StringExtras.cpp revision 7462b5dd99163785480a11a7c999569b349a137c
1//===-- StringExtras.cpp - Implement the StringExtras header --------------===//
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 implements the StringExtras.h header
11//
12//===----------------------------------------------------------------------===//
13
14#include "llvm/ADT/SmallVector.h"
15#include "llvm/ADT/STLExtras.h"
16#include "llvm/ADT/StringExtras.h"
17using namespace llvm;
18
19/// StrInStrNoCase - Portable version of strcasestr.  Locates the first
20/// occurrence of string 's1' in string 's2', ignoring case.  Returns
21/// the offset of s2 in s1 or npos if s2 cannot be found.
22StringRef::size_type llvm::StrInStrNoCase(StringRef s1, StringRef s2) {
23  size_t N = s2.size(), M = s1.size();
24  if (N > M)
25    return StringRef::npos;
26  for (size_t i = 0, e = M - N + 1; i != e; ++i)
27    if (s1.substr(i, N).equals_lower(s2))
28      return i;
29  return StringRef::npos;
30}
31
32/// getToken - This function extracts one token from source, ignoring any
33/// leading characters that appear in the Delimiters string, and ending the
34/// token at any of the characters that appear in the Delimiters string.  If
35/// there are no tokens in the source string, an empty string is returned.
36/// The function returns a pair containing the extracted token and the
37/// remaining tail string.
38std::pair<StringRef, StringRef> llvm::getToken(StringRef Source,
39                                               StringRef Delimiters) {
40  // Figure out where the token starts.
41  StringRef::size_type Start = Source.find_first_not_of(Delimiters);
42  if (Start == StringRef::npos) Start = Source.size();
43
44  // Find the next occurrence of the delimiter.
45  StringRef::size_type End = Source.find_first_of(Delimiters, Start);
46  if (End == StringRef::npos) End = Source.size();
47
48  return std::make_pair(Source.substr(Start, End), Source.substr(End));
49}
50
51/// SplitString - Split up the specified string according to the specified
52/// delimiters, appending the result fragments to the output list.
53void llvm::SplitString(StringRef Source,
54                       SmallVectorImpl<StringRef> &OutFragments,
55                       StringRef Delimiters) {
56  StringRef S2, S;
57  tie(S2, S) = getToken(Source, Delimiters);
58  while (!S2.empty()) {
59    OutFragments.push_back(S2);
60    tie(S2, S) = getToken(S, Delimiters);
61  }
62}
63
64// FIXME: remove when llvm-gcc doesn't use this anymore
65void llvm::SplitString(StringRef Source,
66                       std::vector<std::string> &OutFragments,
67                       StringRef Delimiters) {
68  StringRef S2, S;
69  tie(S2, S) = getToken(Source, Delimiters);
70  while (!S2.empty()) {
71    OutFragments.push_back(S2);
72    tie(S2, S) = getToken(S, Delimiters);
73  }
74}
75
76void llvm::StringRef::split(SmallVectorImpl<StringRef> &A,
77                            StringRef Separators, int MaxSplit,
78                            bool KeepEmpty) const {
79  StringRef rest = *this;
80
81  // rest.data() is used to distinguish cases like "a," that splits into
82  // "a" + "" and "a" that splits into "a" + 0.
83  for (int splits = 0;
84       rest.data() != NULL && (MaxSplit < 0 || splits < MaxSplit);
85       ++splits) {
86    std::pair<llvm::StringRef, llvm::StringRef> p = rest.split(Separators);
87
88    if (p.first.size() != 0 || KeepEmpty)
89      A.push_back(p.first);
90    rest = p.second;
91  }
92  // If we have a tail left, add it.
93  if (rest.data() != NULL && (rest.size() != 0 || KeepEmpty))
94    A.push_back(rest);
95}
96