1 | //===- llvm/MC/MCAsmLexer.h - Abstract Asm Lexer Interface ------*- 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 | #ifndef LLVM_MC_MCPARSER_MCASMLEXER_H |
10 | #define LLVM_MC_MCPARSER_MCASMLEXER_H |
11 | |
12 | #include "llvm/ADT/ArrayRef.h" |
13 | #include "llvm/ADT/SmallVector.h" |
14 | #include "llvm/MC/MCAsmMacro.h" |
15 | #include <cassert> |
16 | #include <cstddef> |
17 | #include <string> |
18 | |
19 | namespace llvm { |
20 | |
21 | /// A callback class which is notified of each comment in an assembly file as |
22 | /// it is lexed. |
23 | class { |
24 | public: |
25 | virtual () = default; |
26 | |
27 | /// Callback function for when a comment is lexed. Loc is the start of the |
28 | /// comment text (excluding the comment-start marker). CommentText is the text |
29 | /// of the comment, excluding the comment start and end markers, and the |
30 | /// newline for single-line comments. |
31 | virtual void HandleComment(SMLoc Loc, StringRef ) = 0; |
32 | }; |
33 | |
34 | |
35 | /// Generic assembler lexer interface, for use by target specific assembly |
36 | /// lexers. |
37 | class MCAsmLexer { |
38 | /// The current token, stored in the base class for faster access. |
39 | SmallVector<AsmToken, 1> CurTok; |
40 | |
41 | /// The location and description of the current error |
42 | SMLoc ErrLoc; |
43 | std::string Err; |
44 | |
45 | protected: // Can only create subclasses. |
46 | const char *TokStart = nullptr; |
47 | bool SkipSpace = true; |
48 | bool AllowAtInIdentifier = false; |
49 | bool AllowHashInIdentifier = false; |
50 | bool IsAtStartOfStatement = true; |
51 | bool LexMasmHexFloats = false; |
52 | bool LexMasmIntegers = false; |
53 | bool LexMasmStrings = false; |
54 | bool LexMotorolaIntegers = false; |
55 | bool UseMasmDefaultRadix = false; |
56 | unsigned DefaultRadix = 10; |
57 | bool LexHLASMIntegers = false; |
58 | bool LexHLASMStrings = false; |
59 | AsmCommentConsumer * = nullptr; |
60 | |
61 | MCAsmLexer(); |
62 | |
63 | virtual AsmToken LexToken() = 0; |
64 | |
65 | void SetError(SMLoc errLoc, const std::string &err) { |
66 | ErrLoc = errLoc; |
67 | Err = err; |
68 | } |
69 | |
70 | public: |
71 | MCAsmLexer(const MCAsmLexer &) = delete; |
72 | MCAsmLexer &operator=(const MCAsmLexer &) = delete; |
73 | virtual ~MCAsmLexer(); |
74 | |
75 | /// Consume the next token from the input stream and return it. |
76 | /// |
77 | /// The lexer will continuously return the end-of-file token once the end of |
78 | /// the main input file has been reached. |
79 | const AsmToken &Lex() { |
80 | assert(!CurTok.empty()); |
81 | // Mark if we parsing out a EndOfStatement. |
82 | IsAtStartOfStatement = CurTok.front().getKind() == AsmToken::EndOfStatement; |
83 | CurTok.erase(CI: CurTok.begin()); |
84 | // LexToken may generate multiple tokens via UnLex but will always return |
85 | // the first one. Place returned value at head of CurTok vector. |
86 | if (CurTok.empty()) { |
87 | AsmToken T = LexToken(); |
88 | CurTok.insert(I: CurTok.begin(), Elt: T); |
89 | } |
90 | return CurTok.front(); |
91 | } |
92 | |
93 | void UnLex(AsmToken const &Token) { |
94 | IsAtStartOfStatement = false; |
95 | CurTok.insert(I: CurTok.begin(), Elt: Token); |
96 | } |
97 | |
98 | bool isAtStartOfStatement() { return IsAtStartOfStatement; } |
99 | |
100 | virtual StringRef LexUntilEndOfStatement() = 0; |
101 | |
102 | /// Get the current source location. |
103 | SMLoc getLoc() const; |
104 | |
105 | /// Get the current (last) lexed token. |
106 | const AsmToken &getTok() const { |
107 | return CurTok[0]; |
108 | } |
109 | |
110 | /// Look ahead at the next token to be lexed. |
111 | const AsmToken peekTok(bool ShouldSkipSpace = true) { |
112 | AsmToken Tok; |
113 | |
114 | MutableArrayRef<AsmToken> Buf(Tok); |
115 | size_t ReadCount = peekTokens(Buf, ShouldSkipSpace); |
116 | |
117 | assert(ReadCount == 1); |
118 | (void)ReadCount; |
119 | |
120 | return Tok; |
121 | } |
122 | |
123 | /// Look ahead an arbitrary number of tokens. |
124 | virtual size_t peekTokens(MutableArrayRef<AsmToken> Buf, |
125 | bool ShouldSkipSpace = true) = 0; |
126 | |
127 | /// Get the current error location |
128 | SMLoc getErrLoc() { |
129 | return ErrLoc; |
130 | } |
131 | |
132 | /// Get the current error string |
133 | const std::string &getErr() { |
134 | return Err; |
135 | } |
136 | |
137 | /// Get the kind of current token. |
138 | AsmToken::TokenKind getKind() const { return getTok().getKind(); } |
139 | |
140 | /// Check if the current token has kind \p K. |
141 | bool is(AsmToken::TokenKind K) const { return getTok().is(K); } |
142 | |
143 | /// Check if the current token has kind \p K. |
144 | bool isNot(AsmToken::TokenKind K) const { return getTok().isNot(K); } |
145 | |
146 | /// Set whether spaces should be ignored by the lexer |
147 | void setSkipSpace(bool val) { SkipSpace = val; } |
148 | |
149 | bool getAllowAtInIdentifier() { return AllowAtInIdentifier; } |
150 | void setAllowAtInIdentifier(bool v) { AllowAtInIdentifier = v; } |
151 | |
152 | void setAllowHashInIdentifier(bool V) { AllowHashInIdentifier = V; } |
153 | |
154 | void (AsmCommentConsumer *) { |
155 | this->CommentConsumer = CommentConsumer; |
156 | } |
157 | |
158 | /// Set whether to lex masm-style binary (e.g., 0b1101) and radix-specified |
159 | /// literals (e.g., 0ABCh [hex], 576t [decimal], 77o [octal], 1101y [binary]). |
160 | void setLexMasmIntegers(bool V) { LexMasmIntegers = V; } |
161 | |
162 | /// Set whether to use masm-style default-radix integer literals. If disabled, |
163 | /// assume decimal unless prefixed (e.g., 0x2c [hex], 077 [octal]). |
164 | void useMasmDefaultRadix(bool V) { UseMasmDefaultRadix = V; } |
165 | |
166 | unsigned getMasmDefaultRadix() const { return DefaultRadix; } |
167 | void setMasmDefaultRadix(unsigned Radix) { DefaultRadix = Radix; } |
168 | |
169 | /// Set whether to lex masm-style hex float literals, such as 3f800000r. |
170 | void setLexMasmHexFloats(bool V) { LexMasmHexFloats = V; } |
171 | |
172 | /// Set whether to lex masm-style string literals, such as 'Can''t find file' |
173 | /// and "This ""value"" not found". |
174 | void setLexMasmStrings(bool V) { LexMasmStrings = V; } |
175 | |
176 | /// Set whether to lex Motorola-style integer literals, such as $deadbeef or |
177 | /// %01010110. |
178 | void setLexMotorolaIntegers(bool V) { LexMotorolaIntegers = V; } |
179 | |
180 | /// Set whether to lex HLASM-flavour integers. For now this is only [0-9]* |
181 | void setLexHLASMIntegers(bool V) { LexHLASMIntegers = V; } |
182 | |
183 | /// Set whether to "lex" HLASM-flavour character and string literals. For now, |
184 | /// setting this option to true, will disable lexing for character and string |
185 | /// literals. |
186 | void setLexHLASMStrings(bool V) { LexHLASMStrings = V; } |
187 | }; |
188 | |
189 | } // end namespace llvm |
190 | |
191 | #endif // LLVM_MC_MCPARSER_MCASMLEXER_H |
192 | |