1 //===--- Encoding.h - Format C++ code ---------------------------*- C++ -*-===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 ///
9 /// \file
10 /// Contains functions for text encoding manipulation. Supports UTF-8,
11 /// 8-bit encodings and escape sequences in C++ string literals.
12 ///
13 //===----------------------------------------------------------------------===//
14 
15 #ifndef LLVM_CLANG_LIB_FORMAT_ENCODING_H
16 #define LLVM_CLANG_LIB_FORMAT_ENCODING_H
17 
18 #include "clang/Basic/LLVM.h"
19 #include "llvm/ADT/StringRef.h"
20 #include "llvm/Support/ConvertUTF.h"
21 #include "llvm/Support/Unicode.h"
22 
23 namespace clang {
24 namespace format {
25 namespace encoding {
26 
27 enum Encoding {
28   Encoding_UTF8,
29   Encoding_Unknown // We treat all other encodings as 8-bit encodings.
30 };
31 
32 /// Detects encoding of the Text. If the Text can be decoded using UTF-8,
33 /// it is considered UTF8, otherwise we treat it as some 8-bit encoding.
34 inline Encoding detectEncoding(StringRef Text) {
35   const llvm::UTF8 *Ptr = reinterpret_cast<const llvm::UTF8 *>(Text.begin());
36   const llvm::UTF8 *BufEnd = reinterpret_cast<const llvm::UTF8 *>(Text.end());
37   if (llvm::isLegalUTF8String(&Ptr, BufEnd))
38     return Encoding_UTF8;
39   return Encoding_Unknown;
40 }
41 
42 /// Returns the number of columns required to display the \p Text on a
43 /// generic Unicode-capable terminal. Text is assumed to use the specified
44 /// \p Encoding.
45 inline unsigned columnWidth(StringRef Text, Encoding Encoding) {
46   if (Encoding == Encoding_UTF8) {
47     int ContentWidth = llvm::sys::unicode::columnWidthUTF8(Text);
48     // FIXME: Figure out the correct way to handle this in the presence of both
49     // printable and unprintable multi-byte UTF-8 characters. Falling back to
50     // returning the number of bytes may cause problems, as columnWidth suddenly
51     // becomes non-additive.
52     if (ContentWidth >= 0)
53       return ContentWidth;
54   }
55   return Text.size();
56 }
57 
58 /// Returns the number of columns required to display the \p Text,
59 /// starting from the \p StartColumn on a terminal with the \p TabWidth. The
60 /// text is assumed to use the specified \p Encoding.
61 inline unsigned columnWidthWithTabs(StringRef Text, unsigned StartColumn,
62                                     unsigned TabWidth, Encoding Encoding) {
63   unsigned TotalWidth = 0;
64   StringRef Tail = Text;
65   for (;;) {
66     StringRef::size_type TabPos = Tail.find('\t');
67     if (TabPos == StringRef::npos)
68       return TotalWidth + columnWidth(Tail, Encoding);
69     TotalWidth += columnWidth(Tail.substr(0, TabPos), Encoding);
70     if (TabWidth)
71       TotalWidth += TabWidth - (TotalWidth + StartColumn) % TabWidth;
72     Tail = Tail.substr(TabPos + 1);
73   }
74 }
75 
76 /// Gets the number of bytes in a sequence representing a single
77 /// codepoint and starting with FirstChar in the specified Encoding.
78 inline unsigned getCodePointNumBytes(char FirstChar, Encoding Encoding) {
79   switch (Encoding) {
80   case Encoding_UTF8:
81     return llvm::getNumBytesForUTF8(FirstChar);
82   default:
83     return 1;
84   }
85 }
86 
87 inline bool isOctDigit(char c) { return '0' <= c && c <= '7'; }
88 
89 inline bool isHexDigit(char c) {
90   return ('0' <= c && c <= '9') || ('a' <= c && c <= 'f') ||
91          ('A' <= c && c <= 'F');
92 }
93 
94 /// Gets the length of an escape sequence inside a C++ string literal.
95 /// Text should span from the beginning of the escape sequence (starting with a
96 /// backslash) to the end of the string literal.
97 inline unsigned getEscapeSequenceLength(StringRef Text) {
98   assert(Text[0] == '\\');
99   if (Text.size() < 2)
100     return 1;
101 
102   switch (Text[1]) {
103   case 'u':
104     return 6;
105   case 'U':
106     return 10;
107   case 'x': {
108     unsigned I = 2; // Point after '\x'.
109     while (I < Text.size() && isHexDigit(Text[I]))
110       ++I;
111     return I;
112   }
113   default:
114     if (isOctDigit(Text[1])) {
115       unsigned I = 1;
116       while (I < Text.size() && I < 4 && isOctDigit(Text[I]))
117         ++I;
118       return I;
119     }
120     return 1 + llvm::getNumBytesForUTF8(Text[1]);
121   }
122 }
123 
124 } // namespace encoding
125 } // namespace format
126 } // namespace clang
127 
128 #endif
129