From 428fe84ab74044520043dd613b492c7cdc04d48c Mon Sep 17 00:00:00 2001
From: Elena Mokeeva <25465835+ryzheboka@users.noreply.github.com>
Date: Mon, 26 Sep 2022 16:02:14 +0200
Subject: [PATCH] Add support for edge markers (#46)

* support \b and \A

* support \B and \z
---
 pkg/dialect/base/chars.go      | 41 ++++++++++++++++++++++++++++++----
 pkg/dialect/base/chars_test.go | 16 +++++++++++++
 2 files changed, 53 insertions(+), 4 deletions(-)

diff --git a/pkg/dialect/base/chars.go b/pkg/dialect/base/chars.go
index 9ae8964..676d98b 100644
--- a/pkg/dialect/base/chars.go
+++ b/pkg/dialect/base/chars.go
@@ -128,6 +128,13 @@ func (CharsBaseDialect) Begin() ClassToken {
 	return newClassToken(helper.ByteToken('^')).withoutBrackets()
 }
 
+// Begin of text (even if the flag EnableMultiline is set)
+//
+// Regex: `\A`.
+func (CharsBaseDialect) BeginOfText() ClassToken {
+	return newClassToken(helper.StringToken(`\A`)).withoutBrackets()
+}
+
 // End of text or line if the flag EnableMultiline is set.
 //
 // Regex: `$`.
@@ -135,6 +142,31 @@ func (CharsBaseDialect) End() ClassToken {
 	return newClassToken(helper.ByteToken('$')).withoutBrackets()
 }
 
+// End of text (even if the flag EnableMultiline is set).
+//
+// Regex: `\z`.
+func (CharsBaseDialect) EndOfText() ClassToken {
+	return newClassToken(helper.StringToken(`\z`)).withoutBrackets()
+}
+
+// A word boundary for ACII words. Following positions count as word boundaries:
+//   - Beginning of string: If the first character is an ASCII word character.
+//   - End of string: If the last character is an ASCII word character.
+//   - Between a word and a non-word character.
+//
+// Regex: `\b`.
+func (CharsBaseDialect) ASCIIWordBoundary() ClassToken {
+	return newClassToken(helper.StringToken(`\b`)).withoutBrackets()
+}
+
+// A non-word boundary:
+// A position between two word characters or two non-word characters.
+//
+// Regex: `\B`.
+func (CharsBaseDialect) NotASCIIWordBoundary() ClassToken {
+	return newClassToken(helper.StringToken(`\B`)).withoutBrackets()
+}
+
 // Any character, possibly including newline if the flag AnyIncludeNewLine() is set.
 //
 // Regex: `.`.
@@ -146,8 +178,9 @@ func (CharsBaseDialect) Any() ClassToken {
 // It is safe to pass unicode characters.
 //
 // Example usage:
-//   Runes("a") // == Chars.Single('a')
-//   Runes("ab") // == Common.Class(Chars.Single('a'), Chars.Single('b'))
+//
+//	Runes("a") // == Chars.Single('a')
+//	Runes("ab") // == Common.Class(Chars.Single('a'), Chars.Single('b'))
 //
 // Regex: `[abc]`.
 func (CharsBaseDialect) Runes(val string) ClassToken {
@@ -204,7 +237,7 @@ func (CharsBaseDialect) Single(r rune) ClassToken {
 //
 // Example usage:
 //
-//   Chars.Unicode(unicode.Greek)
+//	Chars.Unicode(unicode.Greek)
 //
 // Regex: `\p{Greek}`.
 func (d CharsBaseDialect) Unicode(table *unicode.RangeTable) ClassToken {
@@ -230,7 +263,7 @@ func (d CharsBaseDialect) Unicode(table *unicode.RangeTable) ClassToken {
 //
 // Example usage:
 //
-//   Chars.UnicodeByName("Greek")
+//	Chars.UnicodeByName("Greek")
 //
 // Regex: `\p{Greek}`.
 func (CharsBaseDialect) UnicodeByName(name string) ClassToken {
diff --git a/pkg/dialect/base/chars_test.go b/pkg/dialect/base/chars_test.go
index 8c43c48..5c03bb1 100644
--- a/pkg/dialect/base/chars_test.go
+++ b/pkg/dialect/base/chars_test.go
@@ -78,10 +78,26 @@ func TestRexChars_base(t *testing.T) {
 		Name:     "begin",
 		Chain:    []dialect.Token{base.Chars.Begin()},
 		Expected: `^`,
+	}, {
+		Name:     "beginOfText",
+		Chain:    []dialect.Token{base.Chars.BeginOfText()},
+		Expected: `\A`,
 	}, {
 		Name:     "end",
 		Chain:    []dialect.Token{base.Chars.End()},
 		Expected: `$`,
+	}, {
+		Name:     "endOfText",
+		Chain:    []dialect.Token{base.Chars.EndOfText()},
+		Expected: `\z`,
+	}, {
+		Name:     "ASCIIWordBoundary",
+		Chain:    []dialect.Token{base.Chars.ASCIIWordBoundary()},
+		Expected: `\b`,
+	}, {
+		Name:     "notASCIIWordBoundary",
+		Chain:    []dialect.Token{base.Chars.NotASCIIWordBoundary()},
+		Expected: `\B`,
 	}, {
 		Name:     "single",
 		Chain:    []dialect.Token{base.Chars.Single('a')},