Merge pull request #596 from rctcwyvrn/consume-non-newline

[Optimization] Add instructions for consuming non-newlines and advancing in scalar view

Author: rctcwyvrn

Sources/_StringProcessing/ByteCodeGen.swift

@@ -220,27 +220,16 @@ fileprivate extension Compiler.ByteCodeGen {
     case .graphemeCluster:
       builder.buildAdvance(1)
     case .unicodeScalar:
-      // TODO: builder.buildAdvanceUnicodeScalar(1)
-      builder.buildConsume { input, bounds in
-        input.unicodeScalars.index(after: bounds.lowerBound)
-      }
+      builder.buildAdvanceUnicodeScalar(1)
     }
   }
 
   mutating func emitAnyNonNewline() {
     switch options.semanticLevel {
     case .graphemeCluster:
-      builder.buildConsume { input, bounds in
-        input[bounds.lowerBound].isNewline
-        ? nil
-        : input.index(after: bounds.lowerBound)
-      }
+      builder.buildConsumeNonNewline()
     case .unicodeScalar:
-      builder.buildConsume { input, bounds in
-        input[bounds.lowerBound].isNewline
-        ? nil
-        : input.unicodeScalars.index(after: bounds.lowerBound)
-      }
+      builder.buildConsumeScalarNonNewline()
     }
   }
 

Sources/_StringProcessing/Engine/InstPayload.swift

@@ -196,11 +196,19 @@ extension Instruction.Payload {
     interpret()
   }
 
-  init(distance: Distance) {
-    self.init(distance)
+  init(distance: Distance, isScalarDistance: Bool = false) {
+    self.init(isScalarDistance ? 1 : 0, distance)
   }
-  var distance: Distance {
-    interpret()
+  var distance: (isScalarDistance: Bool, Distance) {
+    let pair: (UInt64, Distance) = interpretPair()
+    return (isScalarDistance: pair.0 == 1, pair.1)
+  }
+  
+  init(isScalar: Bool) {
+    self.init(isScalar ? 1 : 0)
+  }
+  var isScalar: Bool {
+    self.rawValue == 1
   }
 
   init(bool: BoolRegister) {

Sources/_StringProcessing/Engine/Instruction.swift

@@ -122,6 +122,10 @@ extension Instruction {
     /// - If it is inverted
     /// - If it strictly matches only ascii values
     case matchBuiltin
+    
+    /// Matches any non newline character
+    /// Operand: If we are in scalar mode or not
+    case matchAnyNonNewline
 
     // MARK: Extension points
 

Sources/_StringProcessing/Engine/MEBuilder.swift

@@ -142,6 +142,19 @@ extension MEProgram.Builder {
   mutating func buildAdvance(_ n: Distance) {
     instructions.append(.init(.advance, .init(distance: n)))
   }
+  
+  mutating func buildAdvanceUnicodeScalar(_ n: Distance) {
+    instructions.append(
+      .init(.advance, .init(distance: n, isScalarDistance: true)))
+  }
+  
+  mutating func buildConsumeNonNewline() {
+    instructions.append(.init(.matchAnyNonNewline, .init(isScalar: false)))
+  }
+                        
+  mutating func buildConsumeScalarNonNewline() {
+    instructions.append(.init(.matchAnyNonNewline, .init(isScalar: true)))
+  }
 
   mutating func buildMatch(_ e: Character, isCaseInsensitive: Bool) {
     instructions.append(.init(

Sources/_StringProcessing/Engine/Processor.swift

@@ -181,6 +181,18 @@ extension Processor {
     currentPosition = idx
     return true
   }
+  
+  // Advances in unicode scalar view
+  mutating func consumeScalar(_ n: Distance) -> Bool {
+    guard let idx = input.unicodeScalars.index(
+      currentPosition, offsetBy: n.rawValue, limitedBy: end
+    ) else {
+      signalFailure()
+      return false
+    }
+    currentPosition = idx
+    return true
+  }
 
   /// Continue matching at the specified index.
   ///
@@ -329,6 +341,26 @@ extension Processor {
     return true
   }
 
+  // Matches the next character if it is not a newline
+  mutating func matchAnyNonNewline() -> Bool {
+    guard let c = load(), !c.isNewline else {
+      signalFailure()
+      return false
+    }
+    _uncheckedForcedConsumeOne()
+    return true
+  }
+  
+  // Matches the next scalar if it is not a newline
+  mutating func matchAnyNonNewlineScalar() -> Bool {
+    guard let s = loadScalar(), !s.isNewline else {
+      signalFailure()
+      return false
+    }
+    input.unicodeScalars.formIndex(after: &currentPosition)
+    return true
+  }
+
   mutating func signalFailure() {
     guard !savePoints.isEmpty else {
       state = .fail
@@ -477,10 +509,26 @@ extension Processor {
       signalFailure()
 
     case .advance:
-      if consume(payload.distance) {
-        controller.step()
+      let (isScalar, distance) = payload.distance
+      if isScalar {
+        if consumeScalar(distance) {
+          controller.step()
+        }
+      } else {
+        if consume(distance) {
+          controller.step()
+        }
+      }
+    case .matchAnyNonNewline:
+      if payload.isScalar {
+        if matchAnyNonNewlineScalar() {
+          controller.step()
+        }
+      } else {
+        if matchAnyNonNewline() {
+          controller.step()
+        }
       }
-
     case .match:
       let (isCaseInsensitive, reg) = payload.elementPayload
       if isCaseInsensitive {

Tests/RegexTests/CompileTests.swift

@@ -37,6 +37,7 @@ enum DecodedInstr {
   case matchScalarCaseInsensitive
   case matchScalarUnchecked
   case matchBitsetScalar
+  case matchAnyNonNewline
   case matchBitset
   case matchBuiltin
   case consumeBy
@@ -116,7 +117,9 @@ extension DecodedInstr {
         return .matchBitset
       }
     case .consumeBy:
-      return consumeBy
+      return .consumeBy
+    case .matchAnyNonNewline:
+      return .matchAnyNonNewline
     case .assertBy:
       return .assertBy
     case .matchBy: