Fix dot to match anything in ascii mode (#135)

fix dot to match anything in ascii mode

Author: nitely

src/regex.nim

@@ -319,7 +319,7 @@ This flag makes ascii mode ``(?-u)`` the default.
     :test:
     let flags = {regexArbitraryBytes}
     doAssert match("\xff", re2(r"\xff", flags))
-    #doAssert match("\xf8\xa1\xa1\xa1\xa1", re2(r".+", flags))
+    doAssert match("\xf8\xa1\xa1\xa1\xa1", re2(r".+", flags))
 
 Beware of (un)expected behaviour when mixin UTF-8 characters.
 
@@ -1581,6 +1581,7 @@ when isMainModule:
     doAssert match("aΪ", re2"\w(?<=a)Ϊ")
     doAssert match("Ϊb", re2"\w(?<=Ϊ)b")
     doAssert match("弢Ⓐ", re2"\w(?<=弢)Ⓐ")
+    doAssert match("弢", re2"(?-u).+")
     block:  # Follows Nim re's behaviour
       doAssert match("abc", re2"(?<=a)bc", m, start = 1)
       doAssert(not match("abc", re2"(?<=x)bc", m, start = 1))

src/regex/exptransformation.nim

@@ -118,10 +118,6 @@ func toAsciiKind(k: NodeKind): NodeKind =
     reNotDigitAscii
   of reNotWhiteSpace:
     reNotWhiteSpaceAscii
-  of reAny:
-    reAnyAscii
-  of reAnyNL:
-    reAnyNLAscii
   else:
     k
 

src/regex/nfamacro.nim

@@ -120,12 +120,8 @@ func genMatch(c: NimNode, n: Node): NimNode =
       quote do: not `whiteSpaceMatch`
     of reAny:
       quote do: `c` != '\L'.ord
-    of reAnyAscii:
-      quote do: `c` <= 128 and `c` != '\L'.ord
     of reAnyNL:
       quote do: true
-    of reAnyNlAscii:
-      quote do: `c` <= 128
     of reCharCI:
       let cp2Lit = newLit n.cp.swapCase().int32
       quote do: `c` == `cpLit` or `c` == `cp2Lit`

src/regex/nodematch.nim

@@ -95,10 +95,6 @@ func isDigitAscii(r: Rune): bool {.inline.} =
   else:
     false
 
-func isAnyAscii(r: Rune): bool {.inline.} =
-  (r.int <= int8.high and
-   r != lineBreakRune)
-
 # todo: can not use unicodeplus due to
 # https://github.com/nim-lang/Nim/issues/7059
 func swapCase*(r: Rune): Rune =
@@ -160,10 +156,6 @@ func match*(n: Node, r: Rune): bool {.inline.} =
     not r.isWhiteSpaceAscii()
   of reNotUCC:
     r.unicodeCategory() notin n.cc
-  of reAnyAscii:
-    r.isAnyAscii()
-  of reAnyNLAscii:
-    r.isAnyAscii() or r == lineBreakRune
   else:
     assert n.kind == reChar
     n.cp == r

src/regex/types.nim

@@ -84,8 +84,6 @@ type
     reNotAlphaNumAscii,  # \W ascii only
     reNotDigitAscii,  # \D ascii only
     reNotWhiteSpaceAscii,  # \S ascii only
-    reAnyAscii,  # . ascii only
-    reAnyNlAscii,  # . new-line ascii only
     reInSet,  # [abc]
     reNotSet,  # [^abc]
     reLookahead,  # (?=...)
@@ -261,9 +259,7 @@ const
     reWhiteSpaceAscii,
     reNotAlphaNumAscii,
     reNotDigitAscii,
-    reNotWhiteSpaceAscii,
-    reAnyAscii,
-    reAnyNLAscii}
+    reNotWhiteSpaceAscii}
   repetitionKind* = {
     reZeroOrMore,
     reOneOrMore,
@@ -310,7 +306,7 @@ func `$`*(n: Node): string =
   of reNotDigit, reNotDigitAscii: r"\D"
   of reNotWhiteSpace, reNotWhiteSpaceAscii: r"\S"
   of reNotUCC: r"\PN"
-  of reAny, reAnyNl, reAnyAscii, reAnyNlAscii: "."
+  of reAny, reAnyNl: "."
   of reInSet, reNotSet:
     var str = ""
     str.add '['

tests/tests2.nim

@@ -1232,6 +1232,8 @@ test "tflags":
   check(not "Ǝ".isMatch(re2"(?-u)[\w]"))
   check(not "\t".isMatch(re2"(?-u)[\w]"))
   check "ƎƎ".isMatch(re2"(?-u)[^\w](?u)\w")
+  check isMatch("弢", re2"(?u).+")
+  check isMatch("弢", re2"(?-u).+")
 
   check "a".isMatch(re2"(?x)a")
   check "a".isMatch(re2"(?x)a ")
@@ -3156,9 +3158,9 @@ when not defined(js) or NimMajor >= 2:
     check match("abcd", re2(r"(?-su).{4}", flags))
     check match("abcd", re2(r"(?s-u).{4}", flags))
     check match("abcd", re2(r"(?u-s).{4}", flags))
-    #check match("弢", re2(r".{4}", flags))  # XXX should match
+    check match("弢", re2(r".{4}", flags))
     check match("弢", re2(r"(?u).{4}", flags))
-    check(not match("弢", re2(r"(?-u).{4}", flags)))
+    check match("弢", re2(r"(?-u).{4}", flags))
     check(not match("\n", re2(r".", flags)))
     check match("\n", re2(r"(?s).", flags))
     check(not match("\n", re2(r"(?u).", flags)))