# LiteMarkup Syntax (version: 0.1)
## Primitives
- *LF* = U+000A `\n` (line feed)
- *EOF* = end of input
- *EOL* = *LF* or *EOF*
- *SPACE* = U+0020 ` `
- *TAB* = U+0009 `\t`
- *WS* = one or more *SPACE* or *TAB*
- *INDENT* = 0-3 spaces
A `?` suffix indicates the element is optional (e.g., *WS*? means optional whitespace).
## Document Structure
*DOCUMENT* = zero or more *BLOCK*
*BLOCK* = *BLANKLINES* | *THEMATICBREAK* | *HEADING* | *HTMLBLOCK* | *CODEBLOCK* | *TABLE* | *BLOCKQUOTE* | *LIST* | *PARAGRAPH*
Blocks are matched in the order listed above. The first matching rule wins.
## Leaf Blocks
### Blank Lines
*BLANKLINES* = one or more lines containing only whitespace
Blank lines are consumed and produce no AST node. They serve as separators between blocks.
### Thematic Break
*THEMATICBREAK* = *INDENT* + three or more of (`-`, `_`, or `*`) each optionally followed by spaces + *EOL*
Examples:
- `---`
- `- - -`
- `***`
- `___`
### Heading
*HEADING* = *INDENT* + 1-6 `#` characters + *WS* + content + *EOL*
The number of `#` characters determines the heading level (1-6). The content is parsed as inline.
Examples:
- `# Heading 1`
- `## Heading 2`
- `###### Heading 6`
### Code Block (Fenced)
*CODEBLOCK* = opening fence + content + (*EOF* | closing fence)
- Opening fence: *INDENT* + 3+ backticks + optional info text (no backticks allowed) + *LF*
- Content: any characters until closing fence (or *EOF*)
- Closing fence: same indent + same number of backticks + *WS*? + *EOL*
The indent of the opening fence is stripped from each content line. The final newline before the closing fence is considered part of the content.
### HTML Block
*HTMLBLOCK* = `<` + tagname + optional attributes + `>` + *WS*? + *LF* + content + *LF* + `` + *WS*? + *EOL* + *WS*? + *EOL*
- Tag name: lowercase letter followed by lowercase letters, digits, or hyphens (`[a-z][a-z0-9-]*`)
- Opening and closing tags must match exactly
- Must be followed by a blank line or end of input
### Table
*TABLE* = *HEADERROW* + *LF* + *DELIMITERROW* + *EOL* + zero or more (*BODYROW* + *EOL*)
*HEADERROW* = `|` + one or more (*CELLCONTENT* + `|`) + *WS*?
*DELIMITERROW* = `|` + one or more (*WS*? + one or more `-` + *WS*? + `|`) + *WS*?
*BODYROW* = `|` + one or more (*CELLCONTENT* + `|`) + *WS*?
*CELLCONTENT* = any characters except unescaped `|` and newlines (escaped pipes `\|` are literal)
Cell content is parsed as inline. Each row should have the same number of columns. Implementations may decide to truncate or pad rows with too many/few cells, or end the table parsing after last valid row.
Leading and trailing `|` are always required on every row.
Example:
```
| Variable | Value |
| --------- | ----- |
| X | 30 |
| Y | 25 |
```
## Container Blocks
### Blockquote
*BLOCKQUOTE* = one or more consecutive lines starting with *INDENT* + `>` + optional content
Each line must be prefixed with `>`. The `> ` prefix (with optional space) is stripped, and the remaining content is parsed recursively as a document. Blockquotes are not lazy — every line must have the `>` prefix.
Example:
```
> This is a quote.
> It continues here.
>
> > Nested quote.
```
### List
*LIST* = one or more *LISTITEM*s with consistent marker style
*LISTMARKER* = *INDENT* + (`-`, `+`, or `*` for unordered) or (1-9 digits + `.` or `)` for ordered) + 1-3 spaces
*LISTITEM* = *LISTMARKER* + content + continuation lines
Continuation lines must be indented to align with the first content character after the marker.
Examples:
```
- Item one
- Item two
with continuation
1. First
2. Second
1) Also valid
2) Ordered list
```
Nested lists and other blocks can appear inside list items with proper indentation.
## Paragraph
*PARAGRAPH* = one or more lines of text not matching any other block rule
A paragraph continues until:
- A blank line
- A heading start (`# `)
- A code block start (three backticks)
- A table (header + delimiter row)
- A blockquote start (`>`)
- A list marker (`- `, `* `, `1. `, etc.)
Paragraph content is parsed as inline.
## Inline Elements
Inline parsing applies to heading content, paragraph content, and table cell content. Inline rules are matched in the order listed below; the first matching rule wins.
### Code Span
*CODESPAN* = *n* backticks + content + *n* backticks (same count)
Content is rendered verbatim. Newlines in content become spaces. If content starts and ends with a space (and has other non-space content), those spaces are trimmed.
### Hard Line Break
*LINEBREAK* = `\` + *EOL*
A backslash at end of line creates a `
`.
### Escape
*ESCAPE* = `\` + ASCII punctuation character
Escapable characters: ``- ! " # $ % & ' ( ) * + , . : ; < = > ? @ ^ _ ` { | } ~ / \ [ ]``
The backslash is removed and the character is rendered literally.
### Emphasis (LiteMarkup mode)
*ITALIC* = `_` + content + `_`\
*BOLD* = `*` + content + `*`\
*STRIKETHROUGH* = `~` + content + `~`
For emphasis delimiters (`_`, `*`, `~`) to open and close, they must be adjacent to non-space content:
- opening delimiter must not be followed by a space
- closing delimiter must not be preceded by a space
Content cannot contain unescaped backticks.
Examples:
- `_italic_` → italic
- `*bold*` → bold
- `~deleted~` → strikethrough
- `*bold with _italic_ inside*`
### Emphasis (markdown mode)
When `markdownMode: true`:
*ITALIC* = `_` or `*` + content + same delimiter\
*BOLD* = `__` or `**` + content + same delimiter\
*STRIKETHROUGH* = `~~` + content + `~~`
The same non-space-adjacent delimiter rule applies in markdown mode.
### Link
*LINK* = `[` + text + `]` + (`<` + url + `>` | `(` + url + `)`)
Text is parsed as inline, but square brackets must be escaped, except for embedded images.
URL is parsed verbatim. URL cannot be empty.
Examples:
- `[click here]`
- `[click here](https://example.com/path?a=1)`
The angle bracket syntax is useful for URLs containing parentheses.
### Image
*IMAGE* = `!` + `[` + alt text + `]` + (`<` + url + `>` | `(` + url + `)`)
Examples:
- `![alt text]`
- ``
Alt text must escape any square brackets.
URL is parsed verbatim. URL cannot be empty.
### Text
Any characters not matching the above rules are rendered as plain text.