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 LLVM_ADT_STRINGEXTRAS_H
16
17#include <iterator>
18#include "llvm/ADT/StringRef.h"
19#include "llvm/Support/DataTypes.h"
20
21namespace llvm {
22template<typename T> class SmallVectorImpl;
23
24/// hexdigit - Return the hexadecimal character for the
25/// given number \p X (which should be less than 16).
26static inline char hexdigit(unsigned X, bool LowerCase = false) {
27  const char HexChar = LowerCase ? 'a' : 'A';
28  return X < 10 ? '0' + X : HexChar + X - 10;
29}
30
31/// Interpret the given character \p C as a hexadecimal digit and return its
32/// value.
33///
34/// If \p C is not a valid hex digit, -1U is returned.
35static inline unsigned hexDigitValue(char C) {
36  if (C >= '0' && C <= '9') return C-'0';
37  if (C >= 'a' && C <= 'f') return C-'a'+10U;
38  if (C >= 'A' && C <= 'F') return C-'A'+10U;
39  return -1U;
40}
41
42/// utohex_buffer - Emit the specified number into the buffer specified by
43/// BufferEnd, returning a pointer to the start of the string.  This can be used
44/// like this: (note that the buffer must be large enough to handle any number):
45///    char Buffer[40];
46///    printf("0x%s", utohex_buffer(X, Buffer+40));
47///
48/// This should only be used with unsigned types.
49///
50template<typename IntTy>
51static inline char *utohex_buffer(IntTy X, char *BufferEnd) {
52  char *BufPtr = BufferEnd;
53  *--BufPtr = 0;      // Null terminate buffer.
54  if (X == 0) {
55    *--BufPtr = '0';  // Handle special case.
56    return BufPtr;
57  }
58
59  while (X) {
60    unsigned char Mod = static_cast<unsigned char>(X) & 15;
61    *--BufPtr = hexdigit(Mod);
62    X >>= 4;
63  }
64  return BufPtr;
65}
66
67static inline std::string utohexstr(uint64_t X) {
68  char Buffer[17];
69  return utohex_buffer(X, Buffer+17);
70}
71
72static inline std::string utostr_32(uint32_t X, bool isNeg = false) {
73  char Buffer[11];
74  char *BufPtr = Buffer+11;
75
76  if (X == 0) *--BufPtr = '0';  // Handle special case...
77
78  while (X) {
79    *--BufPtr = '0' + char(X % 10);
80    X /= 10;
81  }
82
83  if (isNeg) *--BufPtr = '-';   // Add negative sign...
84
85  return std::string(BufPtr, Buffer+11);
86}
87
88static inline std::string utostr(uint64_t X, bool isNeg = false) {
89  char Buffer[21];
90  char *BufPtr = Buffer+21;
91
92  if (X == 0) *--BufPtr = '0';  // Handle special case...
93
94  while (X) {
95    *--BufPtr = '0' + char(X % 10);
96    X /= 10;
97  }
98
99  if (isNeg) *--BufPtr = '-';   // Add negative sign...
100  return std::string(BufPtr, Buffer+21);
101}
102
103
104static inline std::string itostr(int64_t X) {
105  if (X < 0)
106    return utostr(static_cast<uint64_t>(-X), true);
107  else
108    return utostr(static_cast<uint64_t>(X));
109}
110
111/// StrInStrNoCase - Portable version of strcasestr.  Locates the first
112/// occurrence of string 's1' in string 's2', ignoring case.  Returns
113/// the offset of s2 in s1 or npos if s2 cannot be found.
114StringRef::size_type StrInStrNoCase(StringRef s1, StringRef s2);
115
116/// getToken - This function extracts one token from source, ignoring any
117/// leading characters that appear in the Delimiters string, and ending the
118/// token at any of the characters that appear in the Delimiters string.  If
119/// there are no tokens in the source string, an empty string is returned.
120/// The function returns a pair containing the extracted token and the
121/// remaining tail string.
122std::pair<StringRef, StringRef> getToken(StringRef Source,
123                                         StringRef Delimiters = " \t\n\v\f\r");
124
125/// SplitString - Split up the specified string according to the specified
126/// delimiters, appending the result fragments to the output list.
127void SplitString(StringRef Source,
128                 SmallVectorImpl<StringRef> &OutFragments,
129                 StringRef Delimiters = " \t\n\v\f\r");
130
131/// HashString - Hash function for strings.
132///
133/// This is the Bernstein hash function.
134//
135// FIXME: Investigate whether a modified bernstein hash function performs
136// better: http://eternallyconfuzzled.com/tuts/algorithms/jsw_tut_hashing.aspx
137//   X*33+c -> X*33^c
138static inline unsigned HashString(StringRef Str, unsigned Result = 0) {
139  for (unsigned i = 0, e = Str.size(); i != e; ++i)
140    Result = Result * 33 + (unsigned char)Str[i];
141  return Result;
142}
143
144/// Returns the English suffix for an ordinal integer (-st, -nd, -rd, -th).
145static inline StringRef getOrdinalSuffix(unsigned Val) {
146  // It is critically important that we do this perfectly for
147  // user-written sequences with over 100 elements.
148  switch (Val % 100) {
149  case 11:
150  case 12:
151  case 13:
152    return "th";
153  default:
154    switch (Val % 10) {
155      case 1: return "st";
156      case 2: return "nd";
157      case 3: return "rd";
158      default: return "th";
159    }
160  }
161}
162
163template <typename IteratorT>
164inline std::string join_impl(IteratorT Begin, IteratorT End,
165                             StringRef Separator, std::input_iterator_tag) {
166  std::string S;
167  if (Begin == End)
168    return S;
169
170  S += (*Begin);
171  while (++Begin != End) {
172    S += Separator;
173    S += (*Begin);
174  }
175  return S;
176}
177
178template <typename IteratorT>
179inline std::string join_impl(IteratorT Begin, IteratorT End,
180                             StringRef Separator, std::forward_iterator_tag) {
181  std::string S;
182  if (Begin == End)
183    return S;
184
185  size_t Len = (std::distance(Begin, End) - 1) * Separator.size();
186  for (IteratorT I = Begin; I != End; ++I)
187    Len += (*Begin).size();
188  S.reserve(Len);
189  S += (*Begin);
190  while (++Begin != End) {
191    S += Separator;
192    S += (*Begin);
193  }
194  return S;
195}
196
197/// Joins the strings in the range [Begin, End), adding Separator between
198/// the elements.
199template <typename IteratorT>
200inline std::string join(IteratorT Begin, IteratorT End, StringRef Separator) {
201  typedef typename std::iterator_traits<IteratorT>::iterator_category tag;
202  return join_impl(Begin, End, Separator, tag());
203}
204
205} // End llvm namespace
206
207#endif
208