From f960e39f8b6f67c23f1d67c8ccd09c2d7bd6568d Mon Sep 17 00:00:00 2001
From: Brummel <mould@mycelium.de>
Date: Sun, 10 May 2026 15:04:50 +0200
Subject: [PATCH] floats iter 2.1 fixup: drop task-tags + refresh stale prose +
 uppercase-E + leading-dot tests

---
 crates/ailang-surface/src/lex.rs | 40 ++++++++++++++++++++++++++++----
 1 file changed, 36 insertions(+), 4 deletions(-)

diff --git a/crates/ailang-surface/src/lex.rs b/crates/ailang-surface/src/lex.rs
index a67df2e..1bfd013 100644
--- a/crates/ailang-surface/src/lex.rs
+++ b/crates/ailang-surface/src/lex.rs
@@ -10,7 +10,9 @@
 //! 3. After tokenisation, classify by first character:
 //!    - `"` ⇒ string literal (consume until closing `"`; `\"` and
 //!      `\n` escapes supported).
-//!    - digit, or `-` followed by digit ⇒ integer literal.
+//!    - digit, or `-` followed by digit ⇒ integer or float literal
+//!      (float if the token contains `.`, `e`, or `E`; integer
+//!      otherwise).
 //!    - otherwise ⇒ ident.
 //!
 //! Operators (`+`, `==`, `<=`, `**`), qualified names like
@@ -40,7 +42,7 @@ pub enum Tok {
     /// Integer literal. Original textual form preserved for error messages;
     /// parsed `i64` value carried for direct use.
     Int(i64),
-    /// IEEE-754 binary64 literal (Floats milestone iter 2). The
+    /// IEEE-754 binary64 literal. The
     /// payload is the bit pattern produced by `f64::to_bits`. The
     /// surface forms recognised by the lexer are
     /// `<digits>.<digits>(e[+-]?<digits>)?` and
@@ -49,7 +51,7 @@ pub enum Tok {
     Float(u64),
     /// String literal contents (after escape processing).
     Str(String),
-    /// Identifier. Anything that isn't a paren, integer, or string.
+    /// Identifier. Anything that isn't a paren, integer, float, or string.
     Ident(String),
 }
 
@@ -347,7 +349,7 @@ mod tests {
         assert!(matches!(&toks[0].tok, Tok::Str(s) if s == "em — dash and naïve"));
     }
 
-    /// Iter 22-floats.2 RED: a `<digits>.<digits>` token lexes as
+    /// A `<digits>.<digits>` token lexes as
     /// `Tok::Float` carrying the IEEE-754 binary64 bit pattern of
     /// `f64::from_str(raw).unwrap().to_bits()`. Pinned literal:
     /// `1.5_f64` → bit pattern `0x3FF8000000000000`.
@@ -475,4 +477,34 @@ mod tests {
             "expected InvalidFloat for `1..5`, got {err:?}"
         );
     }
+
+    /// Spec A2 grammar accepts both `e` and `E` for the exponent
+    /// marker. The `looks_like_float` validator handles both bytes;
+    /// pin the uppercase form explicitly.
+    #[test]
+    fn float_uppercase_exponent() {
+        let toks = tokenize("1.5E3 2E-5").unwrap();
+        assert_eq!(toks.len(), 2);
+        match toks[0].tok {
+            Tok::Float(b) => assert_eq!(b, 1.5e3_f64.to_bits()),
+            ref other => panic!("expected Tok::Float, got {other:?}"),
+        }
+        match toks[1].tok {
+            Tok::Float(b) => assert_eq!(b, 2e-5_f64.to_bits()),
+            ref other => panic!("expected Tok::Float, got {other:?}"),
+        }
+    }
+
+    /// Spec A2 rejects bare leading dot at the *float* level — `.5` is
+    /// not a float literal. The lexer's digit-lead rule does not fire
+    /// (first byte `.` is not a digit and not `-`-followed-by-digit),
+    /// so `.5` falls through to the maximal-non-paren-non-whitespace
+    /// run and lexes as `Tok::Ident(".5")`. Downstream parser /
+    /// typecheck stages produce the appropriate diagnostic.
+    #[test]
+    fn float_leading_dot_lexes_as_ident() {
+        let toks = tokenize(".5").unwrap();
+        assert_eq!(toks.len(), 1);
+        assert!(matches!(&toks[0].tok, Tok::Ident(s) if s == ".5"));
+    }
 }