Overhaul the Block parser

z/mark.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"regexp"
 	"strconv"
+	"unicode"
 
 	"github.com/rwxrob/scan"
 	"github.com/rwxrob/term"
@@ -67,109 +68,113 @@ func (s *Block) String() string { return string(s.V) }
 func Blocks(in string) []*Block {
 
 	var blocks []*Block
-	verbpre := regexp.MustCompile(` {4,}`)
+
 	s := scan.R{Buf: []byte(to.Dedented(in))}
+	//s.Trace++
 
-MAIN:
 	for s.Scan() {
 
-		if s.Rune == '*' { // bulleted list
+		// bulleted list
-			if !s.Peek(" ") {
+		if s.Peek("* ") {
-				goto PARA
+			var beg, end int
-			}
+			beg = s.Pos - 1
-			m := s.Pos - 1
+
 			for s.Scan() {
 				if s.Peek("\n\n") {
-					blocks = append(blocks, &Block{Bulleted, s.Buf[m:s.Pos]})
+					end = s.Pos - 1
-					s.Pos += 2
+					s.Pos++
-					continue MAIN
+					break
 				}
 			}
+
+			blocks = append(blocks, &Block{Bulleted, s.Buf[beg:end]})
+			continue
 		}
 
-		if s.Rune == '1' { // numbered list
+		// numbered list
-			if !s.Peek(". ") {
+		if s.Peek("1. ") {
-				goto PARA
+			var beg, end int
-			}
+			beg = s.Pos - 1
-			m := s.Pos - 1
+
 			for s.Scan() {
 				if s.Peek("\n\n") {
-					blocks = append(blocks, &Block{Numbered, s.Buf[m:s.Pos]})
+					end = s.Pos - 1
-					s.Pos += 2
+					s.Pos++
-					continue MAIN
+					break
 				}
 			}
+
+			blocks = append(blocks, &Block{Numbered, s.Buf[beg:end]})
+			continue
 		}
 
-		if s.Rune == ' ' { // verbatim
+		// verbatim
-			s.Pos -= 1
+		if ln := s.Match(begVerbatim); ln >= 4 {
-			ln := s.Match(verbpre)
+			s.Pos--
-			s.Pos++
 
-			if ln < 0 {
+			var beg, end int
-				continue
+			beg = s.Pos
-			}
-			pre := s.Buf[s.Pos-1 : s.Pos+ln-1]
-			s.Pos += len(pre) - 1
 
-			var block []byte
 			for s.Scan() {
 
-				if s.Rune == '\n' {
+				if s.Peek("\n\n") {
-
+					s.Pos++
-					// add in indented lines
+					end = s.Pos - 2
-					if s.Peek(string(pre)) {
+					break
-						block = append(block, '\n')
-						s.Pos += len(pre)
-						continue
-					}
-
-					// end of the block
-					blocks = append(blocks, &Block{Verbatim, block})
-					continue MAIN
 				}
 
-				block = append(block, []byte(string(s.Rune))...)
 			}
 
-		}
+			dedented := to.Dedented(string(s.Buf[beg:end]))
-
+			blocks = append(blocks, &Block{Verbatim, []byte(dedented)})
-		if s.Rune == '\n' || s.Rune == '\r' || s.Rune == '\t' {
 			continue
 		}
 
-	PARA:
+		// paragraph (default)
-		{
+		if !unicode.IsSpace(s.Rune) {
-			var block []byte
+
-			block = append(block, []byte(string(s.Rune))...)
+			buf := []byte(string(s.Rune))
+
 			for s.Scan() {
 
 				if s.Peek("\n\n") {
-					block = append(block, []byte(string(s.Rune))...)
+					s.Pos++
-					blocks = append(blocks, &Block{Paragraph, block})
+					break
-					s.Scan()
-					s.Scan()
-					continue MAIN
 				}
 
-				if s.Rune == '\n' || s.Rune == '\r' {
+				if ln := s.Match(ws); ln > 0 {
-					block = append(block, ' ')
+					buf = append(buf, ' ')
+					s.Pos += ln - 1
 					continue
 				}
 
-				block = append(block, []byte(string(s.Rune))...)
+				buf = append(buf, []byte(string(s.Rune))...)
-			}
 
-			if len(block) > 0 {
-				blocks = append(blocks, &Block{Paragraph, block})
 			}
 
-		} // PARA
+			if len(buf) > 0 {
+				blocks = append(blocks, &Block{Paragraph, buf})
+			}
+			continue
+		}
 
 	}
 	return blocks
 }
 
+// don't expose these until mark has own package
+
+var begVerbatim = regexp.MustCompile(`^ {4,}`)
+var ws = regexp.MustCompile(`^[\s\r\n]+`)
+var begUnder = regexp.MustCompile(`^<\p{L}`)
+var endUnder = regexp.MustCompile(`^\p{L}>`)
+var begBoldItalic = regexp.MustCompile(`^\*{3}\p{L}`)
+var endBoldItalic = regexp.MustCompile(`^\p{L}\*{3}`)
+var begBold = regexp.MustCompile(`^\*{2}\p{L}`)
+var endBold = regexp.MustCompile(`^\p{L}\*{2}`)
+var begItalic = regexp.MustCompile(`^\*\p{L}`)
+var endItalic = regexp.MustCompile(`^\p{L}\*`)
+
 // Emph renders BonzaiMark emphasis spans specifically for
 // VT100-compatible terminals (which almost all are today):
 //
@@ -187,13 +192,15 @@ func Emph[T string | []byte | []rune](buf T) string {
 	for s.Scan() {
 
 		// <under>
-		if s.Rune == '<' {
+		if s.Match(begUnder) > 0 {
 			nbuf = append(nbuf, '<')
 			nbuf = append(nbuf, []rune(term.Under)...)
 			for s.Scan() {
-				if s.Rune == '>' {
+				if s.Match(endUnder) > 0 {
+					nbuf = append(nbuf, s.Rune)
 					nbuf = append(nbuf, []rune(term.Reset)...)
 					nbuf = append(nbuf, '>')
+					s.Pos++
 					break
 				}
 				nbuf = append(nbuf, s.Rune)
@@ -202,13 +209,15 @@ func Emph[T string | []byte | []rune](buf T) string {
 		}
 
 		// ***BoldItalic***
-		if s.Rune == '*' && s.Peek("**") {
+		if s.Match(begBoldItalic) > 0 {
-			s.Pos += 2
+			s.Scan()
+			s.Scan()
 			nbuf = append(nbuf, []rune(term.BoldItalic)...)
 			for s.Scan() {
-				if s.Rune == '*' && s.Peek("**") {
+				if s.Match(endBoldItalic) > 0 {
-					s.Pos += 2
+					nbuf = append(nbuf, s.Rune)
 					nbuf = append(nbuf, []rune(term.Reset)...)
+					s.Pos += 3
 					break
 				}
 				nbuf = append(nbuf, s.Rune)
@@ -217,12 +226,13 @@ func Emph[T string | []byte | []rune](buf T) string {
 		}
 
 		// **Bold**
-		if s.Rune == '*' && s.Peek("*") {
+		if s.Match(begBold) > 0 {
-			s.Pos++
+			s.Pos += 1
 			nbuf = append(nbuf, []rune(term.Bold)...)
 			for s.Scan() {
-				if s.Rune == '*' && s.Peek("*") {
+				if s.Match(endBold) > 0 {
-					s.Pos++
+					nbuf = append(nbuf, s.Rune)
+					s.Pos += 2
 					nbuf = append(nbuf, []rune(term.Reset)...)
 					break
 				}
@@ -232,11 +242,13 @@ func Emph[T string | []byte | []rune](buf T) string {
 		}
 
 		// *Italic*
-		if s.Rune == '*' {
+		if s.Match(begItalic) > 0 {
 			nbuf = append(nbuf, []rune(term.Italic)...)
 			for s.Scan() {
-				if s.Rune == '*' {
+				if s.Match(endItalic) > 0 {
+					nbuf = append(nbuf, s.Rune)
 					nbuf = append(nbuf, []rune(term.Reset)...)
+					s.Pos++
 					break
 				}
 				nbuf = append(nbuf, s.Rune)

z/mark_test.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 
 	Z "github.com/rwxrob/bonzai/z"
+	//	"github.com/rwxrob/scan"
 	"github.com/rwxrob/term"
 )
 
@@ -13,6 +14,7 @@ func init() {
 	term.BoldItalic = `<bolditalic>`
 	term.Under = `<under>`
 	term.Reset = `<reset>`
+	//	scan.Trace++
 }
 
 func ExampleLines() {
@@ -37,12 +39,13 @@ func ExampleBlocks_bulleted() {
 			`
 
 	blocks := Z.Blocks(in)
-	fmt.Printf("%v %q\n", blocks[1].T, blocks[1])
+	fmt.Printf("%q\n", blocks[0])
-	fmt.Printf("%v %q\n", blocks[2].T, blocks[2])
+	fmt.Printf("%q\n", blocks[1])
 
 	//Output:
-	// 3 "* another block\n* here"
+	// "* some thing\n* another thing"
-	// 1 "*boldnotbullet*"
+	// "* another block\n* here"
+
 }
 
 func ExampleBlocks_numbered() {
@@ -56,10 +59,14 @@ func ExampleBlocks_numbered() {
 
 			`
 
-	fmt.Printf("%q\n", Z.Blocks(in)[1])
+	blocks := Z.Blocks(in)
+	fmt.Printf("%q\n", blocks[0])
+	fmt.Printf("%q\n", blocks[1])
 
 	//Output:
+	// "1. some thing\n2. another thing"
 	// "1. another block\n2. here"
+
 }
 
 func ExampleBlocks_paragraph() {
@@ -73,17 +80,19 @@ func ExampleBlocks_paragraph() {
 			here on multiple
 			lines.
 
-			And another one here
+			And another   one here
 			with just a bit more.
 
 			`
 
-	fmt.Printf("%q\n", Z.Blocks(in)[0])
+	blocks := Z.Blocks(in)
-	fmt.Printf("%q\n", Z.Blocks(in)[1])
+	fmt.Printf("%q\n", blocks[0])
+	fmt.Printf("%q\n", blocks[1])
 
 	// Output:
 	// "Simple paragraph here on multiple lines."
 	// "And another one here with just a bit more."
+
 }
 
 func ExampleBlocks_verbatim() {
@@ -107,9 +116,10 @@ func ExampleBlocks_verbatim() {
 
 			`
 
-	fmt.Printf("%q\n", Z.Blocks(in)[0])
+	blocks := Z.Blocks(in)
-	fmt.Printf("%q\n", Z.Blocks(in)[1])
+	fmt.Printf("%q\n", blocks[0])
-	fmt.Printf("%q\n", Z.Blocks(in)[2])
+	fmt.Printf("%q\n", blocks[1])
+	fmt.Printf("%q\n", blocks[2])
 
 	//Output:
 	// "Must have another block type first."
@@ -124,8 +134,10 @@ func ExampleEmph_under() {
 	term.Under = `<under>`
 	term.Reset = `<reset>`
 	fmt.Println(Z.Emph("<UNDER>"))
+	fmt.Println(Z.Emph("< UNDER >"))
 	// Output:
 	// <<under>UNDER<reset>>
+	// < UNDER >
 }
 
 func ExampleEmph_boldItalic() {
@@ -140,16 +152,20 @@ func ExampleEmph_bold() {
 	term.Bold = `<bold>`
 	term.Reset = `<reset>`
 	fmt.Println(Z.Emph("**Bold**"))
+	fmt.Println(Z.Emph("** Bold **"))
 	// Output:
 	// <bold>Bold<reset>
+	// ** Bold **
 }
 
 func ExampleEmph_italic() {
 	term.Italic = `<italic>`
 	term.Reset = `<reset>`
 	fmt.Println(Z.Emph("*Italic*"))
+	fmt.Println(Z.Emph("* Italic *"))
 	// Output:
 	// <italic>Italic<reset>
+	// * Italic *
 }
 
 func ExampleEmph_basics() {