rust_markdown_parser

Simplified markdown parser implemented in Rust. Converts markdown into HTML.

Features

This implementation can handle these markdown features:

• Headers
• Thematic breaks
• Lists:
  • Unordered lists
  • Ordered lists
• Code blocks
• Paragraphs
• Links
• Bold and italic text styles

Limitations

For features described above to work, some limitations are applied:

• Text indentaion is not supported. Any line that starts with a space or a tab will be converted to <p> element;
• Nested lists are not supported;
• Nesting of elements in general is not well supported;

Implementation details

Parsing process

The library used for parsing in the project is pest. It is used to build a DAG based on markdown string, and later this DAG is used in construction of an HTML string. To build a DAG it uses grammar, provided in markdown.pest.

Talking about parsing logic - there are two main steps:

1. Parsing all input into block elements.
2. Parsing inline elements of each block.

As we can see, looking at the grammar file, block rule is responsible for the first step, while line and line_pars are responsible for the second step.

Grammar rules

1. space: whitespaces
2. newline: new line symbols
3. blankline: one or more lines that are empty(only spaces allowed)
4. markdown: whole file structure
5. block: block element
6. special_block: all blocks except paragraph
7. header: header element
8. header_hashtags: used to select header variant
9. thematic_break: thematic break element
10. list: combined rule for both lists' variants
11. unordered_list: unordered list element
12. ordered_list: ordered list element
13. code_block: code block element
14. paragraph: paragraph element
15. line: captures all elements to the end of line
16. line_pars: same as line, but also parses it into inline elements
17. inline: inline element
18. link = link element
19. link_text = captures link text
20. link_href = captures link href
21. bold = bold element
22. italic = italic element
23. text = works like line, but stops when other inline elements can be captured