· 13,472 chars · 28 deps · depth 13 Reason: Phase B2b: proof by the good-sets argument, with continuity from above derived from continuity from below on complements, and the Lipschitz function built by truncating a multiple of the distance to the inner closed set.
The sets approximable from inside by closed sets and from outside by open sets form a sigma-algebra containing the closed sets, whose neighbourhoods shrink to them by continuity from above; the regularity statements follow, and the Lipschitz function is built by truncating a multiple of the distance to the inner closed set.
Step 1 (Continuity from above). Let (Am)m∈N be a sequence in B(X) with Am+1⊆Am for every m, and let A=⋂m∈NAm. Writing Bm=X∖Am, one has A=X∖⋃m∈NBm, so A∈B(X) because a σ-algebra is closed under complements and countable unions. The sets Bm increase, ⋃mBm=X∖A, and every μ(Bm) is real and at most μ(X), so claim 5 of Basic Properties of a Measure applies and (μ(Bm))m∈Nconverges to μ(X∖A). By claim 3 of Basic Properties of a Measure, μ(Am)=μ(X)−μ(Bm) for every m and μ(A)=μ(X)−μ(X∖A); hence, by claim 3 of Arithmetic of Limits of Real Sequences applied to the constant sequence with value μ(X) and to (μ(Bm))m, the sequence (μ(Am))m∈N converges to μ(A).
Step 2 (A closed set contains its closure points). Let F∈C and let x belong to the closure clX(F). If x∈/F then x∈X∖F, which is open because F is closed, so there is a positive real ε with Bd(x,ε)⊆X∖F, that is, Bd(x,ε)∩F=∅; this contradicts claim 2 of Characterization of the Closure in a Metric Space by Open Balls. Hence x∈F.
Step 3 (Shrinking neighbourhoods of a nonempty closed set). Let F∈C be nonempty, let distd(⋅,F) be the distance to F, and for m∈N put
Step 4 (Closed sets are approximable). Let A be the family of those B∈B(X) such that for every positive real ε there are F∈C and U∈U with F⊆B⊆U and μ(U∖F)≤ε. Then C⊆A. Indeed, let F∈C and let ε be a positive real number. If F=∅, take F itself and U=∅, which is both closed and open by claim 1 of Complements, Unions and Intersections of Closed Sets in a Topological Space and Metric Open Sets Form a Topology, so that μ(U∖F)=μ(∅)=0≤ε. If F=∅, let Vm be as in Step 3; since (μ(Vm))m converges to μ(F) there is m with μ(Vm)−μ(F)≤ε, and F⊆F⊆Vm with μ(Vm∖F)=μ(Vm)−μ(F)≤ε.
Complements. Let B∈A and let ε be a positive real number; choose F⊆B⊆U as in the definition of A. Then X∖U⊆X∖B⊆X∖F, the set X∖U is closed and X∖F is open by Closed Subset of a Topological Space, and (X∖F)∖(X∖U)=U∖F, so μ((X∖F)∖(X∖U))≤ε. Hence X∖B∈A.
Step 6 (Claims 2 and 3). Let B∈B(X) and let S={μ(F):F⊆B,F∈C}. It is nonempty, since ∅∈C and ∅⊆B, and μ(B) is an upper bound for it by claim 2 of Basic Properties of a Measure; so S is bounded above. Let b be an upper bound for S and suppose, for a contradiction, that b<μ(B). Then ε=μ(B)−b is positive, and claim 1, applied with the positive real number 2ε, gives F∈C and U∈U with F⊆B⊆U and μ(U∖F)≤2ε. Since B∖F⊆U∖F, claim 2 of Basic Properties of a Measure gives μ(B)−μ(F)=μ(B∖F)≤2ε, so μ(F)≥μ(B)−2ε=b+2ε; since 0<2ε by claim 8 of Elementary Order Arithmetic in an Ordered Field, claim 1 of that lemma gives b<b+2ε and claim 2 then gives b<μ(F). As μ(F)∈S, this contradicts the assumption that b is an upper bound for S. Hence μ(B)≤b for every upper bound b of S, and μ(B) is the least upper bound of S. This is claim 2.
For claim 3 let S′={μ(U):B⊆U,U∈U}. It is nonempty, since X is open and B⊆X, is bounded below by 0, and μ(B) is a lower bound for it by claim 2 of Basic Properties of a Measure. Let b′ be a lower bound for S′ and suppose μ(B)<b′. Then ε′=b′−μ(B) is positive, and claim 1, applied with 2ε′, gives F∈C and U∈U with F⊆B⊆U and μ(U∖F)≤2ε′. Since U∖B⊆U∖F, claim 2 of Basic Properties of a Measure gives μ(U)−μ(B)=μ(U∖B)≤2ε′, so μ(U)≤μ(B)+2ε′=b′−2ε′; since 0<2ε′ by claim 8 of Elementary Order Arithmetic in an Ordered Field, claims 1 and 2 of that lemma give μ(U)<b′, contradicting that b′ is a lower bound for S′. Hence b′≤μ(B) for every lower bound b′ of S′, and μ(B) is the greatest lower bound of S′.
Step 7 (Claim 4). Let B∈B(X) and let ε be a positive real number. By claim 1, applied with the positive real number 2ε, there are F∈C and U∈U with F⊆B⊆U and μ(U∖F)≤2ε; in particular μ(B∖F)≤2ε, since B∖F⊆U∖F.
If F=∅, take h to be the constant function with value 0, which is Lipschitz with constant 0 and takes values in {0}, and take N=B; then μ(N)=μ(B∖F)≤ε and h(x)=0=1B(x) for every x∈X∖N.
Suppose F=∅ and let Vm be as in Step 3. Since (μ(Vm))m converges to μ(F), choose k∈N with μ(Vk)−μ(F)≤2ε, that is μ(Vk∖F)≤2ε; the order of choice is ε, then F and U, then k. Define
h(x)=max(0,1−ι(k)distd(x,F))(x∈X),
the maximum of two real numbers. Then 0≤h(x), and h(x)≤1 because 1−ι(k)distd(x,F)≤1, the distance being nonnegative and ι(k) positive; so h takes values in [0,1], the lower bound by claim 1 of Elementary Properties of the Maximum of Two Elements and the upper bound by claim 3 of that lemma, applied with the upper bound 1 for both 0 and 1−ι(k)distd(x,F).
Put N=(B∖F)∪(Vk∖F), a member of B(X) with μ(N)≤ε by claim 4 of Basic Properties of a Measure. Let x∈X∖N. If x∈F then distd(x,F)=0 by Vanishing of the Distance to a Set Characterizes the Closure, so h(x)=max(0,1)=1, and 1B(x)=1 because F⊆B. If x∈/F then x∈/Vk, since x∈/Vk∖F, so ι(k)−1≤distd(x,F), and multiplying by the nonnegative number ι(k) gives 1≤ι(k)distd(x,F) by claim 5 of Elementary Arithmetic in an Ordered Field, giving h(x)=max(0,c) with c≤0, that is h(x)=0; and x∈/B, since x∈/B∖F and x∈/F, so 1B(x)=0. In both cases h(x)=1B(x), which proves claim 4.