Remove RegexConsumer and fix its dependencies (#617)

* Remove `RegexConsumer` and fix its dependencies

This eliminates the RegexConsumer type and rewrites its users to call
through to other, existing functionality on Regex or in the Algorithms
implementations. RegexConsumer doesn't take account of the dual
subranges required for matching, so it can produce results that are
inconsistent with matches(of:) and ranges(of:), which were rewritten
earlier.

rdar://102841216

* Remove remaining from-end algorithm methods

This removes methods that are left over from when we were considering
from-end algorithms. These aren't tested and may not have the correct
semantics, so it's safer to remove them entirely.

Author: natecook1000

Sources/_StringProcessing/Algorithms/Algorithms/FirstRange.swift

@@ -20,15 +20,6 @@ extension Collection {
   }
 }
 
-extension BidirectionalCollection {
-  func _lastRange<S: BackwardCollectionSearcher>(
-    of searcher: S
-  ) -> Range<Index>? where S.BackwardSearched == Self {
-    var state = searcher.backwardState(for: self, in: startIndex..<endIndex)
-    return searcher.searchBack(self, &state)
-  }
-}
-
 // MARK: Fixed pattern algorithms
 
 extension Collection where Element: Equatable {
@@ -78,11 +69,7 @@ extension BidirectionalCollection where SubSequence == Substring {
   @_disfavoredOverload
   @available(SwiftStdlib 5.7, *)
   public func firstRange(of regex: some RegexComponent) -> Range<Index>? {
-    _firstRange(of: RegexConsumer(regex))
-  }
-
-  @available(SwiftStdlib 5.7, *)
-  func _lastRange<R: RegexComponent>(of regex: R) -> Range<Index>? {
-    _lastRange(of: RegexConsumer(regex))
+    let s = self[...]
+    return try? regex.regex.firstMatch(in: s)?.range
   }
 }

Sources/_StringProcessing/Algorithms/Algorithms/Ranges.swift

@@ -115,43 +115,6 @@ extension RangesCollection.Index: Comparable {
   }
 }
 
-// MARK: `ReversedRangesCollection`
-
-struct ReversedRangesCollection<Searcher: BackwardCollectionSearcher> {
-  typealias Base = Searcher.BackwardSearched
-  
-  let base: Base
-  let searcher: Searcher
-  
-  init(base: Base, searcher: Searcher) {
-    self.base = base
-    self.searcher = searcher
-  }
-}
-
-extension ReversedRangesCollection: Sequence {
-  public struct Iterator: IteratorProtocol {
-    let base: Base
-    let searcher: Searcher
-    var state: Searcher.BackwardState
-    
-    init(base: Base, searcher: Searcher) {
-      self.base = base
-      self.searcher = searcher
-      self.state = searcher.backwardState(
-        for: base, in: base.startIndex..<base.endIndex)
-    }
-    
-    public mutating func next() -> Range<Base.Index>? {
-      searcher.searchBack(base, &state)
-    }
-  }
-  
-  public func makeIterator() -> Iterator {
-    Iterator(base: base, searcher: searcher)
-  }
-}
-
 // TODO: `Collection` conformance
 
 // MARK: `CollectionSearcher` algorithms
@@ -164,14 +127,6 @@ extension Collection {
   }
 }
 
-extension BidirectionalCollection {
-  func _rangesFromBack<S: BackwardCollectionSearcher>(
-    of searcher: S
-  ) -> ReversedRangesCollection<S> where S.BackwardSearched == Self {
-    ReversedRangesCollection(base: self, searcher: searcher)
-  }
-}
-
 // MARK: Fixed pattern algorithms
 
 extension Collection where Element: Equatable {
@@ -195,37 +150,6 @@ extension Collection where Element: Equatable {
   }
 }
 
-extension BidirectionalCollection where Element: Equatable {
-  // FIXME
-//  public func rangesFromBack<S: Sequence>(
-//    of other: S
-//  ) -> ReversedRangesCollection<ZSearcher<SubSequence>>
-//    where S.Element == Element
-//  {
-//    fatalError()
-//  }
-}
-
-extension BidirectionalCollection where Element: Comparable {
-  func _ranges<C: Collection>(
-    of other: C
-  ) -> RangesCollection<PatternOrEmpty<TwoWaySearcher<Self>>>
-    where C.Element == Element
-  {
-    _ranges(of: PatternOrEmpty(searcher: TwoWaySearcher(pattern: Array(other))))
-  }
-  
-  // FIXME
-//  public func rangesFromBack<S: Sequence>(
-//    of other: S
-//  ) -> ReversedRangesCollection<PatternOrEmpty<TwoWaySearcher<SubSequence>>>
-//    where S.Element == Element
-//  {
-//    rangesFromBack(
-//      of: PatternOrEmpty(searcher: TwoWaySearcher(pattern: Array(other))))
-//  }
-}
-
 @available(SwiftStdlib 5.7, *)
 struct RegexRangesCollection<Output> {
   let base: RegexMatchesCollection<Output>
@@ -299,13 +223,6 @@ extension Collection where SubSequence == Substring {
 }
 
 extension BidirectionalCollection where SubSequence == Substring {
-  @available(SwiftStdlib 5.7, *)
-  func _rangesFromBack<R: RegexComponent>(
-    of regex: R
-  ) -> ReversedRangesCollection<RegexConsumer<R, Self>> {
-    _rangesFromBack(of: RegexConsumer(regex))
-  }
-
   // FIXME: Return `some Collection<Range<Index>>` for SE-0346
   /// Finds and returns the ranges of the all occurrences of a given sequence
   /// within the collection.

Sources/_StringProcessing/Algorithms/Algorithms/Split.swift

@@ -106,121 +106,6 @@ extension SplitCollection: Sequence {
   }
 }
 
-//extension SplitCollection: Collection {
-//  public struct Index {
-//    var start: Base.Index
-//    var base: RangesCollection<Searcher>.Index
-//    var isEndIndex: Bool
-//  }
-//
-//  public var startIndex: Index {
-//    let base = ranges.startIndex
-//    return Index(start: ranges.base.startIndex, base: base, isEndIndex: false)
-//  }
-//
-//  public var endIndex: Index {
-//    Index(start: ranges.base.endIndex, base: ranges.endIndex, isEndIndex: true)
-//  }
-//
-//  public func formIndex(after index: inout Index) {
-//    guard !index.isEndIndex else { fatalError("Cannot advance past endIndex") }
-//
-//    if let range = index.base.range {
-//      let newStart = range.upperBound
-//      ranges.formIndex(after: &index.base)
-//      index.start = newStart
-//    } else {
-//      index.isEndIndex = true
-//    }
-//  }
-//
-//  public func index(after index: Index) -> Index {
-//    var index = index
-//    formIndex(after: &index)
-//    return index
-//  }
-//
-//  public subscript(index: Index) -> Base.SubSequence {
-//    guard !index.isEndIndex else {
-//      fatalError("Cannot subscript using endIndex")
-//    }
-//    let end = index.base.range?.lowerBound ?? ranges.base.endIndex
-//    return ranges.base[index.start..<end]
-//  }
-//}
-//
-//extension SplitCollection.Index: Comparable {
-//   static func == (lhs: Self, rhs: Self) -> Bool {
-//    switch (lhs.isEndIndex, rhs.isEndIndex) {
-//    case (false, false):
-//      return lhs.start == rhs.start
-//    case (let lhs, let rhs):
-//      return lhs == rhs
-//    }
-//  }
-//
-//  static func < (lhs: Self, rhs: Self) -> Bool {
-//    switch (lhs.isEndIndex, rhs.isEndIndex) {
-//    case (true, _):
-//      return false
-//    case (_, true):
-//      return true
-//    case (false, false):
-//      return lhs.start < rhs.start
-//    }
-//  }
-//}
-
-// MARK: `ReversedSplitCollection`
-
-struct ReversedSplitCollection<Searcher: BackwardCollectionSearcher> {
-  public typealias Base = Searcher.BackwardSearched
-  
-  let ranges: ReversedRangesCollection<Searcher>
-  
-  init(ranges: ReversedRangesCollection<Searcher>) {
-    self.ranges = ranges
-  }
-
-  init(base: Base, searcher: Searcher) {
-    self.ranges = base._rangesFromBack(of: searcher)
-  }
-}
-
-extension ReversedSplitCollection: Sequence {
-  public struct Iterator: IteratorProtocol {
-    let base: Base
-    var index: Base.Index
-    var ranges: ReversedRangesCollection<Searcher>.Iterator
-    var isDone: Bool
-    
-    init(ranges: ReversedRangesCollection<Searcher>) {
-      self.base = ranges.base
-      self.index = base.endIndex
-      self.ranges = ranges.makeIterator()
-      self.isDone = false
-    }
-    
-    public mutating func next() -> Base.SubSequence? {
-      guard !isDone else { return nil }
-      
-      guard let range = ranges.next() else {
-        isDone = true
-        return base[..<index]
-      }
-      
-      defer { index = range.lowerBound }
-      return base[range.upperBound..<index]
-    }
-  }
-  
-  public func makeIterator() -> Iterator {
-    Iterator(ranges: ranges)
-  }
-}
-
-// TODO: `Collection` conformance
-
 // MARK: `CollectionSearcher` algorithms
 
 extension Collection {
@@ -237,16 +122,6 @@ extension Collection {
   }
 }
 
-extension BidirectionalCollection {
-  func splitFromBack<Searcher: BackwardCollectionSearcher>(
-    by separator: Searcher
-  ) -> ReversedSplitCollection<Searcher>
-    where Searcher.BackwardSearched == Self
-  {
-    ReversedSplitCollection(base: self, searcher: separator)
-  }
-}
-
 // MARK: Predicate algorithms
 
 extension Collection {
@@ -260,14 +135,6 @@ extension Collection {
   }
 }
 
-extension BidirectionalCollection where Element: Equatable {
-  func splitFromBack(
-    whereSeparator predicate: @escaping (Element) -> Bool
-  ) -> ReversedSplitCollection<PredicateConsumer<Self>> {
-    splitFromBack(by: PredicateConsumer(predicate: predicate))
-  }
-}
-
 // MARK: Single element algorithms
 
 extension Collection where Element: Equatable {
@@ -280,14 +147,6 @@ extension Collection where Element: Equatable {
   }
 }
 
-extension BidirectionalCollection where Element: Equatable {
-  func splitFromBack(
-    by separator: Element
-  ) -> ReversedSplitCollection<PredicateConsumer<Self>> {
-    splitFromBack(whereSeparator: { $0 == separator })
-  }
-}
-
 // MARK: Fixed pattern algorithms
 
 extension Collection where Element: Equatable {
@@ -399,21 +258,6 @@ extension StringProtocol where SubSequence == Substring {
 
 @available(SwiftStdlib 5.7, *)
 extension BidirectionalCollection where SubSequence == Substring {
-  @_disfavoredOverload
-  func split<R: RegexComponent>(
-    by separator: R,
-    maxSplits: Int,
-    omittingEmptySubsequences: Bool
-  ) -> SplitCollection<RegexConsumer<R, Self>> {
-    split(by: RegexConsumer(separator), maxSplits: maxSplits, omittingEmptySubsequences: omittingEmptySubsequences)
-  }
-  
-  func splitFromBack<R: RegexComponent>(
-    by separator: R
-  ) -> ReversedSplitCollection<RegexConsumer<R, Self>> {
-    splitFromBack(by: RegexConsumer(separator))
-  }
-
   // TODO: Is this @_disfavoredOverload necessary?
   // It prevents split(separator: String) from choosing this overload instead
   // of the collection-based version when String has RegexComponent conformance
@@ -431,9 +275,34 @@ extension BidirectionalCollection where SubSequence == Substring {
     maxSplits: Int = .max,
     omittingEmptySubsequences: Bool = true
   ) -> [SubSequence] {
-    Array(split(
-      by: RegexConsumer(separator),
-      maxSplits: maxSplits,
-      omittingEmptySubsequences: omittingEmptySubsequences))
+    var result: [SubSequence] = []
+    var subSequenceStart = startIndex
+    
+    func appendSubsequence(end: Index) -> Bool {
+      if subSequenceStart == end && omittingEmptySubsequences {
+        return false
+      }
+      result.append(self[subSequenceStart..<end])
+      return true
+    }
+    
+    guard maxSplits > 0 && !isEmpty else {
+      _ = appendSubsequence(end: endIndex)
+      return result
+    }
+
+    for match in _matches(of: separator) {
+      defer { subSequenceStart = match.range.upperBound }
+      let didAppend = appendSubsequence(end: match.range.lowerBound)
+      if didAppend && result.count == maxSplits {
+        break
+      }
+    }
+    
+    if subSequenceStart != endIndex || !omittingEmptySubsequences {
+      result.append(self[subSequenceStart..<endIndex])
+    }
+    
+    return result
   }
 }