Escaping

2026-07-15 13:09:56 +02:00
parent 5cc4e43431
commit b244a89eae
3 changed files with 116 additions and 46 deletions
@@ -8,7 +8,7 @@ Lorem ipsum dolor sit amet, consectetur adipiscing elit. Nam blandit dolor nibh,
 #[Link to nekoweb.org in the middle](https://n0va.nekoweb.org) ^^
 #[https://krzak.org]
 #(https://krzak.org)
-^ Links with no alt text
+\^ Links with no alt text

 $[https://krzak.org/krzak-icon.png|img]
 $[/staircase.mp3]
@@ -56,6 +56,8 @@ Here is some __underlined__ text

 [Go to Top](#Heading)

+! Escaped star: \*
+
 [&&1] Ordered list item 2 not included
 [&&named] This is the definition for a named footnote

@@ -65,6 +67,7 @@ $(counter=0)
 $(link=https://git.krzak.org/N0VA/TypoML)

 `tlm.counter` = $(counter)
+\^ This is JS!!!

 #($(link))[Repo link]

@@ -5,8 +5,10 @@ import "core:strings"
 import "core:unicode"

 escape_html :: proc(s: string, allocator := context.temp_allocator) -> string {
+	escaped := decode_escapes(s, allocator)
+
 	b := strings.builder_make(allocator)
-	for ch in s {
+	for ch in escaped {
 		switch ch {
 		case '&':
 			strings.write_string(&b, "&amp;")
@@ -121,6 +121,15 @@ push_text :: proc(parent: ^Token, text: string) {
 	push_token(target, TokenType.Text, cleaned)
 }

+push_text_line :: proc(target: ^Token, text: string) {
+	if len(target.children) > 0 {
+		br := new(Token)
+		br^ = Token{TokenType.BreakLine, nil, nil, target}
+		append(&target.children, br)
+	}
+	push_text(target, text)
+}
+
 strip_alignment_markers :: proc(text: string) -> string {
 	trimmed := strings.trim_space(text)
 	if strings.has_suffix(trimmed, ">>") {
@@ -144,6 +153,92 @@ strip_alignment_markers :: proc(text: string) -> string {
 	return trimmed
 }

+ESCAPE_PLACEHOLDER_BASE :: 0xE000 // Unicode Private Use Area
+
+is_escapable_char :: proc(b: byte) -> bool {
+	switch b {
+	case '\\',
+	     '*',
+	     '_',
+	     '~',
+	     '^',
+	     '`',
+	     '[',
+	     ']',
+	     '(',
+	     ')',
+	     '$',
+	     '&',
+	     '!',
+	     '>',
+	     '|',
+	     '-',
+	     '=',
+	     '#',
+	     '.':
+		return true
+	}
+	return false
+}
+
+escape_placeholder_rune :: proc(b: byte) -> rune {
+	return rune(ESCAPE_PLACEHOLDER_BASE + int(b))
+}
+
+is_escape_placeholder :: proc(r: rune) -> bool {
+	return r >= ESCAPE_PLACEHOLDER_BASE && r < ESCAPE_PLACEHOLDER_BASE + 256
+}
+
+escape_placeholder_to_byte :: proc(r: rune) -> byte {
+	return byte(int(r) - ESCAPE_PLACEHOLDER_BASE)
+}
+
+unescape_backslashes :: proc(s: string, allocator := context.temp_allocator) -> string {
+	if !strings.contains(s, "\\") {
+		return s
+	}
+
+	b := strings.builder_make(allocator)
+	i := 0
+	for i < len(s) {
+		if s[i] == '\\' && i + 1 < len(s) && is_escapable_char(s[i + 1]) {
+			strings.write_rune(&b, escape_placeholder_rune(s[i + 1]))
+			i += 2
+		} else {
+			strings.write_byte(&b, s[i])
+			i += 1
+		}
+	}
+	return strings.to_string(b)
+}
+
+decode_escapes :: proc(s: string, allocator := context.temp_allocator) -> string {
+	has_placeholder := false
+	for r in s {
+		if is_escape_placeholder(r) {
+			has_placeholder = true
+			break
+		}
+	}
+	if !has_placeholder {
+		return s
+	}
+
+	b := strings.builder_make(allocator)
+	for r in s {
+		if is_escape_placeholder(r) {
+			strings.write_byte(&b, escape_placeholder_to_byte(r))
+		} else {
+			strings.write_rune(&b, r)
+		}
+	}
+	return strings.to_string(b)
+}
+
+is_space_byte :: proc(b: byte) -> bool {
+	return b == ' ' || b == '\t'
+}
+
 parse_header :: proc(line: string) -> (int, string) {
 	if len(line) == 0 {
 		return 0, line
@@ -221,6 +316,14 @@ parse_inline_style :: proc(token: ^Token, symbol: byte, toggles: []StyleToggle)
 					i += 1
 				}

+				preceded_by_space := symbol_start > 0 && is_space_byte(str[symbol_start - 1])
+				followed_by_space := i < len(str) && is_space_byte(str[i])
+				if preceded_by_space && followed_by_space {
+					push_token(current_node, TokenType.Text, str[symbol_start:i])
+					start = i
+					continue
+				}
+
 				for count > 0 {
 					matched := false
 					for toggle in toggles {
@@ -827,40 +930,6 @@ resolve_variables :: proc(token: ^Token, vars: ^map[string]string) {
 	}
 }

-add_breaklines :: proc(token: ^Token) {
-	if token.type == .CodeBlock {
-		return
-	}
-	if token.children == nil {
-		return
-	}
-
-	new_children: [dynamic]^Token
-	for i := 0; i < len(token.children); i += 1 {
-		child := token.children[i]
-		add_breaklines(child)
-		append(&new_children, child)
-		is_child_inline :=
-			child.type == TokenType.Text ||
-			child.type == TokenType.Left ||
-			child.type == TokenType.Right ||
-			child.type == TokenType.Middle
-		is_next_inline :=
-			i + 1 < len(token.children) &&
-			(token.children[i + 1].type == TokenType.Text ||
-					token.children[i + 1].type == TokenType.Left ||
-					token.children[i + 1].type == TokenType.Right ||
-					token.children[i + 1].type == TokenType.Middle)
-
-		if is_child_inline && is_next_inline {
-			br := new(Token)
-			br^ = Token{TokenType.BreakLine, nil, nil, token}
-			append(&new_children, br)
-		}
-	}
-	token.children = new_children
-}
-
 parse :: proc(lines: []string, allocator := context.allocator) -> Token {
 	root := Token{TokenType.Root, nil, nil, nil}
 	current_token := &root
@@ -900,6 +969,8 @@ parse :: proc(lines: []string, allocator := context.allocator) -> Token {
 			continue
 		}

+		line = unescape_backslashes(line)
+
 		if strings.has_prefix(line, "```") {
 			lang := strings.trim_space(line[3:])
 			current_token = push_token(&root, TokenType.CodeBlock, lang)
@@ -965,7 +1036,7 @@ parse :: proc(lines: []string, allocator := context.allocator) -> Token {
 				push_token(current_token, TokenType.Paragraph, nil)
 			}
 			last_p := current_token.children[len(current_token.children) - 1]
-			push_text(last_p, text[2:])
+			push_text_line(last_p, text[2:])
 			continue
 		}

@@ -1025,15 +1096,15 @@ parse :: proc(lines: []string, allocator := context.allocator) -> Token {

 		if current_token.type == .Blockquote {
 			last_p := current_token.children[len(current_token.children) - 1]
-			push_text(last_p, line)
+			push_text_line(last_p, line)
 		} else if current_token.type == .UnorderedList || current_token.type == .OrderedList {
 			last_li := current_token.children[len(current_token.children) - 1]
-			push_text(last_li, line)
+			push_text_line(last_li, line)
 		} else {
 			if current_token.type != .Paragraph {
 				current_token = push_token(&root, TokenType.Paragraph, nil)
 			}
-			push_text(current_token, line)
+			push_text_line(current_token, line)
 		}
 	}

@@ -1048,12 +1119,6 @@ parse :: proc(lines: []string, allocator := context.allocator) -> Token {
 		pass_through_children(token)
 	}

-	// THIRD PASS
-
-	for token in root.children {
-		add_breaklines(token)
-	}
-
 	// VARIABLE RESOLUTION PASS

 	vars: map[string]string