume its scrutinee Many times. And all the argument variables must have multiplicity Many.

mkBigTupleSelectorSolo is like mkBigTupleSelector but one-tuples are NOT flattened (see Note [Flattening one-tuples])

Builds a selector which scrutinises the given expression and extracts the one name from the list given. If you want the no-shadowing rule to apply, the caller is responsible for making sure that none of these names are in scope.

If there is just one Id in the tuple, then the selector is just the identity.

If necessary, we pattern match on a "big" tuple.

A tuple selector is not linear in its argument. Consequently, the case expression built by mkBigTupleSelector must consume its scrutinee Many times. And all the argument variables must have multiplicity Many.

A generalization of mkBigTupleSelector, allowing the body of the case to be an arbitrary expression.

To avoid shadowing, we use uniques to invent new variables.

If necessary we pattern match on a "big" tuple.

Makes a list [] for lists of the specified type

Makes a list (:) for lists of the specified type

Make a list containing the given expressions, where the list has the given type

Make a fully applied foldr expression

Make a build expression applied to a locally-bound worker function

Makes a Nothing for the specified type

Makes a Just from a value of the specified type