UCNs handling - proposal from subcommision WG14/N842 Antoine Leca 1998-06-24, revised 1998-06-26 Reference text: N837 Reference text: N828 Editings: In 5.1.1.2p1, point 1 in the list, delete the sentence about replacement of multibyte source file character by universal character names. Delete the footnote 6 as well. Delete the "Then" of the next sentence. In point 2, insert the following sentence as the 3rd sentence: "If, as a result, a character sequence that matches the syntax of an universal character name is produced, the behavior is undefined." In 5.2.1p3, last sentence, add "identifiers" to the enumeration "(except in..." In 5.2.1.2p2, add "identifier" to both enumeration. Replace 6.4.2.1p1 with identifier: non-basic-digit identifier non-basic-digit identifier basic-digit non-basic-digit: basic-letter universal-character-name any multibyte character from the extended set basic-letter: _ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z a b c d e f g h i j k l m n o p q r s t u v w x y z basic-digit: 0 1 2 3 4 5 6 7 8 9 In 6.4.2.1p2, replace "specified using universal character names" with "and other characters specified using either members of the extended source character set or universal character name". Add a sentence just after the closing parenthesis: "The relationship between universal character names and members of the extended set is implementation defined." In the next sentence, replace "universal character name" by "extended source character and universal character name". Next, the reference to ISO10646 ought to be "ISO/IEC 10646". Next sentence, replace "nondigit" with "non basic digit". In the forward references: remove the references to 6.2.2 In 6.4.3p2, first sentence, replace from "range ..." to the end of the end of the sentence with "ranges 00000000 to 00000020, 0000007F to 0000009F and 0000D800 to 0000DFFF inclusive." Paragraph 3, delete the word "source". Paragraph 4, the reference to ISO10646 ought to be "ISO/IEC 10646". In 6.4.4.2p1, replace digit-sequence syntax with digit-sequence: basic-digit digit-sequence basic-digit In 6.4.4.4p1, remove the universal-character-name production from c-char syntax. Add universal-character-name to escape-sequence. Replace paragraph 8 with: "In addition, members outside of the basic execution character set can be designated using the universal character name notation of escape sequences. Certain nongraphic characters are representable by escape sequences consisting of the backslash \ followed by a lowercase letter: \a, \b, \f, \n, \r, \t and \v. 54" In 6.4.5p1, remove the universal-character-name production from s-char syntax. In 6.4.8p1, replace "digit" with "basic-digit" (3 occurences). Replace "nondigit" with "non-basic-digit". Remove 6.10p4, and the attached footnote 124 as well. In 6.10.3.2p2, before "If the replacement...", insert the following sentence "If there are any universal character name notation that appears in a character constant or string litteral argument, the behavior is unspecified." In 7.20.1.5p3 (and 7.24.4.1.1), replace "nondigit" with "basic-letter", and replace "digit" with "basic-digit" in the syntax of n-char-sequence (2 occurences each).