Adjust emphasis parser for CJK characters.

Typst documentation says that `*` strong emphasis
"only works at word boundaries." However, in typst/typst#2648
this was changed for CJK.

Closes #49.

src/Typst/Parse.hs

@@ -588,11 +588,50 @@ pText :: P Markup
 pText = Text . mconcat <$> some
   ((do xs <- some alphaNum
        T.pack . (xs <>) <$>
-             try (some (char '*' <|> char '_') <* lookAhead alphaNum)
+             try (some (char '*' <|> char '_') <* lookAhead (satisfy nonCJKAlphaNum))
         <|> pure (T.pack xs))
  <|> (T.pack <$> some (satisfy (\c -> not (isSpace c || isSpecial c))))
   )
 
+nonCJKAlphaNum :: Char -> Bool
+nonCJKAlphaNum c = isAlphaNum c && not (isCJK c)
+
+isCJK :: Char -> Bool
+isCJK c | c < '\x2e80' = False
+isCJK c =
+          (c >= '\x2e80' && c <= '\x2eff') -- CJK Radicals Supplement
+       || (c >= '\x2f00' && c <= '\x2fdf') -- Kangxi Radicals
+       || (c >= '\x2ff0' && c <= '\x2fff') -- Ideographic Description Characters
+       || (c >= '\x3000' && c <= '\x303f') -- JK Symbols and Punctuation
+       || (c >= '\x3040' && c <= '\x309f') -- Hiragana
+       || (c >= '\x30a0' && c <= '\x30ff') -- Katakana
+       || (c >= '\x3100' && c <= '\x312f') -- Bopomofo
+       || (c >= '\x3130' && c <= '\x318f') -- Kanbun
+       || (c >= '\x3190' && c <= '\x319f') -- Kanbun
+       || (c >= '\x31c0' && c <= '\x31ef') -- CJK Strokes
+       || (c >= '\x31f0' && c <= '\x31ff') -- Katakana Phonetic Extensions
+       || (c >= '\x3200' && c <= '\x32ff') -- Enclosed CJK Letters & Months
+       || (c >= '\x3300' && c <= '\x33ff') -- CJK Compatibility
+       || (c >= '\x3400' && c <= '\x4dbf') -- CJK Unified Ideographs Extension A
+       || (c >= '\x4e00' && c <= '\x9fff') -- CJK Unified Ideographs
+       || (c >= '\xa000' && c <= '\xa48f') -- Yi Syllables
+       || (c >= '\xa490' && c <= '\xa4cf') -- Yi Radicals
+       || (c >= '\xf900' && c <= '\xfaff') -- CJK Compatibility Ideographs
+       || (c >= '\xfe10' && c <= '\xfe1f') -- Vertical forms
+       || (c >= '\xfe30' && c <= '\xfe4f') -- CJK Compatibility Forms
+       || (c >= '\xFE50' && c <= '\xFE6F') -- Small Form Variants
+       || (c >= '\xFF00' && c <= '\xFFEE') -- Halfwidth and Fullwidth Forms
+       || (c >= '\x1B000' && c <= '\x1B0FF') -- Kana Supplement
+       || (c >= '\x1B100' && c <= '\x1B12F') -- Kana Extended-A
+       || (c >= '\x1B130' && c <= '\x1B16F') -- Small Kana Extension
+       || (c >= '\x20000' && c <= '\x2A6DF') -- CJK Unified Ideographs Extension B
+       || (c >= '\x2A700' && c <= '\x2B73F') -- CJK Unified Ideographs Extension C
+       || (c >= '\x2B740' && c <= '\x2B81F') -- CJK Unified Ideographs Extension D
+       || (c >= '\x2B820' && c <= '\x2CEAF') -- CJK Unified Ideographs Extension E
+       || (c >= '\x2CEB0' && c <= '\x2EBEF') -- CJK Unified Ideographs Extension F
+       || (c >= '\x2F800' && c <= '\x2FA1F') -- CJK Compatibility Ideographs Supp
+       || (c >= '\x30000' && c <= '\x3134F') -- CJK Unified Ideographs Exten
+
 pEscaped :: P Markup
 pEscaped = Text . T.singleton <$> pEsc
 

test/out/regression/issue49.out

@@ -0,0 +1,52 @@
+--- parse tree ---
+[ Code
+    "test/typ/regression/issue49.typ"
+    ( line 1 , column 2 )
+    (Let
+       (BasicBind (Just (Identifier "test")))
+       (FuncExpr
+          [ NormalParam (Identifier "x") , NormalParam (Identifier "y") ]
+          (Block
+             (CodeBlock
+                [ If
+                    [ ( Equals (Ident (Identifier "x")) (Ident (Identifier "y"))
+                      , Block (Content [ Text "\9989" ])
+                      )
+                    , ( Literal (Boolean True)
+                      , Block
+                          (Content
+                             [ Text "\10060"
+                             , Text "("
+                             , Code
+                                 "test/typ/regression/issue49.typ"
+                                 ( line 1 , column 47 )
+                                 (FuncCall
+                                    (Ident (Identifier "repr"))
+                                    [ NormalArg (Ident (Identifier "x")) ])
+                             , Space
+                             , Text "/"
+                             , Text "="
+                             , Space
+                             , Code
+                                 "test/typ/regression/issue49.typ"
+                                 ( line 1 , column 59 )
+                                 (FuncCall
+                                    (Ident (Identifier "repr"))
+                                    [ NormalArg (Ident (Identifier "y")) ])
+                             , Text ")"
+                             ])
+                      )
+                    ]
+                ]))))
+, SoftBreak
+, Text "\27979\35797\25991\26412"
+, Strong [ Text "\21152\31895" ]
+, Text "\12290"
+, ParBreak
+]
+--- evaluated ---
+document(body: { text(body: [
+测试文本]), 
+                 strong(body: text(body: [加粗])), 
+                 text(body: [。]), 
+                 parbreak() })

test/typ/regression/issue49.typ

@@ -0,0 +1 @@
+测试文本*加粗*。