1 | //===- llvm/ADT/StringExtras.h - Useful string functions --------*- C++ -*-===// |
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 | // This file contains some functions that are useful when dealing with strings. |
11 | // |
12 | //===----------------------------------------------------------------------===// |
13 | |
14 | #ifndef LLVM_ADT_STRINGEXTRAS_H |
15 | #define |
16 | |
17 | #include "llvm/ADT/ArrayRef.h" |
18 | #include "llvm/ADT/SmallString.h" |
19 | #include "llvm/ADT/StringRef.h" |
20 | #include "llvm/ADT/Twine.h" |
21 | #include <cassert> |
22 | #include <cstddef> |
23 | #include <cstdint> |
24 | #include <cstdlib> |
25 | #include <cstring> |
26 | #include <iterator> |
27 | #include <string> |
28 | #include <utility> |
29 | |
30 | namespace llvm { |
31 | |
32 | template<typename T> class SmallVectorImpl; |
33 | class raw_ostream; |
34 | |
35 | /// hexdigit - Return the hexadecimal character for the |
36 | /// given number \p X (which should be less than 16). |
37 | inline char hexdigit(unsigned X, bool LowerCase = false) { |
38 | const char HexChar = LowerCase ? 'a' : 'A'; |
39 | return X < 10 ? '0' + X : HexChar + X - 10; |
40 | } |
41 | |
42 | /// Given an array of c-style strings terminated by a null pointer, construct |
43 | /// a vector of StringRefs representing the same strings without the terminating |
44 | /// null string. |
45 | inline std::vector<StringRef> toStringRefArray(const char *const *Strings) { |
46 | std::vector<StringRef> Result; |
47 | while (*Strings) |
48 | Result.push_back(*Strings++); |
49 | return Result; |
50 | } |
51 | |
52 | /// Construct a string ref from a boolean. |
53 | inline StringRef toStringRef(bool B) { return StringRef(B ? "true" : "false" ); } |
54 | |
55 | /// Construct a string ref from an array ref of unsigned chars. |
56 | inline StringRef toStringRef(ArrayRef<uint8_t> Input) { |
57 | return StringRef(reinterpret_cast<const char *>(Input.begin()), Input.size()); |
58 | } |
59 | |
60 | /// Construct a string ref from an array ref of unsigned chars. |
61 | inline ArrayRef<uint8_t> arrayRefFromStringRef(StringRef Input) { |
62 | return {Input.bytes_begin(), Input.bytes_end()}; |
63 | } |
64 | |
65 | /// Interpret the given character \p C as a hexadecimal digit and return its |
66 | /// value. |
67 | /// |
68 | /// If \p C is not a valid hex digit, -1U is returned. |
69 | inline unsigned hexDigitValue(char C) { |
70 | if (C >= '0' && C <= '9') return C-'0'; |
71 | if (C >= 'a' && C <= 'f') return C-'a'+10U; |
72 | if (C >= 'A' && C <= 'F') return C-'A'+10U; |
73 | return -1U; |
74 | } |
75 | |
76 | /// Checks if character \p C is one of the 10 decimal digits. |
77 | inline bool isDigit(char C) { return C >= '0' && C <= '9'; } |
78 | |
79 | /// Checks if character \p C is a hexadecimal numeric character. |
80 | inline bool isHexDigit(char C) { return hexDigitValue(C) != -1U; } |
81 | |
82 | /// Checks if character \p C is a valid letter as classified by "C" locale. |
83 | inline bool isAlpha(char C) { |
84 | return ('a' <= C && C <= 'z') || ('A' <= C && C <= 'Z'); |
85 | } |
86 | |
87 | /// Checks whether character \p C is either a decimal digit or an uppercase or |
88 | /// lowercase letter as classified by "C" locale. |
89 | inline bool isAlnum(char C) { return isAlpha(C) || isDigit(C); } |
90 | |
91 | /// Checks whether character \p C is valid ASCII (high bit is zero). |
92 | inline bool isASCII(char C) { return static_cast<unsigned char>(C) <= 127; } |
93 | |
94 | /// Checks whether all characters in S are ASCII. |
95 | inline bool isASCII(llvm::StringRef S) { |
96 | for (char C : S) |
97 | if (LLVM_UNLIKELY(!isASCII(C))) |
98 | return false; |
99 | return true; |
100 | } |
101 | |
102 | /// Checks whether character \p C is printable. |
103 | /// |
104 | /// Locale-independent version of the C standard library isprint whose results |
105 | /// may differ on different platforms. |
106 | inline bool isPrint(char C) { |
107 | unsigned char UC = static_cast<unsigned char>(C); |
108 | return (0x20 <= UC) && (UC <= 0x7E); |
109 | } |
110 | |
111 | /// Returns the corresponding lowercase character if \p x is uppercase. |
112 | inline char toLower(char x) { |
113 | if (x >= 'A' && x <= 'Z') |
114 | return x - 'A' + 'a'; |
115 | return x; |
116 | } |
117 | |
118 | /// Returns the corresponding uppercase character if \p x is lowercase. |
119 | inline char toUpper(char x) { |
120 | if (x >= 'a' && x <= 'z') |
121 | return x - 'a' + 'A'; |
122 | return x; |
123 | } |
124 | |
125 | inline std::string utohexstr(uint64_t X, bool LowerCase = false) { |
126 | char Buffer[17]; |
127 | char *BufPtr = std::end(Buffer); |
128 | |
129 | if (X == 0) *--BufPtr = '0'; |
130 | |
131 | while (X) { |
132 | unsigned char Mod = static_cast<unsigned char>(X) & 15; |
133 | *--BufPtr = hexdigit(Mod, LowerCase); |
134 | X >>= 4; |
135 | } |
136 | |
137 | return std::string(BufPtr, std::end(Buffer)); |
138 | } |
139 | |
140 | /// Convert buffer \p Input to its hexadecimal representation. |
141 | /// The returned string is double the size of \p Input. |
142 | inline std::string toHex(StringRef Input, bool LowerCase = false) { |
143 | static const char *const LUT = "0123456789ABCDEF" ; |
144 | const uint8_t Offset = LowerCase ? 32 : 0; |
145 | size_t Length = Input.size(); |
146 | |
147 | std::string Output; |
148 | Output.reserve(2 * Length); |
149 | for (size_t i = 0; i < Length; ++i) { |
150 | const unsigned char c = Input[i]; |
151 | Output.push_back(LUT[c >> 4] | Offset); |
152 | Output.push_back(LUT[c & 15] | Offset); |
153 | } |
154 | return Output; |
155 | } |
156 | |
157 | inline std::string toHex(ArrayRef<uint8_t> Input, bool LowerCase = false) { |
158 | return toHex(toStringRef(Input), LowerCase); |
159 | } |
160 | |
161 | inline uint8_t hexFromNibbles(char MSB, char LSB) { |
162 | unsigned U1 = hexDigitValue(MSB); |
163 | unsigned U2 = hexDigitValue(LSB); |
164 | assert(U1 != -1U && U2 != -1U); |
165 | |
166 | return static_cast<uint8_t>((U1 << 4) | U2); |
167 | } |
168 | |
169 | /// Convert hexadecimal string \p Input to its binary representation. |
170 | /// The return string is half the size of \p Input. |
171 | inline std::string fromHex(StringRef Input) { |
172 | if (Input.empty()) |
173 | return std::string(); |
174 | |
175 | std::string Output; |
176 | Output.reserve((Input.size() + 1) / 2); |
177 | if (Input.size() % 2 == 1) { |
178 | Output.push_back(hexFromNibbles('0', Input.front())); |
179 | Input = Input.drop_front(); |
180 | } |
181 | |
182 | assert(Input.size() % 2 == 0); |
183 | while (!Input.empty()) { |
184 | uint8_t Hex = hexFromNibbles(Input[0], Input[1]); |
185 | Output.push_back(Hex); |
186 | Input = Input.drop_front(2); |
187 | } |
188 | return Output; |
189 | } |
190 | |
191 | /// Convert the string \p S to an integer of the specified type using |
192 | /// the radix \p Base. If \p Base is 0, auto-detects the radix. |
193 | /// Returns true if the number was successfully converted, false otherwise. |
194 | template <typename N> bool to_integer(StringRef S, N &Num, unsigned Base = 0) { |
195 | return !S.getAsInteger(Base, Num); |
196 | } |
197 | |
198 | namespace detail { |
199 | template <typename N> |
200 | inline bool to_float(const Twine &T, N &Num, N (*StrTo)(const char *, char **)) { |
201 | SmallString<32> Storage; |
202 | StringRef S = T.toNullTerminatedStringRef(Storage); |
203 | char *End; |
204 | N Temp = StrTo(S.data(), &End); |
205 | if (*End != '\0') |
206 | return false; |
207 | Num = Temp; |
208 | return true; |
209 | } |
210 | } |
211 | |
212 | inline bool to_float(const Twine &T, float &Num) { |
213 | return detail::to_float(T, Num, strtof); |
214 | } |
215 | |
216 | inline bool to_float(const Twine &T, double &Num) { |
217 | return detail::to_float(T, Num, strtod); |
218 | } |
219 | |
220 | inline bool to_float(const Twine &T, long double &Num) { |
221 | return detail::to_float(T, Num, strtold); |
222 | } |
223 | |
224 | inline std::string utostr(uint64_t X, bool isNeg = false) { |
225 | char Buffer[21]; |
226 | char *BufPtr = std::end(Buffer); |
227 | |
228 | if (X == 0) *--BufPtr = '0'; // Handle special case... |
229 | |
230 | while (X) { |
231 | *--BufPtr = '0' + char(X % 10); |
232 | X /= 10; |
233 | } |
234 | |
235 | if (isNeg) *--BufPtr = '-'; // Add negative sign... |
236 | return std::string(BufPtr, std::end(Buffer)); |
237 | } |
238 | |
239 | inline std::string itostr(int64_t X) { |
240 | if (X < 0) |
241 | return utostr(static_cast<uint64_t>(-X), true); |
242 | else |
243 | return utostr(static_cast<uint64_t>(X)); |
244 | } |
245 | |
246 | /// StrInStrNoCase - Portable version of strcasestr. Locates the first |
247 | /// occurrence of string 's1' in string 's2', ignoring case. Returns |
248 | /// the offset of s2 in s1 or npos if s2 cannot be found. |
249 | StringRef::size_type StrInStrNoCase(StringRef s1, StringRef s2); |
250 | |
251 | /// getToken - This function extracts one token from source, ignoring any |
252 | /// leading characters that appear in the Delimiters string, and ending the |
253 | /// token at any of the characters that appear in the Delimiters string. If |
254 | /// there are no tokens in the source string, an empty string is returned. |
255 | /// The function returns a pair containing the extracted token and the |
256 | /// remaining tail string. |
257 | std::pair<StringRef, StringRef> getToken(StringRef Source, |
258 | StringRef Delimiters = " \t\n\v\f\r" ); |
259 | |
260 | /// SplitString - Split up the specified string according to the specified |
261 | /// delimiters, appending the result fragments to the output list. |
262 | void SplitString(StringRef Source, |
263 | SmallVectorImpl<StringRef> &OutFragments, |
264 | StringRef Delimiters = " \t\n\v\f\r" ); |
265 | |
266 | /// Returns the English suffix for an ordinal integer (-st, -nd, -rd, -th). |
267 | inline StringRef getOrdinalSuffix(unsigned Val) { |
268 | // It is critically important that we do this perfectly for |
269 | // user-written sequences with over 100 elements. |
270 | switch (Val % 100) { |
271 | case 11: |
272 | case 12: |
273 | case 13: |
274 | return "th" ; |
275 | default: |
276 | switch (Val % 10) { |
277 | case 1: return "st" ; |
278 | case 2: return "nd" ; |
279 | case 3: return "rd" ; |
280 | default: return "th" ; |
281 | } |
282 | } |
283 | } |
284 | |
285 | /// Print each character of the specified string, escaping it if it is not |
286 | /// printable or if it is an escape char. |
287 | void printEscapedString(StringRef Name, raw_ostream &Out); |
288 | |
289 | /// Print each character of the specified string, escaping HTML special |
290 | /// characters. |
291 | void printHTMLEscaped(StringRef String, raw_ostream &Out); |
292 | |
293 | /// printLowerCase - Print each character as lowercase if it is uppercase. |
294 | void printLowerCase(StringRef String, raw_ostream &Out); |
295 | |
296 | namespace detail { |
297 | |
298 | template <typename IteratorT> |
299 | inline std::string join_impl(IteratorT Begin, IteratorT End, |
300 | StringRef Separator, std::input_iterator_tag) { |
301 | std::string S; |
302 | if (Begin == End) |
303 | return S; |
304 | |
305 | S += (*Begin); |
306 | while (++Begin != End) { |
307 | S += Separator; |
308 | S += (*Begin); |
309 | } |
310 | return S; |
311 | } |
312 | |
313 | template <typename IteratorT> |
314 | inline std::string join_impl(IteratorT Begin, IteratorT End, |
315 | StringRef Separator, std::forward_iterator_tag) { |
316 | std::string S; |
317 | if (Begin == End) |
318 | return S; |
319 | |
320 | size_t Len = (std::distance(Begin, End) - 1) * Separator.size(); |
321 | for (IteratorT I = Begin; I != End; ++I) |
322 | Len += (*Begin).size(); |
323 | S.reserve(Len); |
324 | S += (*Begin); |
325 | while (++Begin != End) { |
326 | S += Separator; |
327 | S += (*Begin); |
328 | } |
329 | return S; |
330 | } |
331 | |
332 | template <typename Sep> |
333 | inline void join_items_impl(std::string &Result, Sep Separator) {} |
334 | |
335 | template <typename Sep, typename Arg> |
336 | inline void join_items_impl(std::string &Result, Sep Separator, |
337 | const Arg &Item) { |
338 | Result += Item; |
339 | } |
340 | |
341 | template <typename Sep, typename Arg1, typename... Args> |
342 | inline void join_items_impl(std::string &Result, Sep Separator, const Arg1 &A1, |
343 | Args &&... Items) { |
344 | Result += A1; |
345 | Result += Separator; |
346 | join_items_impl(Result, Separator, std::forward<Args>(Items)...); |
347 | } |
348 | |
349 | inline size_t join_one_item_size(char C) { return 1; } |
350 | inline size_t join_one_item_size(const char *S) { return S ? ::strlen(S) : 0; } |
351 | |
352 | template <typename T> inline size_t join_one_item_size(const T &Str) { |
353 | return Str.size(); |
354 | } |
355 | |
356 | inline size_t join_items_size() { return 0; } |
357 | |
358 | template <typename A1> inline size_t join_items_size(const A1 &A) { |
359 | return join_one_item_size(A); |
360 | } |
361 | template <typename A1, typename... Args> |
362 | inline size_t join_items_size(const A1 &A, Args &&... Items) { |
363 | return join_one_item_size(A) + join_items_size(std::forward<Args>(Items)...); |
364 | } |
365 | |
366 | } // end namespace detail |
367 | |
368 | /// Joins the strings in the range [Begin, End), adding Separator between |
369 | /// the elements. |
370 | template <typename IteratorT> |
371 | inline std::string join(IteratorT Begin, IteratorT End, StringRef Separator) { |
372 | using tag = typename std::iterator_traits<IteratorT>::iterator_category; |
373 | return detail::join_impl(Begin, End, Separator, tag()); |
374 | } |
375 | |
376 | /// Joins the strings in the range [R.begin(), R.end()), adding Separator |
377 | /// between the elements. |
378 | template <typename Range> |
379 | inline std::string join(Range &&R, StringRef Separator) { |
380 | return join(R.begin(), R.end(), Separator); |
381 | } |
382 | |
383 | /// Joins the strings in the parameter pack \p Items, adding \p Separator |
384 | /// between the elements. All arguments must be implicitly convertible to |
385 | /// std::string, or there should be an overload of std::string::operator+=() |
386 | /// that accepts the argument explicitly. |
387 | template <typename Sep, typename... Args> |
388 | inline std::string join_items(Sep Separator, Args &&... Items) { |
389 | std::string Result; |
390 | if (sizeof...(Items) == 0) |
391 | return Result; |
392 | |
393 | size_t NS = detail::join_one_item_size(Separator); |
394 | size_t NI = detail::join_items_size(std::forward<Args>(Items)...); |
395 | Result.reserve(NI + (sizeof...(Items) - 1) * NS + 1); |
396 | detail::join_items_impl(Result, Separator, std::forward<Args>(Items)...); |
397 | return Result; |
398 | } |
399 | |
400 | } // end namespace llvm |
401 | |
402 | #endif // LLVM_ADT_STRINGEXTRAS_H |
403 | |