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