Warn if we find a Unicode homoglyph for a symbol in an identifier.

Specifically, warn if:
 * we find a character that the language standard says we must treat as an
   identifier, and
 * that character is not reasonably an identifier character (it's a punctuation
   character or similar), and 
 * it renders identically to a valid non-identifier character in common
   fixed-width fonts.

Some tools "helpfully" substitute the surprising characters for the expected
characters, and replacing semicolons with Greek question marks is a common
"prank".

llvm-svn: 320697
This commit is contained in:
Richard Smith 2017-12-14 13:15:08 +00:00
parent bf484aa0f1
commit 77091b167f
3 changed files with 86 additions and 1 deletions
clang
include/clang/Basic
lib/Lex
test/Lexer

View File

@ -119,6 +119,9 @@ def err_non_ascii : Error<
def ext_unicode_whitespace : ExtWarn<
"treating Unicode character as whitespace">,
InGroup<DiagGroup<"unicode-whitespace">>;
def warn_utf8_symbol_homoglyph : Warning<
"treating Unicode character <U+%0> as identifier character rather than "
"as '%1' symbol">, InGroup<DiagGroup<"unicode-homoglyph">>;
def err_hex_escape_no_digits : Error<
"\\%0 used with no following hex digits">;

View File

@ -37,6 +37,7 @@
#include "llvm/Support/ConvertUTF.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Support/MemoryBuffer.h"
#include "llvm/Support/NativeFormatting.h"
#include "llvm/Support/UnicodeCharRanges.h"
#include <algorithm>
#include <cassert>
@ -1500,6 +1501,75 @@ static void maybeDiagnoseIDCharCompat(DiagnosticsEngine &Diags, uint32_t C,
}
}
/// After encountering UTF-8 character C and interpreting it as an identifier
/// character, check whether it's a homoglyph for a common non-identifier
/// source character that is unlikely to be an intentional identifier
/// character and warn if so.
static void maybeDiagnoseUTF8Homoglyph(DiagnosticsEngine &Diags, uint32_t C,
CharSourceRange Range) {
// FIXME: Handle Unicode quotation marks (smart quotes, fullwidth quotes).
struct HomoglyphPair {
uint32_t Character;
char LooksLike;
bool operator<(HomoglyphPair R) const { return Character < R.Character; }
};
static constexpr HomoglyphPair SortedHomoglyphs[] = {
{U'\u01c3', '!'}, // LATIN LETTER RETROFLEX CLICK
{U'\u037e', ';'}, // GREEK QUESTION MARK
{U'\u2212', '-'}, // MINUS SIGN
{U'\u2215', '/'}, // DIVISION SLASH
{U'\u2216', '\\'}, // SET MINUS
{U'\u2217', '*'}, // ASTERISK OPERATOR
{U'\u2223', '|'}, // DIVIDES
{U'\u2227', '^'}, // LOGICAL AND
{U'\u2236', ':'}, // RATIO
{U'\u223c', '~'}, // TILDE OPERATOR
{U'\ua789', ':'}, // MODIFIER LETTER COLON
{U'\uff01', '!'}, // FULLWIDTH EXCLAMATION MARK
{U'\uff03', '#'}, // FULLWIDTH NUMBER SIGN
{U'\uff04', '$'}, // FULLWIDTH DOLLAR SIGN
{U'\uff05', '%'}, // FULLWIDTH PERCENT SIGN
{U'\uff06', '&'}, // FULLWIDTH AMPERSAND
{U'\uff08', '('}, // FULLWIDTH LEFT PARENTHESIS
{U'\uff09', ')'}, // FULLWIDTH RIGHT PARENTHESIS
{U'\uff0a', '*'}, // FULLWIDTH ASTERISK
{U'\uff0b', '+'}, // FULLWIDTH ASTERISK
{U'\uff0c', ','}, // FULLWIDTH COMMA
{U'\uff0d', '-'}, // FULLWIDTH HYPHEN-MINUS
{U'\uff0e', '.'}, // FULLWIDTH FULL STOP
{U'\uff0f', '/'}, // FULLWIDTH SOLIDUS
{U'\uff1a', ':'}, // FULLWIDTH COLON
{U'\uff1b', ';'}, // FULLWIDTH SEMICOLON
{U'\uff1c', '<'}, // FULLWIDTH LESS-THAN SIGN
{U'\uff1d', '='}, // FULLWIDTH EQUALS SIGN
{U'\uff1e', '>'}, // FULLWIDTH GREATER-THAN SIGN
{U'\uff1f', '?'}, // FULLWIDTH QUESTION MARK
{U'\uff20', '@'}, // FULLWIDTH COMMERCIAL AT
{U'\uff3b', '['}, // FULLWIDTH LEFT SQUARE BRACKET
{U'\uff3c', '\\'}, // FULLWIDTH REVERSE SOLIDUS
{U'\uff3d', ']'}, // FULLWIDTH RIGHT SQUARE BRACKET
{U'\uff3e', '^'}, // FULLWIDTH CIRCUMFLEX ACCENT
{U'\uff5b', '{'}, // FULLWIDTH LEFT CURLY BRACKET
{U'\uff5c', '|'}, // FULLWIDTH VERTICAL LINE
{U'\uff5d', '}'}, // FULLWIDTH RIGHT CURLY BRACKET
{U'\uff5e', '~'}, // FULLWIDTH TILDE
{0, 0}
};
auto Homoglyph =
std::lower_bound(std::begin(SortedHomoglyphs),
std::end(SortedHomoglyphs) - 1, HomoglyphPair{C, '\0'});
if (Homoglyph->Character == C) {
llvm::SmallString<5> CharBuf;
{
llvm::raw_svector_ostream CharOS(CharBuf);
llvm::write_hex(CharOS, C, llvm::HexPrintStyle::Upper, 4);
}
const char LooksLikeStr[] = {Homoglyph->LooksLike, 0};
Diags.Report(Range.getBegin(), diag::warn_utf8_symbol_homoglyph)
<< Range << CharBuf << LooksLikeStr;
}
}
bool Lexer::tryConsumeIdentifierUCN(const char *&CurPtr, unsigned Size,
Token &Result) {
const char *UCNPtr = CurPtr + Size;
@ -1534,10 +1604,13 @@ bool Lexer::tryConsumeIdentifierUTF8Char(const char *&CurPtr) {
!isAllowedIDChar(static_cast<uint32_t>(CodePoint), LangOpts))
return false;
if (!isLexingRawMode())
if (!isLexingRawMode()) {
maybeDiagnoseIDCharCompat(PP->getDiagnostics(), CodePoint,
makeCharRange(*this, CurPtr, UnicodePtr),
/*IsFirst=*/false);
maybeDiagnoseUTF8Homoglyph(PP->getDiagnostics(), CodePoint,
makeCharRange(*this, CurPtr, UnicodePtr));
}
CurPtr = UnicodePtr;
return true;
@ -3737,6 +3810,7 @@ LexNextToken:
// We can't just reset CurPtr to BufferPtr because BufferPtr may point to
// an escaped newline.
--CurPtr;
const char *UTF8StartPtr = CurPtr;
llvm::ConversionResult Status =
llvm::convertUTF8Sequence((const llvm::UTF8 **)&CurPtr,
(const llvm::UTF8 *)BufferEnd,
@ -3751,6 +3825,9 @@ LexNextToken:
// (We manually eliminate the tail call to avoid recursion.)
goto LexNextToken;
}
if (!isLexingRawMode())
maybeDiagnoseUTF8Homoglyph(PP->getDiagnostics(), CodePoint,
makeCharRange(*this, UTF8StartPtr, CurPtr));
return LexUnicode(Result, CodePoint, CurPtr);
}

View File

@ -33,3 +33,8 @@ int main () {
int 🌷 = 🌵(🌹);
return 🌷;
}
int n; = 3; // expected-warning {{treating Unicode character <U+037E> as identifier character rather than as ';' symbol}}
int *nv = &n;; // expected-warning 2{{treating Unicode character <U+A789> as identifier character rather than as ':' symbol}}
// expected-warning@-1 {{treating Unicode character <U+037E> as identifier character rather than as ';' symbol}}
int vautoreturnx; // expected-warning 12{{treating Unicode character}}