Skip to content

Implement \R, \v, \h for character/scalar modes #384

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 10 commits into from
May 9, 2022
Merged

Implement \R, \v, \h for character/scalar modes #384

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
60456b1
Implement \R, \v, \h for character/scalar modes
natecook1000 May 6, 2022
42414d4
Extra \R vs \v testing
natecook1000 May 6, 2022
3a9e9e1
Merge branch 'main' into newline_handling
natecook1000 May 9, 2022
e71ac84
print("here")
natecook1000 May 9, 2022
6034a90
print("here, really")
natecook1000 May 9, 2022
a49854f
print("1")
natecook1000 May 9, 2022
ab30b49
Calm the line break tests back down
natecook1000 May 9, 2022
57cd4f6
Audit assertions for semantic level
natecook1000 May 9, 2022
646210c
Merge branch 'main' into newline_handling
natecook1000 May 9, 2022
cfc7aed
Allow \v and \V in validation
natecook1000 May 9, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions Sources/_RegexParser/Regex/Parse/Sema.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -182,15 +182,16 @@ extension RegexValidator {
_ esc: AST.Atom.EscapedBuiltin, at loc: SourceLocation
) throws {
switch esc {
case .resetStartOfMatch, .singleDataUnit, .verticalTab, .notVerticalTab,
case .resetStartOfMatch, .singleDataUnit,
// '\N' needs to be emitted using 'emitAny'.
.notNewline:
throw error(.unsupported("'\\\(esc.character)'"), at: loc)

// Character classes.
case .decimalDigit, .notDecimalDigit, .whitespace, .notWhitespace,
.wordCharacter, .notWordCharacter, .graphemeCluster, .trueAnychar,
.horizontalWhitespace, .notHorizontalWhitespace:
.horizontalWhitespace, .notHorizontalWhitespace,
.verticalTab, .notVerticalTab:
break

case .newlineSequence:
Expand Down
32 changes: 25 additions & 7 deletions Sources/_StringProcessing/ByteCodeGen.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,10 +80,16 @@ extension Compiler.ByteCodeGen {
}

case .endOfSubjectBeforeNewline:
builder.buildAssert { (input, pos, bounds) in
builder.buildAssert { [semanticLevel = options.semanticLevel] (input, pos, bounds) in
if pos == input.endIndex { return true }
return input.index(after: pos) == input.endIndex
&& input[pos].isNewline
switch semanticLevel {
case .graphemeCluster:
return input.index(after: pos) == input.endIndex
&& input[pos].isNewline
case .unicodeScalar:
return input.unicodeScalars.index(after: pos) == input.endIndex
&& input.unicodeScalars[pos].isNewline
}
}

case .endOfSubject:
Expand Down Expand Up @@ -115,8 +121,14 @@ extension Compiler.ByteCodeGen {

case .startOfLine:
if options.anchorsMatchNewlines {
builder.buildAssert { (input, pos, bounds) in
pos == input.startIndex || input[input.index(before: pos)].isNewline
builder.buildAssert { [semanticLevel = options.semanticLevel] (input, pos, bounds) in
if pos == input.startIndex { return true }
switch semanticLevel {
case .graphemeCluster:
return input[input.index(before: pos)].isNewline
case .unicodeScalar:
return input.unicodeScalars[input.unicodeScalars.index(before: pos)].isNewline
}
}
} else {
builder.buildAssert { (input, pos, bounds) in
Expand All @@ -126,8 +138,14 @@ extension Compiler.ByteCodeGen {

case .endOfLine:
if options.anchorsMatchNewlines {
builder.buildAssert { (input, pos, bounds) in
pos == input.endIndex || input[pos].isNewline
builder.buildAssert { [semanticLevel = options.semanticLevel] (input, pos, bounds) in
if pos == input.endIndex { return true }
switch semanticLevel {
case .graphemeCluster:
return input[pos].isNewline
case .unicodeScalar:
return input.unicodeScalars[pos].isNewline
}
}
} else {
builder.buildAssert { (input, pos, bounds) in
Expand Down
16 changes: 16 additions & 0 deletions Sources/_StringProcessing/Unicode/ScalarProps.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,3 +46,19 @@ extension Unicode.Script {
return result
}
}

extension UnicodeScalar {
var isHorizontalWhitespace: Bool {
value == 0x09 || properties.generalCategory == .spaceSeparator
}

var isNewline: Bool {
switch value {
case 0x000A...0x000D /* LF ... CR */: return true
case 0x0085 /* NEXT LINE (NEL) */: return true
case 0x2028 /* LINE SEPARATOR */: return true
case 0x2029 /* PARAGRAPH SEPARATOR */: return true
default: return false
}
}
}
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Character has a isNewline. Can we share implementation and/or SPI for this?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also, is this the same list?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's the same list — we should ultimately have this on UnicodeScalar as well; probably with a name to match Character's.

28 changes: 19 additions & 9 deletions Sources/_StringProcessing/_CharacterClassModel.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -178,15 +178,18 @@ public struct _CharacterClassModel: Hashable {
matched = c.isNumber && (c.isASCII || !options.usesASCIIDigits)
case .hexDigit:
matched = c.isHexDigit && (c.isASCII || !options.usesASCIIDigits)
case .horizontalWhitespace: fatalError("Not implemented")
case .newlineSequence:
matched = c.isNewline && (c.isASCII || !options.usesASCIISpaces)
case .verticalWhitespace: fatalError("Not implemented")
case .horizontalWhitespace:
matched = c.unicodeScalars.first?.isHorizontalWhitespace == true
&& (c.isASCII || !options.usesASCIISpaces)
case .newlineSequence, .verticalWhitespace:
matched = c.unicodeScalars.first?.isNewline == true
&& (c.isASCII || !options.usesASCIISpaces)
case .whitespace:
matched = c.isWhitespace && (c.isASCII || !options.usesASCIISpaces)
case .word:
matched = c.isWordCharacter && (c.isASCII || !options.usesASCIIWord)
case .custom(let set): matched = set.any { $0.matches(c, with: options) }
case .custom(let set):
matched = set.any { $0.matches(c, with: options) }
}
if isInverted {
matched.toggle()
Expand All @@ -206,14 +209,21 @@ public struct _CharacterClassModel: Hashable {
matched = c.properties.numericType != nil && (c.isASCII || !options.usesASCIIDigits)
case .hexDigit:
matched = Character(c).isHexDigit && (c.isASCII || !options.usesASCIIDigits)
case .horizontalWhitespace: fatalError("Not implemented")
case .newlineSequence: fatalError("Not implemented")
case .verticalWhitespace: fatalError("Not implemented")
case .horizontalWhitespace:
matched = c.isHorizontalWhitespace && (c.isASCII || !options.usesASCIISpaces)
case .verticalWhitespace:
matched = c.isNewline && (c.isASCII || !options.usesASCIISpaces)
case .newlineSequence:
matched = c.isNewline && (c.isASCII || !options.usesASCIISpaces)
if c == "\r" && nextIndex != str.endIndex && str.unicodeScalars[nextIndex] == "\n" {
str.unicodeScalars.formIndex(after: &nextIndex)
}
case .whitespace:
matched = c.properties.isWhitespace && (c.isASCII || !options.usesASCIISpaces)
case .word:
matched = (c.properties.isAlphabetic || c == "_") && (c.isASCII || !options.usesASCIIWord)
case .custom: fatalError("Not supported")
case .custom(let set):
matched = set.any { $0.matches(Character(c), with: options) }
}
if isInverted {
matched.toggle()
Expand Down
30 changes: 25 additions & 5 deletions Tests/RegexTests/UTS18Tests.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,14 @@ import XCTest
@testable // for internal `matches(of:)`
import _StringProcessing

extension UnicodeScalar {
var value4Digits: String {
let valueString = String(value, radix: 16, uppercase: true)
if valueString.count >= 4 { return valueString }
return String(repeating: "0", count: 4 - valueString.count) + valueString
}
}

class UTS18Tests: XCTestCase {
var input: String {
"ABCdefghîøu\u{308}\u{FFF0} -–—[]123"
Expand Down Expand Up @@ -262,21 +270,33 @@ extension UTS18Tests {
09\u{85}\
10\u{2028}\
11\u{2029}\

12
"""
// Check the input counts
var lines = lineInput.matches(of: regex(#"\d{2}"#))
XCTAssertEqual(lines.count, 11)
XCTAssertEqual(lines.count, 12)
// Test \R - newline sequence
lines = lineInput.matches(of: regex(#"\d{2}\R"#))
lines = lineInput.matches(of: regex(#"\d{2}\R^"#).anchorsMatchLineEndings())
XCTAssertEqual(lines.count, 11)
// Test \v - vertical space
lines = lineInput.matches(of: regex(#"\d{2}\v^"#).anchorsMatchLineEndings())
XCTAssertEqual(lines.count, 11)
// Test anchors as line boundaries
lines = lineInput.matches(of: regex(#"^\d{2}$"#).anchorsMatchLineEndings())
XCTAssertEqual(lines.count, 11)
XCTAssertEqual(lines.count, 12)
// Test that dot does not match line endings
lines = lineInput.matches(of: regex(#".+"#))
XCTAssertEqual(lines.count, 11)
XCTAssertEqual(lines.count, 12)

// Unicode scalar semantics - \R still matches all, including \r\n sequence
lines = lineInput.matches(
of: regex(#"\d{2}\R(?=\d)"#).matchingSemantics(.unicodeScalar).anchorsMatchLineEndings())
XCTAssertEqual(lines.count, 11)
// Unicode scalar semantics - \v matches all except for \r\n sequence
lines = lineInput.matches(
of: regex(#"\d{2}\v(?=\d)"#).matchingSemantics(.unicodeScalar).anchorsMatchLineEndings())
XCTAssertEqual(lines.count, 10)

// Does not contain an empty line
XCTAssertFalse(lineInput.contains(regex(#"^$"#)))
// Does contain an empty line (between \n and \r, which are reversed here)
Expand Down