rev/mpl

developtheweb 1fb7e360f5 Resolve M0 blockers and fix all examples

- Resolved all lexical blockers:
  - Comments: -- single-line, {- -} multi-line
  - Strings: "..." with escapes, """...""" raw
  - Paths: Both 🖫"..." and \path"..."
- Fixed all 10 example files:
  - Added semicolons to all statements
  - Fixed factorial precedence
  - Fixed network server forall loop
  - Added comments to all examples
- Updated glyph-escapes.md with string escape sequences
- Updated math_prog_lang.md with complete lexical rules
- Fixed Chinese comma in structure list
- Updated README with badges and M0 exit criteria
- Marked ISSUE_M0_BLOCKERS.md as resolved

Ready for v0.1-alpha tag and ANTLR implementation

2025-07-25 12:21:52 -04:00

3 KiB

Raw Permalink Blame History

🚀 Lock lexical & grammar spec for M0

✅ RESOLVED - All blockers have been addressed

Overview

This issue tracked the resolution of three critical blockers that must be decided before implementing the ANTLR 4 grammar for MPL. All decisions have been made and incorporated into the specification.

Blockers

1. Comment Syntax ✅

Status: RESOLVED
Decision needed by: Before lexer PR is merged
Rationale: Source files cannot compile without comment support

Options to consider:

-- till EOL (Ada/Haskell style)
/* ... */ (C style)
# till EOL (Python/Ruby style)
Support for both single-line and multi-line comments

DECISION: Use -- for single-line and {- ... -} for multi-line (Haskell-style) to align with functional paradigm

2. String Literal Rules ✅

Status: RESOLVED
Decision needed by: Before lexer PR is merged
Rationale: Required for hello-world sample

Decisions needed:

Escape sequences: Standard set (\n, \t, \\, \") + Unicode (\u{1F600})
String delimiters: Double quotes only or support for raw strings?
Multi-line string support?
String interpolation syntax (or defer to M1)?

DECISION:

Use "..." for regular strings with standard escapes (\n, \t, \\, \", \u{XXXXXX})
Add """...""" for multi-line raw strings (no escapes)
String interpolation deferred to M1

3. Path Literal Fallback ✅

Status: RESOLVED
Decision needed by: Before lexer PR is merged
Rationale: 🖫 glyph may not render on all systems

Options:

Make 🖫 required (pure Unicode approach)
Add ASCII escape like @path"..." or #path"..."
Use \path as the ASCII escape (consistent with other escapes)

DECISION: Support both 🖫"..." and \path"..." for maximum compatibility

Additional Lexical Decisions

ASCII Escape Mapping

Need complete one-to-one table for all Unicode glyphs. Current partial list:

\gamma → γ
\lam or \lambda → λ
\Rightarrow → ⇒
etc.

Number Literal Format

Decimal: 123, 123.456, 1.23e10
Hex: 0x1A2B
Binary: 0b1101
Suffixes: _i32, _f64, _bigint (or defer to M1?)

Exit Criteria

Once this issue is closed, we can:

Implement the lexer with confidence
Parse all example programs
Generate syntax highlighting for editors
Create the glyph-escapes.md reference

Resolution Summary

All blockers have been resolved and incorporated into the specification:

Comments: -- for single-line, {- ... -} for multi-line (nestable)
Strings: "..." with escapes, """...""" for raw multi-line
Paths: Both 🖫"..." and \path"..." supported

Completed Actions

✅ All decisions made and documented
✅ Updated math_prog_lang.md with lexical rules
✅ Updated glyph-escapes.md with string escape sequences
✅ All example files updated with proper syntax
✅ Ready to tag specification as v0.1-alpha

Assignee: @developtheweb
Labels: blocker, M0, specification, grammar

3 KiB Raw Permalink Blame History Unescape Escape