1 //===--- Encoding.h - Format C++ code -------------------------------------===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 ///
10 /// \file
11 /// \brief Contains functions for text encoding manipulation. Supports UTF-8,
12 /// 8-bit encodings and escape sequences in C++ string literals.
13 ///
14 //===----------------------------------------------------------------------===//
15 
16 #ifndef LLVM_CLANG_LIB_FORMAT_ENCODING_H
17 #define LLVM_CLANG_LIB_FORMAT_ENCODING_H
18 
19 #include "clang/Basic/LLVM.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 /// \brief 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.
detectEncoding(StringRef Text)34 inline Encoding detectEncoding(StringRef Text) {
35   const UTF8 *Ptr = reinterpret_cast<const UTF8 *>(Text.begin());
36   const UTF8 *BufEnd = reinterpret_cast<const UTF8 *>(Text.end());
37   if (::isLegalUTF8String(&Ptr, BufEnd))
38     return Encoding_UTF8;
39   return Encoding_Unknown;
40 }
41 
getCodePointCountUTF8(StringRef Text)42 inline unsigned getCodePointCountUTF8(StringRef Text) {
43   unsigned CodePoints = 0;
44   for (size_t i = 0, e = Text.size(); i < e; i += getNumBytesForUTF8(Text[i])) {
45     ++CodePoints;
46   }
47   return CodePoints;
48 }
49 
50 /// \brief Gets the number of code points in the Text using the specified
51 /// Encoding.
getCodePointCount(StringRef Text,Encoding Encoding)52 inline unsigned getCodePointCount(StringRef Text, Encoding Encoding) {
53   switch (Encoding) {
54   case Encoding_UTF8:
55     return getCodePointCountUTF8(Text);
56   default:
57     return Text.size();
58   }
59 }
60 
61 /// \brief Returns the number of columns required to display the \p Text on a
62 /// generic Unicode-capable terminal. Text is assumed to use the specified
63 /// \p Encoding.
columnWidth(StringRef Text,Encoding Encoding)64 inline unsigned columnWidth(StringRef Text, Encoding Encoding) {
65   if (Encoding == Encoding_UTF8) {
66     int ContentWidth = llvm::sys::unicode::columnWidthUTF8(Text);
67     // FIXME: Figure out the correct way to handle this in the presence of both
68     // printable and unprintable multi-byte UTF-8 characters. Falling back to
69     // returning the number of bytes may cause problems, as columnWidth suddenly
70     // becomes non-additive.
71     if (ContentWidth >= 0)
72       return ContentWidth;
73   }
74   return Text.size();
75 }
76 
77 /// \brief Returns the number of columns required to display the \p Text,
78 /// starting from the \p StartColumn on a terminal with the \p TabWidth. The
79 /// text is assumed to use the specified \p Encoding.
columnWidthWithTabs(StringRef Text,unsigned StartColumn,unsigned TabWidth,Encoding Encoding)80 inline unsigned columnWidthWithTabs(StringRef Text, unsigned StartColumn,
81                                     unsigned TabWidth, Encoding Encoding) {
82   unsigned TotalWidth = 0;
83   StringRef Tail = Text;
84   for (;;) {
85     StringRef::size_type TabPos = Tail.find('\t');
86     if (TabPos == StringRef::npos)
87       return TotalWidth + columnWidth(Tail, Encoding);
88     TotalWidth += columnWidth(Tail.substr(0, TabPos), Encoding);
89     TotalWidth += TabWidth - (TotalWidth + StartColumn) % TabWidth;
90     Tail = Tail.substr(TabPos + 1);
91   }
92 }
93 
94 /// \brief Gets the number of bytes in a sequence representing a single
95 /// codepoint and starting with FirstChar in the specified Encoding.
getCodePointNumBytes(char FirstChar,Encoding Encoding)96 inline unsigned getCodePointNumBytes(char FirstChar, Encoding Encoding) {
97   switch (Encoding) {
98   case Encoding_UTF8:
99     return getNumBytesForUTF8(FirstChar);
100   default:
101     return 1;
102   }
103 }
104 
isOctDigit(char c)105 inline bool isOctDigit(char c) { return '0' <= c && c <= '7'; }
106 
isHexDigit(char c)107 inline bool isHexDigit(char c) {
108   return ('0' <= c && c <= '9') || ('a' <= c && c <= 'f') ||
109          ('A' <= c && c <= 'F');
110 }
111 
112 /// \brief Gets the length of an escape sequence inside a C++ string literal.
113 /// Text should span from the beginning of the escape sequence (starting with a
114 /// backslash) to the end of the string literal.
getEscapeSequenceLength(StringRef Text)115 inline unsigned getEscapeSequenceLength(StringRef Text) {
116   assert(Text[0] == '\\');
117   if (Text.size() < 2)
118     return 1;
119 
120   switch (Text[1]) {
121   case 'u':
122     return 6;
123   case 'U':
124     return 10;
125   case 'x': {
126     unsigned I = 2; // Point after '\x'.
127     while (I < Text.size() && isHexDigit(Text[I]))
128       ++I;
129     return I;
130   }
131   default:
132     if (isOctDigit(Text[1])) {
133       unsigned I = 1;
134       while (I < Text.size() && I < 4 && isOctDigit(Text[I]))
135         ++I;
136       return I;
137     }
138     return 1 + getNumBytesForUTF8(Text[1]);
139   }
140 }
141 
142 } // namespace encoding
143 } // namespace format
144 } // namespace clang
145 
146 #endif
147