is a covering space of
ProvedHatcher.isCoveringMap_circleCoverLet be Hatcher's map onto the unit circle. Then is a covering space of in Hatcher's sense: every point has an open neighbourhood such that is a disjoint union of open subsets of , each mapped homeomorphically onto by (such a is evenly covered). In symbols,
Hatcher records this as the observation that "any open arc in is evenly covered". It is the hypothesis under which the path- and homotopy-lifting properties are applied in the proof of Theorem 1.7, so it is the first brick of the computation .
Formalization Note The covering-space condition is Mathlib's IsCoveringMap: for each , the fibre is discrete and is locally trivial over an open neighbourhood of with that fibre. This is equivalent to Hatcher's condition ; neither formulation requires to be surjective. Mathlib already knows Circle.isCoveringMap_exp; here the map is Circle.exp precomposed with .
import Definitions.Def_Hatcher_Circle import Mathlib open unitInterval Hatcher
namespace Hatcher theorem isCoveringMap_circleCover : IsCoveringMap circleCover := by sorry end Hatcher
Read-back
What the Lean code literally says, in plain math · claude-fable-5-1
Declaration Hatcher.isCoveringMap_circleCover. The statement has no hypotheses and no free variables; it asserts a single closed proposition about one fixed map.
The map. Write for the unit circle in , i.e. the set , carrying the subspace topology inherited from (this is Mathlib's Circle). The custom function (called circleCover in the code) is defined for every real by
where is the real number and the value is regarded as a point of . (Concretely, is Mathlib's Circle.exp applied to the real number , and Circle.exp t is the point of the unit circle.) The domain carries its usual topology.
The assertion. The theorem states that is a covering map in the following precise sense (Mathlib's IsCoveringMap): for every point , the point is evenly covered by with fibre indexed by the set , meaning that all of the following hold:
- the set , given the subspace topology from , is a discrete topological space; and
- there exists a subset such that
- ,
- is open in ,
- the preimage is open in , and
- there exists a homeomorphism
(where $p^{-1}(U)$ has the subspace topology from $\mathbb{R}$, $U$ the subspace topology from $S^1$, $p^{-1}(\{x\})$ the subspace topology from $\mathbb{R}$, and the right-hand side the product topology) such that for every $y \in p^{-1}(U)$, the first coordinate of $H(y)$, viewed as a point of $S^1$, equals $p(y)$; i.e. $H$ commutes with the projections to $U$.
Notes on what is and is not said.
- The fibre type used in the local trivialisation is specifically the preimage of the single point under , not an arbitrary discrete space; it is required to be discrete but is not required to be nonempty, and its cardinality is not specified.
- Nothing is asserted about being surjective, continuous, or a local homeomorphism as separate claims; only the pointwise even-covering condition above is asserted, for every .
- The statement quantifies over all points of ; since is nonempty, the claim is not vacuous.
- The proof of the declaration is omitted (
sorry); the read-back describes only the statement.
Confirmed by the mission captain (proposal self-audit).