Print codepoints in generated guards as character literals (#158)

Author: dkuku

README.md

@@ -68,13 +68,13 @@ clauses, which are shown below:
 defp datetime__0(<<x0, x1, x2, x3, "-", x4, x5, "-", x6, x7, "T",
                    x8, x9, ":", x10, x11, ":", x12, x13, rest::binary>>,
                  acc, stack, comb__context, comb__line, comb__column)
-     when x0 >= 48 and x0 <= 57 and (x1 >= 48 and x1 <= 57) and
-         (x2 >= 48 and x2 <= 57) and (x3 >= 48 and x3 <= 57) and
-         (x4 >= 48 and x4 <= 57) and (x5 >= 48 and x5 <= 57) and
-         (x6 >= 48 and x6 <= 57) and (x7 >= 48 and x7 <= 57) and
-         (x8 >= 48 and x8 <= 57) and (x9 >= 48 and x9 <= 57) and
-         (x10 >= 48 and x10 <= 57) and (x11 >= 48 and x11 <= 57) and
-         (x12 >= 48 and x12 <= 57) and (x13 >= 48 and x13 <= 57) do
+     when x0 >= ?0 and x0 <= ?9 and (x1 >= ?0 and x1 <= ?9) and
+         (x2 >= ?0 and x2 <= ?9) and (x3 >= ?0 and x3 <= ?9) and
+         (x4 >= ?0 and x4 <= ?9) and (x5 >= ?0 and x5 <= ?9) and
+         (x6 >= ?0 and x6 <= ?9) and (x7 >= ?0 and x7 <= ?9) and
+         (x8 >= ?0 and x8 <= ?9) and (x9 >= ?0 and x9 <= ?9) and
+         (x10 >= ?0 and x10 <= ?9) and (x11 >= ?0 and x11 <= ?9) and
+         (x12 >= ?0 and x12 <= ?9) and (x13 >= ?0 and x13 <= ?9) do
   datetime__1(
     rest,
     [(x13 - 48) * 1 + (x12 - 48) * 10, (x11 - 48) * 1 + (x10 - 48) * 10,

lib/nimble_parsec/compiler.ex

@@ -1165,31 +1165,48 @@ defmodule NimbleParsec.Compiler do
   defp bin_range_to_guard(var, range) do
     case range do
       min..min//step when abs(step) == 1 ->
-        quote(do: unquote(var) === unquote(min))
+        quote(do: unquote(var) === unquote(char(min)))
 
       min..max//1 ->
-        quote(do: unquote(var) >= unquote(min) and unquote(var) <= unquote(max))
+        quote(do: unquote(var) >= unquote(char(min)) and unquote(var) <= unquote(char(max)))
 
       min..max//-1 ->
-        quote(do: unquote(var) >= unquote(max) and unquote(var) <= unquote(min))
+        quote(do: unquote(var) >= unquote(char(max)) and unquote(var) <= unquote(char(min)))
 
       min when is_integer(min) ->
-        quote(do: unquote(var) === unquote(min))
+        quote(do: unquote(var) === unquote(char(min)))
 
       {:not, min..min//step} when abs(step) == 1 ->
-        quote(do: unquote(var) !== unquote(min))
+        quote(do: unquote(var) !== unquote(char(min)))
 
       {:not, min..max//1} ->
-        quote(do: unquote(var) < unquote(min) or unquote(var) > unquote(max))
+        quote(do: unquote(var) < unquote(char(min)) or unquote(var) > unquote(char(max)))
 
       {:not, min..max//-1} ->
-        quote(do: unquote(var) < unquote(max) or unquote(var) > unquote(min))
+        quote(do: unquote(var) < unquote(char(max)) or unquote(var) > unquote(char(min)))
 
       {:not, min} when is_integer(min) ->
-        quote(do: unquote(var) !== unquote(min))
+        quote(do: unquote(var) !== unquote(char(min)))
     end
   end
 
+  # `?a` and `97` are the same AST node, so the literal spelling only survives as
+  # `:token` metadata, which `Macro.to_string/1` honours when printing.
+  defp char(codepoint) do
+    token =
+      case codepoint do
+        ?\\ -> "?\\\\"
+        ?\s -> "?\\s"
+        ?\n -> "?\\n"
+        ?\t -> "?\\t"
+        ?\r -> "?\\r"
+        codepoint when codepoint in ?!..?~ -> "?" <> <<codepoint>>
+        codepoint -> Integer.to_string(codepoint)
+      end
+
+    {:__block__, [token: token], [codepoint]}
+  end
+
   defp inspect_bin_range(min..max//_, printable?) do
     {" in the range #{inspect_char(min)} to #{inspect_char(max)}",
      printable? and printable?(min) and printable?(max)}

test/nimble_parsec_test.exs

@@ -1638,6 +1638,27 @@ defmodule NimbleParsecTest do
     end
   end
 
+  describe "generated guards" do
+    test "spell codepoints out as literals" do
+      assert guard_source(ascii_char([?a..?f])) =~ "x0 >= ?a and x0 <= ?f"
+      assert guard_source(ascii_char([?\s, ?\n])) =~ "x0 === ?\\s or x0 === ?\\n"
+      assert guard_source(ascii_char([?\t, ?\r])) =~ "x0 === ?\\t or x0 === ?\\r"
+      assert guard_source(ascii_char([?\\, ?~])) =~ "x0 === ?\\\\ or x0 === ?~"
+      assert guard_source(ascii_char(not: ?q)) =~ "x0 !== ?q"
+
+      # Codepoints with no printable spelling stay as plain integers.
+      assert guard_source(ascii_char([0x00, 0x1B])) =~ "x0 === 0 or x0 === 27"
+      assert guard_source(utf8_char([?é, ?ą])) =~ "x0 === 233 or x0 === 261"
+      assert guard_source(utf8_char([0x1F600, 0x1F601])) =~ "x0 === 128_512 or x0 === 128_513"
+    end
+
+    defp guard_source(combinator) do
+      {defs, _inline} = NimbleParsec.Compiler.compile(:literals, combinator, [])
+
+      Enum.map_join(defs, "\n", fn {_name, _args, guards, _body} -> Macro.to_string(guards) end)
+    end
+  end
+
   describe "continuing parser" do
     defparsecp :digits, [?0..?9] |> ascii_char() |> times(min: 1) |> label("digits")
     defparsecp :chars, [?a..?z] |> ascii_char() |> times(min: 1) |> label("chars")