fix(lang): a template inside another's hole, and integer literals past 2^31 - 1

A template literal nested in a {...} hole crashed the parser: the outer
literal ended at the inner backtick. The lexer (and ludic-fmt's) now
reads a hole as code, taking strings, chars and templates in it whole.
A decimal literal past 2147483647, or a hex one of more than eight
digits, was wrapped into a negative int; it is a long with its value
now, and giving one to an int is refused. Reseed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

LANGUAGE.md

@@ -1764,13 +1764,19 @@ let msg = `hello {name}, you have {count + 1} messages`
 # == "hello " + name + ", you have " + str(count + 1) + " messages"
 ```
 
-`str(x)` is the same conversion on its own. Write a literal brace as `{{` / `}}`.
+`str(x)` is the same conversion on its own. Write a literal brace as `{{` / `}}`. A hole is code,
+so a string, a char or another template literal inside it is taken whole - `` `a {wrap(`b {n}`)} c` ``
+is one literal, and a brace or a backtick inside a string in a hole is text.
 
 **Slicing.** `s[a..b]` is a fresh substring of the bytes `[a, b)`, and `len(s)`
 is a string's byte length — so `path[0..len(path) - 6]` trims an extension and
 `s[i]` still indexes a single byte. `expr with { field: … }`
 is not implemented; records appear only in `spawn`. Char literals (`'w'`) are
-`int` code points; colors are hex ints (`0xff8800`). `null` is the null-pointer
+`int` code points; colors are hex ints (`0xff8800`). An integer literal past `2147483647` is a
+`long` and keeps its value (`let mask: long = 4294967295`); given to an `int` it is refused (`n wants
+an int and 4294967295 does not fit one; it is a long`). A hex literal of up to eight digits is a
+32-bit pattern - `0xFFFFFFFF` is `-1`, and `0xEDB88320` given to a `long` is negative - and one of
+more digits is a `long` (`0x10000000000`). `null` is the null-pointer
 literal; test any pointer/record/slice with `x == null` / `x != null` (an unset
 `Node`/`ptr` field reads back as `null`).
 

changes/lexer-literals.md

@@ -0,0 +1,8 @@
+bump: patch
+type: fix
+**A template literal inside another's `{…}` hole, and integer literals past 2^31 - 1.** The first
+crashed the compiler (the outer literal ended at the inner one's backtick); a hole is now read as
+code, so strings, chars and templates inside it are taken whole. The second was read as a wrapped
+negative int; a decimal literal past `2147483647`, or a hex one of more than eight digits, is now a
+`long` with its value, and giving one to an `int` is refused. Eight hex digits or fewer are still a
+32-bit pattern.

examples/lang/long_literals.ludic

@@ -0,0 +1,16 @@
+# long_literals.ludic — a decimal literal past 2^31 - 1 is a long and keeps its value; so is a hex
+# literal of more than eight digits (past 2^63 it wraps, as the bits say). Eight hex digits or
+# fewer stay a 32-bit pattern (0xFFFFFFFF is -1, as it always was).
+#
+# Running it prints:  4294967295 3988292384 -4294967296 1099511627776 -3750763034362895579 -1 1
+program LongLiterals {
+  const MASK: long = 4294967295
+  var poly: long = 3988292384
+  function neg() -> long { return -4294967296 }
+  entry {
+    let big = 0x10000000000
+    let fnv = 0xcbf29ce484222325
+    let small = 0xFFFFFFFF
+    print(`{MASK} {poly} {neg()} {big} {fnv} {small} {MASK > long(2147483647)}`)
+  }
+}

examples/lang/nested_templates.ludic

@@ -0,0 +1,12 @@
+# nested_templates.ludic — a template literal inside another's hole is one expression: the outer
+# literal ends at its own backtick, not at the inner one's. A brace or a backtick inside a string in
+# a hole is text, and {{ }} are still literal braces.
+#
+# Running it prints:  outer [inner 3 {x}] "`}" end
+program NestedTemplates {
+  function wrap(s: string) -> string { return `[{s}]` }
+  entry {
+    let n = 3
+    print(`outer {wrap(`inner {n} {{x}}`)} {"\"`}\""} end`)
+  }
+}

examples/rejected/long_into_int.ludic

@@ -0,0 +1,7 @@
+# a literal past 2^31 - 1 is a long, and a long does not narrow into an int
+program LongIntoInt {
+  entry {
+    let n: int = 4294967295
+    print(n)
+  }
+}

selfhost/backend/emit_call.ludic

@@ -985,7 +985,10 @@ function emit_call(e: Node) -> Val {
 
 function emit_expr(e: Node) -> Val {
   if (e == null) { return val("0", "int") }
-  if e.kind == E_INT  { return val(itoa(e.ival), "int") }
+  if e.kind == E_INT  {
+    if e.s != null { return val(e.s, "long") }     # a literal past 2^31 - 1
+    return val(itoa(e.ival), "int")
+  }
   if e.kind == E_FLOAT and is_float_file(e.file) {
     let ff = val(fp_lit_code(e.s, "float"), "float")
     ff.lit = e

selfhost/backend/emit_core.ludic

@@ -360,7 +360,10 @@ function static_type(e: Node) -> pointer {
     if (et != null) { return "[]" + et }
   }
   if e.kind == E_STR or e.kind == E_SLICE { return "string" }
-  if e.kind == E_INT { return "int" }
+  if e.kind == E_INT {
+    if e.s != null { return "long" }
+    return "int"
+  }
   if e.kind == E_FLOAT and is_float_file(e.file) { return "float" }
   if e.kind == E_FLOAT { return "fixed" }
   if e.kind == E_BOOL { return "bool" }

selfhost/backend/emit_float.ludic

@@ -39,7 +39,7 @@ function fp_result(a: pointer, b: pointer) -> pointer {
 # a literal-only expression (Val.lit) evaluated exactly in float type t
 function fp_const(e: Node, t: pointer) -> pointer {
   if e.kind == E_FLOAT { return fp_lit_code(e.s, t) }
-  if e.kind == E_INT { return fp_lit_code(itoa(e.ival), t) }
+  if e.kind == E_INT { return fp_lit_code(int_lit_code(e), t) }
   if e.kind == E_UN {
     let x = fp_const(e.a, t)
     return emit_bind(`fneg {t} {x}`)

selfhost/backend/emit_head.ludic

@@ -30,12 +30,18 @@ function emit_escaped(s: pointer) -> void {
   }
 }
 
+# an int literal's LLVM constant: its value, or a long literal's own digits
+function int_lit_code(e: Node) -> pointer {
+  if e.s != null { return e.s }
+  return itoa(e.ival)
+}
 # the constant initializer for a global var: a literal, or 0/null
 function global_init(d: Node) -> pointer {
   if (d.a == null) or is_fp(d.ty) { return zero_of(llty(d.ty)) }   # a float's value is set by L_init_globals
   let e = d.a
-  if e.kind == E_INT or e.kind == E_FLOAT or e.kind == E_BOOL { return itoa(e.ival) }
+  if e.kind == E_INT { return int_lit_code(e) }
-  if e.kind == E_UN and (e.s == ("-")) and e.a.kind == E_INT { return (("-") + itoa(e.a.ival)) }
+  if e.kind == E_FLOAT or e.kind == E_BOOL { return itoa(e.ival) }
+  if e.kind == E_UN and (e.s == ("-")) and e.a.kind == E_INT { return (("-") + int_lit_code(e.a)) }
   if e.kind == E_MEMBER and e.a.kind == E_ID {                  # `Enum.Variant` is a compile-time int
     let ord = enum_ordinal(e.a.s, e.s)
     if ord >= 0 { return itoa(ord) }

selfhost/check/check_env.ludic

@@ -51,6 +51,11 @@ function ck_err(cat: pointer, n: Node, msg: pointer) -> void {
 }
 # a value given where a type is wanted: `what` says where ("argument 2 of f", "x")
 function ck_give(to: pointer, from: pointer, e: Node, what: pointer) -> void {
+  # a literal past 2^31 - 1 given to an int would lose its value: say so, rather than wrap it
+  if e != null and e.kind == E_INT and e.s != null and ((to == "int") or (to == "byte") or (to == "fixed")) {
+    ck_err("literal", e, `{what} wants {ck_a(to)} and {e.s} does not fit one; it is a long`)
+    return
+  }
   let cat = ck_mismatch(to, from, e)
   if cat == null { return }
   if (cat == "slice-pointer") {

selfhost/check/check_expr.ludic

@@ -32,7 +32,10 @@ function ck_expr(e: Node) -> pointer {
   let ex = ck_expect
   ck_expect = null
   let k = e.kind
-  if k == E_INT { return "int" }
+  if k == E_INT {
+    if e.s != null { return "long" }              # a literal past 2^31 - 1
+    return "int"
+  }
   if k == E_FLOAT {
     if is_float_file(e.file) { return "float" }
     return "fixed"

selfhost/frontend/lex.ludic

@@ -60,6 +60,41 @@ function lex_error(line: int, c: int) -> void {
   lex_fail(line, `unexpected character '{ch}' (byte {c})`)
 }
 
+# the closing backtick of the template literal that opens at s[i]: a hole's `{...}` is code, so a
+# string, a char or another template inside it is skipped whole - `a {f(`b {c}`)} d` is one literal
+function tmpl_end(s: pointer, i0: int, n: int) -> int {
+  var i = i0 + 1
+  while i < n and s[i] != '`' {
+    if s[i] == CH_BACKSLASH { i += 2;  continue }
+    if s[i] == '{' and i + 1 < n and s[i + 1] == '{' { i += 2;  continue }
+    if s[i] == '{' {
+      i = hole_end(s, i + 1, n)
+    }
+    i += 1
+  }
+  return i
+}
+# the `}` that closes a hole whose code starts at s[i]; n when there is none
+function hole_end(s: pointer, i0: int, n: int) -> int {
+  var i = i0
+  var depth = 1
+  while i < n {
+    let d = s[i]
+    if d == '"' or d == CH_SQUOTE {
+      i += 1
+      while i < n and s[i] != d { if s[i] == CH_BACKSLASH { i += 1 };  i += 1 }
+    }
+    else if d == '`' { i = tmpl_end(s, i, n) }
+    else if d == '{' { depth += 1 }
+    else if d == '}' {
+      depth -= 1
+      if depth == 0 { return i }
+    }
+    i += 1
+  }
+  return n
+}
+
 # report a lexical error in the standard `file:line: error: msg` shape and stop
 function lex_fail(line: int, msg: pointer) -> void {
   let m = `{g_parse_file}:{line}: error: {msg}\n`
@@ -100,11 +135,16 @@ function lex_at(src: pointer, first_line: int) -> void {
       continue
     }
     if c == '`' {                                   # `interpolated string` — captured raw
-      i += 1
+      let e = tmpl_end(src, i, n)
       let out = bytes(n)
       var j = 0
-      while i < n and src[i] != '`' { out[j] = src[i]; j += 1; i += 1 }
       i += 1
+      while i < e {
+        out[j] = src[i]
+        j += 1
+        i += 1
+      }
+      i = e + 1
       out[j] = 0
       tok_push(TK_INTERP, out, 0, line)
       continue
@@ -122,6 +162,8 @@ function lex_at(src: pointer, first_line: int) -> void {
     if char_is_digit(c) {
       if c == '0' and src[i + 1] == 'x' {     # 0x hex
         var v = 0
+        var lv: long = 0
+        var digits = 0
         i += 2
         while i < n {
           let h = src[i]
@@ -130,14 +172,24 @@ function lex_at(src: pointer, first_line: int) -> void {
           else { if h >= 'a' and h <= 'f' { d = h - 87 }
                  else { if h >= 'A' and h <= 'F' { d = h - 55 } else { break } } }
           v = v * 16 + d
+          lv = lv * long(16) + long(d)
+          if digits > 0 or d > 0 { digits += 1 }
           i += 1
         }
-        tok_push(TK_INT, null, v, line)
+        # up to 8 hex digits is a 32-bit pattern (0xFFFFFFFF is -1, as it always was); more is a long
+        var big: pointer = null
+        if digits > 8 { big = string(lv) }
+        tok_push(TK_INT, big, v, line)
         continue
       }
       var v = 0
+      var lv: long = 0
       let nstart = i
-      while i < n and char_is_digit(src[i]) { v = v * 10 + (src[i] - 48); i += 1 }
+      while i < n and char_is_digit(src[i]) {
+        v = v * 10 + (src[i] - 48)
+        lv = lv * long(10) + long(src[i] - 48)
+        i += 1
+      }
       # a fractional part makes it a Q16.16 fixed literal (its text is kept: in a
       # float context the literal is exactly that decimal instead)
       if i < n and src[i] == '.' and char_is_digit(src[i + 1]) {
@@ -154,7 +206,10 @@ function lex_at(src: pointer, first_line: int) -> void {
         tok_push(TK_FLOAT, src[nstart..i], bits, line)
         continue
       }
-      tok_push(TK_INT, null, v, line)
+      # a decimal past 2^31 - 1 does not fit an int: it is a long, and keeps its value as text
+      var big: pointer = null
+      if lv > long(2147483647) { big = string(lv) }
+      tok_push(TK_INT, big, v, line)
       continue
     }
     if char_is_alpha(c) {

selfhost/frontend/parse.ludic

@@ -199,19 +199,9 @@ function parse_interp(raw: pointer) -> Node {
       if lj > 0 { acc = interp_add(acc, interp_lit(lit, lj));  lj = 0 }
       i += 1
       let hs = i
-      # find the hole's closing brace: braces nest, and a brace inside a
+      # find the hole's closing brace: braces nest, and a brace inside a "string", a 'char' or a
-      # "string" or 'char' literal in the hole is text, not structure
+      # `template` in the hole is text, not structure (lex.ludic, hole_end)
-      var depth = 1
+      i = hole_end(raw, i, n)
-      while i < n and depth > 0 {
-        let d = raw[i]
-        if d == '"' or d == CH_SQUOTE {
-          i += 1
-          while i < n and raw[i] != d { if raw[i] == CH_BACKSLASH { i += 1 };  i += 1 }
-        }
-        else if d == '{' { depth += 1 }
-        else if d == '}' { depth -= 1;  if depth == 0 { break } }
-        i += 1
-      }
       acc = interp_add(acc, interp_str(parse_hole(raw[hs..i])))
       i += 1                                   # skip the closing '}'
     } else {
@@ -259,7 +249,7 @@ function p_primary() -> Node {
   if t.kind == TK_INTERP { pi += 1; g_interp_line = t.line; return parse_interp(t.text) }
   if is_op("[") { return parse_list() }
   if t.kind == TK_ID and (t.text == "emit") and (toks[pi + 1].kind == TK_ID) { return parse_emit() }
-  if t.kind == TK_INT { let n = node(E_INT); n.ival = t.ival; pi += 1; return n }
+  if t.kind == TK_INT { let n = node(E_INT); n.ival = t.ival; n.s = t.text; pi += 1; return n }   # s: a long literal's digits
   if t.kind == TK_FLOAT { let n = node(E_FLOAT); n.ival = t.ival; n.s = t.text; pi += 1; return n }
   if t.kind == TK_STR { let n = node(E_STR); n.s = t.text; pi += 1; return n }
   if t.kind == TK_ID {

tools/ludic-cli/test.ludic

@@ -835,6 +835,9 @@ function cmd_dev_test() -> int {
   reject_case("rejected/generic_mismatch", "words wants a Pool<string> and this is a Pool<int>", "two instances of one generic are two types")
   reject_case("rejected/generic_arity", "Pool takes 1 type argument(s) and Pool<int, string> gives 2", "a generic takes as many type arguments as it has parameters")
   reject_case("rejected/private_generic", "grow is private to module kit", "an instance of a private generic is private to its module")
+  feat_case("lang/long_literals", "", "4294967295 3988292384 -4294967296 1099511627776 -3750763034362895579 -1 1", "long_literals.ludic (a decimal literal past 2^31 - 1, and a hex literal of more than eight digits, is a long)")
+  reject_case("rejected/long_into_int", "4294967295 does not fit one; it is a long", "a long literal given to an int is refused, not wrapped")
+  feat_case("lang/nested_templates", "", "outer [inner 3 {x}] \"`}\" end", "nested_templates.ludic (a template literal inside another's hole)")
   feat_case("lang/aliases", "", "10 5 3", "aliases.ludic (L6: a namespace method declared as an alias of a function, labelled or by its parameters)")
   reject_case("rejected/alias_arity", "this call to Trail.length leaves out to, which has no default", "an alias's arguments are checked against its target")
   reject_case("rejected/alias_twice", "Trail.length is declared twice", "a namespace method is declared once")

tools/ludic-tools/fmt.ludic

@@ -136,6 +136,38 @@ program LudicFmt {
     return c == '&' or c == '|' or c == '^' or c == '~'                                                        # & | ^ ~
   }
 
+
+  # the backtick that closes the template opening at s[i0] (or the end): a `{...}` hole is code, so
+  # a string, a char or another template inside it is skipped whole - as the compiler reads it
+  function fmt_tmpl_end(s: pointer, i0: int) -> int {
+    var i = i0 + 1
+    while s[i] != 0 and s[i] != '`' {
+      if s[i] == '\\' and s[i + 1] != 0 { i += 2;  continue }
+      if s[i] == '{' and s[i + 1] == '{' { i += 2;  continue }
+      if s[i] == '{' { i = fmt_hole_end(s, i + 1) }
+      if s[i] != 0 { i += 1 }
+    }
+    return i
+  }
+  function fmt_hole_end(s: pointer, i0: int) -> int {
+    var i = i0
+    var depth = 1
+    while s[i] != 0 {
+      let d = s[i]
+      if d == '"' or d == '\'' {
+        i += 1
+        while s[i] != 0 and s[i] != d { if s[i] == '\\' and s[i + 1] != 0 { i += 1 };  i += 1 }
+      }
+      else if d == '`' { i = fmt_tmpl_end(s, i) }
+      else if d == '{' { depth += 1 }
+      else if d == '}' {
+        depth -= 1
+        if depth == 0 { return i }
+      }
+      if s[i] != 0 { i += 1 }
+    }
+    return i
+  }
   function lex(s: pointer) -> void {
     src = s
     tk_kind = new []int;  tk_start = new []int;  tk_end = new []int;  tk_line = new []int
@@ -155,9 +187,9 @@ program LudicFmt {
         if s[i] == '"' { i += 1 }
         push_tok(LT_STR, st, i, line);  continue
       }
-      if c == '`' {                                    # `interpolated`
+      if c == '`' {                                    # `interpolated`, holes and all
-        let st = i;  i += 1
+        let st = i
-        while s[i] != 0 and s[i] != '`' { if s[i] == '\\' and s[i + 1] != 0 { i += 2 } else { i += 1 } }
+        i = fmt_tmpl_end(s, i)
         if s[i] == '`' { i += 1 }
         push_tok(LT_STR, st, i, line);  continue
       }