For Loop

# Iterate over collection
for name in names:
    print(name)

# Iterate with range
for i in range(10):
    print(i)

# Enumerate for index and value
for index, name in enumerate(names):
    print(f"{index}: {name}")

A tuple is an iterable, never a list: for x in (1, 2, 3) iterates its members without building a collection. The loop variable's type is the tuple's best common element type — the same rule a collection literal uses (see collection_types.md) — so for x in (1, 2.5) binds x as float, and a tuple with no best common type is refused by name (SPY0227) with the steer to annotate the target:

for x in (1, 2.5):
    print(x)              # 1.0 / 2.5  — python3 prints `1` / `2.5`; see the note below

# for x in (1, "a"):      # SPY0227 — no best common type ('int32', 'str')

Departure from Python, and its scope. Python leaves each member at its own type, so for x in (1, 2.5) prints 1 then 2.5. Sharpy binds ONE loop variable, so the element type is the best common type and the first member prints as 1.0. The widening is local to the ITERATION route: unpacking does not widen (a, b = (1, 2.5) prints 1 and 2.5, matching Python), because each target takes its own member's type.

else-clauses are described in loop_else.md.

Implementation: - Collection: ✅ Native - foreach (var item in collection) - range(): 🔄 Lowered - for (int i = 0; i < n; i++) - enumerate(): 🔄 Lowered - .Select((x, i) => (i, x)) - Tuple: 🔄 Lowered - foreach (var item in new T[] { t.Item1, ... }) from the recorded tuple-as-sequence fact