A line-aware SCL parser that includes line/col info in the final AST.

Overview

This parser has three main stages:

1. Tokenization – Converts raw input string into a list of tokens (with line/col).
2. Parsing – Converts the list of tokens into a "raw" AST that reflects the SCL structure (blocks, key-value lines, etc.).
3. Interpretation – Converts the raw AST into a final typed AST.

Highlights

• Typed Values in the final AST:
  • Booleans (true / false)
  • Numbers (floats/ints)
  • Atoms (:atom)
  • Strings (quoted "...", triple ..., single '...', backtick `...`, or unquoted)
• Block Names can have multiple attributes (e.g., config 1, true, "extra"), stored as [1, true, "extra"] in the final AST.
• Line/Column info is included for every key, block, and value in the AST.

Final AST Form

• Key-Value => {{:key_atom, lineK, colK}, {value, lineV, colV}}
  • If multiple comma-separated values appear, the value is a list of tuples instead: {{:key_atom, lineK, colK}, [{val1, l1, c1}, {val2, l2, c2}, ... ]}.
• Block => {{:block_atom, lineB, colB}, {block_name_or_list, lineN, colN}, block_contents}

Usage

iex> SCLParser.parse("foo 42")
{:ok, [{{:foo, 1, 1}, {42, 1, 5}}]}

iex> SCLParser.parse("my_block 3.14 { nested true }")
{:ok,
 [
   {
     {:my_block, 1, 1},
     {3.14, 1, 15},
     [
       {{:nested, 1, 17}, {true, 1, 24}}
     ]
   }
 ]}