Parse comment blocks out to commentValues

Blocks of comments are now concatenated with their related lines within
the same block comment and added to the commentValues property as one
stripped, concatenated string

If there are n code blocks, there will be n stripped, concatenated elements
in the commentValues (plus x line comments)

Author: adammcarter

Sources/SwiftSyntax/Trivia.swift

@@ -46,15 +46,10 @@ public struct Trivia: Sendable {
   ///
   /// Each element in the array is the trimmed contents of a line comment, or, in the case of a multi-line comment a trimmed, concatenated single string.
   public var commentValues: [String] {
-    pieces.compactMap {
-      switch $0 {
-      case .lineComment(let text), .docLineComment(let text):
-        sanitizingLineComment(text)
-
-      default:
-        nil
-      }
-    }
+    [
+      sanitizedLineCommentValues,
+      sanitizedBlockCommentValues,
+    ].flatMap { $0 }
   }
 
   /// The length of all the pieces in this ``Trivia``.
@@ -231,18 +226,53 @@ extension RawTriviaPiece: CustomDebugStringConvertible {
   }
 }
 
-private func sanitizingLineComment(_ text: String) -> String {
-  // TODO: adammcarter - can we import Foundation instead and use trimmingCharacters(in:)
-
-  var charactersToDrop = 0
+private extension Trivia {
+  var sanitizedLineCommentValues: [String] {
+    compactMap {
+      switch $0 {
+      case .lineComment(let text), .docLineComment(let text):
+        String(sanitizingLineComment(text))
 
-  for character in text {
-    if (character == "/" || character == " ") {
-      charactersToDrop += 1
-    } else {
-      break
+      default:
+        nil
+      }
     }
   }
 
-  return String(text.dropFirst(charactersToDrop))
+  func sanitizingLineComment(_ text: String) -> Substring {
+    text.trimmingPrefix("/ ")
+  }
+
+  var sanitizedBlockCommentValues: [String] {
+    var lines = [String]()
+    var substrings = [Substring]()
+    var foundTerminator = false
+
+    for piece in self {
+      switch piece {
+      case .blockComment(let text), .docBlockComment(let text):
+        let sanitized = text.trimmingCharacters(in: "/* ")
+
+        if substrings.isEmpty || sanitized.isEmpty == false {
+          substrings.append(sanitized)
+        }
+
+        foundTerminator = text.hasSuffix("*/")
+
+      default:
+        break
+      }
+
+      if foundTerminator, substrings.isEmpty == false {
+        lines.append(substrings.joined(separator: " "))
+        substrings.removeAll()
+      }
+    }
+
+    if substrings.isEmpty == false {
+      lines.append(substrings.joined(separator: " "))
+    }
+
+    return lines
+  }
 }

Sources/SwiftSyntax/Utils.swift

@@ -120,3 +120,45 @@ extension RawUnexpectedNodesSyntax {
     self.init(raw: raw)
   }
 }
+
+extension String {
+  func trimmingCharacters(in charactersToTrim: any BidirectionalCollection<Character>) -> Substring {
+    // TODO: adammcarter - this feels a bit dirty
+    self[startIndex...].trimmingPrefix(charactersToTrim).trimmingSuffix(charactersToTrim)
+  }
+
+  func trimmingPrefix(_ charactersToTrim: any BidirectionalCollection<Character>) -> Substring {
+    self[startIndex...].trimmingPrefix(charactersToTrim)
+  }
+
+  func trimmingSuffix(_ charactersToTrim: any BidirectionalCollection<Character>) -> Substring {
+    self[startIndex...].trimmingSuffix(charactersToTrim)
+  }
+}
+
+extension Substring {
+  func trimmingPrefix(_ charactersToTrim: any BidirectionalCollection<Character>) -> Self {
+    dropFirst(countOfSequentialCharacters(charactersToTrim, in: self))
+  }
+
+  func trimmingSuffix(_ charactersToTrim: any BidirectionalCollection<Character>) -> Self {
+    dropLast(countOfSequentialCharacters(charactersToTrim, in: reversed()))
+  }
+}
+
+private func countOfSequentialCharacters(
+  _ charactersToCount: any BidirectionalCollection<Character>,
+  in characters: any BidirectionalCollection<Character>
+) -> Int {
+  var count = 0
+
+  for character in characters {
+    if charactersToCount.contains(character) {
+      count += 1
+    } else {
+      break
+    }
+  }
+
+  return count
+}

Tests/SwiftSyntaxTest/TriviaTests.swift

@@ -150,5 +150,134 @@ class TriviaTests: XCTestCase {
         "Other",
       ]
     )
+
+    // MARK: block comment
+
+    XCTAssertEqual(
+      Trivia(pieces: [.blockComment("")]).commentValues,
+      [""]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [.blockComment("Some block comment")]).commentValues,
+      ["Some block comment"]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [.blockComment("/** Some block comment */")]).commentValues,
+      ["Some block comment"]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .blockComment("/** Some block comment"),
+        .blockComment("* spread on many lines */"),
+      ]).commentValues,
+      [
+        "Some block comment spread on many lines"
+      ]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .blockComment("/** Some block comment"),
+        .blockComment("* spread on many lines"),
+        .blockComment("*/"),
+      ]).commentValues,
+      [
+        "Some block comment spread on many lines"
+      ]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .blockComment("/** Some block comment"),
+        .blockComment("* spread on many lines */"),
+        .blockComment("/** Another block comment */"),
+      ]).commentValues,
+      [
+        "Some block comment spread on many lines",
+        "Another block comment",
+      ]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .blockComment("/** Some block comment"),
+        .blockComment("* spread on many lines */"),
+        .newlines(2),
+        .blockComment("/** Another block comment */"),
+      ]).commentValues,
+      [
+        "Some block comment spread on many lines",
+        "Another block comment",
+      ]
+    )
+
+    // MARK: doc block comment
+
+    XCTAssertEqual(
+      Trivia(pieces: [.docBlockComment("")]).commentValues,
+      [""]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [.docBlockComment("Some doc block comment")]).commentValues,
+      ["Some doc block comment"]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [.docBlockComment("/** Some doc block comment */")]).commentValues,
+      ["Some doc block comment"]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .docBlockComment("/** Some doc block comment"),
+        .docBlockComment("* spread on many lines */"),
+      ]).commentValues,
+      [
+        "Some doc block comment spread on many lines"
+      ]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .docBlockComment("/** Some doc block comment"),
+        .docBlockComment("* spread on many lines"),
+        .docBlockComment("*/"),
+      ]).commentValues,
+      [
+        "Some doc block comment spread on many lines"
+      ]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .docBlockComment("/** Some doc block comment"),
+        .docBlockComment("* spread on many lines */"),
+        .docBlockComment("/** Another doc block comment */"),
+      ]).commentValues,
+      [
+        "Some doc block comment spread on many lines",
+        "Another doc block comment",
+      ]
+    )
+
+    XCTAssertEqual(
+      Trivia(pieces: [
+        .docBlockComment("/** Some doc block comment"),
+        .docBlockComment("* spread on many lines */"),
+        .newlines(2),
+        .docBlockComment("/** Another doc block comment */"),
+      ]).commentValues,
+      [
+        "Some doc block comment spread on many lines",
+        "Another doc block comment",
+      ]
+    )
+
+    // TODO: adammcarter - mixing trivia of lines and blocks
+    // TODO: adammcarter - newline chars in the prefix/suffix of code block too
   }
 }