1 //===- llvm/ADT/StringExtras.h - Useful string functions --------*- C++ -*-===//
3 // The LLVM Compiler Infrastructure
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
10 // This file contains some functions that are useful when dealing with strings.
12 //===----------------------------------------------------------------------===//
14 #ifndef LLVM_ADT_STRINGEXTRAS_H
15 #define LLVM_ADT_STRINGEXTRAS_H
17 #include "llvm/ADT/ArrayRef.h"
18 #include "llvm/ADT/SmallString.h"
19 #include "llvm/ADT/StringRef.h"
31 template<typename T> class SmallVectorImpl;
34 /// hexdigit - Return the hexadecimal character for the
35 /// given number \p X (which should be less than 16).
36 static inline char hexdigit(unsigned X, bool LowerCase = false) {
37 const char HexChar = LowerCase ? 'a' : 'A';
38 return X < 10 ? '0' + X : HexChar + X - 10;
41 /// Construct a string ref from a boolean.
42 static inline StringRef toStringRef(bool B) {
43 return StringRef(B ? "true" : "false");
46 /// Construct a string ref from an array ref of unsigned chars.
47 static inline StringRef toStringRef(ArrayRef<uint8_t> Input) {
48 return StringRef(reinterpret_cast<const char *>(Input.begin()), Input.size());
51 /// Interpret the given character \p C as a hexadecimal digit and return its
54 /// If \p C is not a valid hex digit, -1U is returned.
55 static inline unsigned hexDigitValue(char C) {
56 if (C >= '0' && C <= '9') return C-'0';
57 if (C >= 'a' && C <= 'f') return C-'a'+10U;
58 if (C >= 'A' && C <= 'F') return C-'A'+10U;
62 static inline std::string utohexstr(uint64_t X, bool LowerCase = false) {
64 char *BufPtr = std::end(Buffer);
66 if (X == 0) *--BufPtr = '0';
69 unsigned char Mod = static_cast<unsigned char>(X) & 15;
70 *--BufPtr = hexdigit(Mod, LowerCase);
74 return std::string(BufPtr, std::end(Buffer));
77 /// Convert buffer \p Input to its hexadecimal representation.
78 /// The returned string is double the size of \p Input.
79 inline std::string toHex(StringRef Input) {
80 static const char *const LUT = "0123456789ABCDEF";
81 size_t Length = Input.size();
84 Output.reserve(2 * Length);
85 for (size_t i = 0; i < Length; ++i) {
86 const unsigned char c = Input[i];
87 Output.push_back(LUT[c >> 4]);
88 Output.push_back(LUT[c & 15]);
93 inline std::string toHex(ArrayRef<uint8_t> Input) {
94 return toHex(toStringRef(Input));
97 static inline uint8_t hexFromNibbles(char MSB, char LSB) {
98 unsigned U1 = hexDigitValue(MSB);
99 unsigned U2 = hexDigitValue(LSB);
100 assert(U1 != -1U && U2 != -1U);
102 return static_cast<uint8_t>((U1 << 4) | U2);
105 /// Convert hexadecimal string \p Input to its binary representation.
106 /// The return string is half the size of \p Input.
107 static inline std::string fromHex(StringRef Input) {
109 return std::string();
112 Output.reserve((Input.size() + 1) / 2);
113 if (Input.size() % 2 == 1) {
114 Output.push_back(hexFromNibbles('0', Input.front()));
115 Input = Input.drop_front();
118 assert(Input.size() % 2 == 0);
119 while (!Input.empty()) {
120 uint8_t Hex = hexFromNibbles(Input[0], Input[1]);
121 Output.push_back(Hex);
122 Input = Input.drop_front(2);
127 /// \brief Convert the string \p S to an integer of the specified type using
128 /// the radix \p Base. If \p Base is 0, auto-detects the radix.
129 /// Returns true if the number was successfully converted, false otherwise.
130 template <typename N> bool to_integer(StringRef S, N &Num, unsigned Base = 0) {
131 return !S.getAsInteger(Base, Num);
135 template <typename N>
136 inline bool to_float(const Twine &T, N &Num, N (*StrTo)(const char *, char **)) {
137 SmallString<32> Storage;
138 StringRef S = T.toNullTerminatedStringRef(Storage);
140 N Temp = StrTo(S.data(), &End);
148 inline bool to_float(const Twine &T, float &Num) {
149 return detail::to_float(T, Num, strtof);
152 inline bool to_float(const Twine &T, double &Num) {
153 return detail::to_float(T, Num, strtod);
156 inline bool to_float(const Twine &T, long double &Num) {
157 return detail::to_float(T, Num, strtold);
160 static inline std::string utostr(uint64_t X, bool isNeg = false) {
162 char *BufPtr = std::end(Buffer);
164 if (X == 0) *--BufPtr = '0'; // Handle special case...
167 *--BufPtr = '0' + char(X % 10);
171 if (isNeg) *--BufPtr = '-'; // Add negative sign...
172 return std::string(BufPtr, std::end(Buffer));
175 static inline std::string itostr(int64_t X) {
177 return utostr(static_cast<uint64_t>(-X), true);
179 return utostr(static_cast<uint64_t>(X));
182 /// StrInStrNoCase - Portable version of strcasestr. Locates the first
183 /// occurrence of string 's1' in string 's2', ignoring case. Returns
184 /// the offset of s2 in s1 or npos if s2 cannot be found.
185 StringRef::size_type StrInStrNoCase(StringRef s1, StringRef s2);
187 /// getToken - This function extracts one token from source, ignoring any
188 /// leading characters that appear in the Delimiters string, and ending the
189 /// token at any of the characters that appear in the Delimiters string. If
190 /// there are no tokens in the source string, an empty string is returned.
191 /// The function returns a pair containing the extracted token and the
192 /// remaining tail string.
193 std::pair<StringRef, StringRef> getToken(StringRef Source,
194 StringRef Delimiters = " \t\n\v\f\r");
196 /// SplitString - Split up the specified string according to the specified
197 /// delimiters, appending the result fragments to the output list.
198 void SplitString(StringRef Source,
199 SmallVectorImpl<StringRef> &OutFragments,
200 StringRef Delimiters = " \t\n\v\f\r");
202 /// HashString - Hash function for strings.
204 /// This is the Bernstein hash function.
206 // FIXME: Investigate whether a modified bernstein hash function performs
207 // better: http://eternallyconfuzzled.com/tuts/algorithms/jsw_tut_hashing.aspx
209 static inline unsigned HashString(StringRef Str, unsigned Result = 0) {
210 for (StringRef::size_type i = 0, e = Str.size(); i != e; ++i)
211 Result = Result * 33 + (unsigned char)Str[i];
215 /// Returns the English suffix for an ordinal integer (-st, -nd, -rd, -th).
216 static inline StringRef getOrdinalSuffix(unsigned Val) {
217 // It is critically important that we do this perfectly for
218 // user-written sequences with over 100 elements.
229 default: return "th";
234 /// PrintEscapedString - Print each character of the specified string, escaping
235 /// it if it is not printable or if it is an escape char.
236 void PrintEscapedString(StringRef Name, raw_ostream &Out);
240 template <typename IteratorT>
241 inline std::string join_impl(IteratorT Begin, IteratorT End,
242 StringRef Separator, std::input_iterator_tag) {
248 while (++Begin != End) {
255 template <typename IteratorT>
256 inline std::string join_impl(IteratorT Begin, IteratorT End,
257 StringRef Separator, std::forward_iterator_tag) {
262 size_t Len = (std::distance(Begin, End) - 1) * Separator.size();
263 for (IteratorT I = Begin; I != End; ++I)
264 Len += (*Begin).size();
267 while (++Begin != End) {
274 template <typename Sep>
275 inline void join_items_impl(std::string &Result, Sep Separator) {}
277 template <typename Sep, typename Arg>
278 inline void join_items_impl(std::string &Result, Sep Separator,
283 template <typename Sep, typename Arg1, typename... Args>
284 inline void join_items_impl(std::string &Result, Sep Separator, const Arg1 &A1,
288 join_items_impl(Result, Separator, std::forward<Args>(Items)...);
291 inline size_t join_one_item_size(char C) { return 1; }
292 inline size_t join_one_item_size(const char *S) { return S ? ::strlen(S) : 0; }
294 template <typename T> inline size_t join_one_item_size(const T &Str) {
298 inline size_t join_items_size() { return 0; }
300 template <typename A1> inline size_t join_items_size(const A1 &A) {
301 return join_one_item_size(A);
303 template <typename A1, typename... Args>
304 inline size_t join_items_size(const A1 &A, Args &&... Items) {
305 return join_one_item_size(A) + join_items_size(std::forward<Args>(Items)...);
308 } // end namespace detail
310 /// Joins the strings in the range [Begin, End), adding Separator between
312 template <typename IteratorT>
313 inline std::string join(IteratorT Begin, IteratorT End, StringRef Separator) {
314 using tag = typename std::iterator_traits<IteratorT>::iterator_category;
315 return detail::join_impl(Begin, End, Separator, tag());
318 /// Joins the strings in the range [R.begin(), R.end()), adding Separator
319 /// between the elements.
320 template <typename Range>
321 inline std::string join(Range &&R, StringRef Separator) {
322 return join(R.begin(), R.end(), Separator);
325 /// Joins the strings in the parameter pack \p Items, adding \p Separator
326 /// between the elements. All arguments must be implicitly convertible to
327 /// std::string, or there should be an overload of std::string::operator+=()
328 /// that accepts the argument explicitly.
329 template <typename Sep, typename... Args>
330 inline std::string join_items(Sep Separator, Args &&... Items) {
332 if (sizeof...(Items) == 0)
335 size_t NS = detail::join_one_item_size(Separator);
336 size_t NI = detail::join_items_size(std::forward<Args>(Items)...);
337 Result.reserve(NI + (sizeof...(Items) - 1) * NS + 1);
338 detail::join_items_impl(Result, Separator, std::forward<Args>(Items)...);
342 } // end namespace llvm
344 #endif // LLVM_ADT_STRINGEXTRAS_H