Chris Lattner | 872ccce | 2003-12-29 05:07:02 +0000 | [diff] [blame] | 1 | //===-- StringExtras.cpp - Implement the StringExtras header --------------===// |
Misha Brukman | f976c85 | 2005-04-21 22:55:34 +0000 | [diff] [blame] | 2 | // |
Chris Lattner | 872ccce | 2003-12-29 05:07:02 +0000 | [diff] [blame] | 3 | // The LLVM Compiler Infrastructure |
| 4 | // |
Chris Lattner | 4ee451d | 2007-12-29 20:36:04 +0000 | [diff] [blame] | 5 | // This file is distributed under the University of Illinois Open Source |
| 6 | // License. See LICENSE.TXT for details. |
Misha Brukman | f976c85 | 2005-04-21 22:55:34 +0000 | [diff] [blame] | 7 | // |
Chris Lattner | 872ccce | 2003-12-29 05:07:02 +0000 | [diff] [blame] | 8 | //===----------------------------------------------------------------------===// |
| 9 | // |
| 10 | // This file implements the StringExtras.h header |
| 11 | // |
| 12 | //===----------------------------------------------------------------------===// |
| 13 | |
Reid Spencer | 551ccae | 2004-09-01 22:55:40 +0000 | [diff] [blame] | 14 | #include "llvm/ADT/StringExtras.h" |
Rafael Espindola | c78c0c9 | 2009-11-13 02:18:25 +0000 | [diff] [blame] | 15 | #include "llvm/ADT/SmallVector.h" |
Anton Korobeynikov | ae9f3a3 | 2008-02-20 11:08:44 +0000 | [diff] [blame] | 16 | #include <cstring> |
Chris Lattner | 872ccce | 2003-12-29 05:07:02 +0000 | [diff] [blame] | 17 | using namespace llvm; |
| 18 | |
| 19 | /// getToken - This function extracts one token from source, ignoring any |
| 20 | /// leading characters that appear in the Delimiters string, and ending the |
| 21 | /// token at any of the characters that appear in the Delimiters string. If |
| 22 | /// there are no tokens in the source string, an empty string is returned. |
| 23 | /// The Source source string is updated in place to remove the returned string |
| 24 | /// and any delimiter prefix from it. |
| 25 | std::string llvm::getToken(std::string &Source, const char *Delimiters) { |
Evan Cheng | 34cd4a4 | 2008-05-05 18:30:58 +0000 | [diff] [blame] | 26 | size_t NumDelimiters = std::strlen(Delimiters); |
Chris Lattner | 872ccce | 2003-12-29 05:07:02 +0000 | [diff] [blame] | 27 | |
| 28 | // Figure out where the token starts. |
| 29 | std::string::size_type Start = |
| 30 | Source.find_first_not_of(Delimiters, 0, NumDelimiters); |
| 31 | if (Start == std::string::npos) Start = Source.size(); |
| 32 | |
| 33 | // Find the next occurance of the delimiter. |
| 34 | std::string::size_type End = |
| 35 | Source.find_first_of(Delimiters, Start, NumDelimiters); |
| 36 | if (End == std::string::npos) End = Source.size(); |
| 37 | |
| 38 | // Create the return token. |
| 39 | std::string Result = std::string(Source.begin()+Start, Source.begin()+End); |
Misha Brukman | f976c85 | 2005-04-21 22:55:34 +0000 | [diff] [blame] | 40 | |
Chris Lattner | 872ccce | 2003-12-29 05:07:02 +0000 | [diff] [blame] | 41 | // Erase the token that we read in. |
| 42 | Source.erase(Source.begin(), Source.begin()+End); |
Misha Brukman | f976c85 | 2005-04-21 22:55:34 +0000 | [diff] [blame] | 43 | |
Chris Lattner | 872ccce | 2003-12-29 05:07:02 +0000 | [diff] [blame] | 44 | return Result; |
| 45 | } |
Chris Lattner | 53eca94 | 2006-07-14 22:54:39 +0000 | [diff] [blame] | 46 | |
Chris Lattner | d5b58c2 | 2006-11-28 22:32:35 +0000 | [diff] [blame] | 47 | /// SplitString - Split up the specified string according to the specified |
| 48 | /// delimiters, appending the result fragments to the output list. |
| 49 | void llvm::SplitString(const std::string &Source, |
| 50 | std::vector<std::string> &OutFragments, |
| 51 | const char *Delimiters) { |
| 52 | std::string S = Source; |
| 53 | |
| 54 | std::string S2 = getToken(S, Delimiters); |
| 55 | while (!S2.empty()) { |
| 56 | OutFragments.push_back(S2); |
| 57 | S2 = getToken(S, Delimiters); |
| 58 | } |
| 59 | } |
Rafael Espindola | 5ccac24 | 2009-11-13 01:24:40 +0000 | [diff] [blame] | 60 | |
Rafael Espindola | c78c0c9 | 2009-11-13 02:18:25 +0000 | [diff] [blame] | 61 | void llvm::StringRef::split(SmallVectorImpl<StringRef> &A, |
| 62 | StringRef Separators, int MaxSplit, |
Rafael Espindola | 5ccac24 | 2009-11-13 01:24:40 +0000 | [diff] [blame] | 63 | bool KeepEmpty) const { |
| 64 | StringRef rest = *this; |
| 65 | |
Rafael Espindola | 20fd4ec | 2009-11-13 04:55:09 +0000 | [diff] [blame^] | 66 | // rest.data() is used to distinguish cases like "a," that splits into |
| 67 | // "a" + "" and "a" that splits into "a" + 0. |
Rafael Espindola | c78c0c9 | 2009-11-13 02:18:25 +0000 | [diff] [blame] | 68 | for (int splits = 0; |
Rafael Espindola | 20fd4ec | 2009-11-13 04:55:09 +0000 | [diff] [blame^] | 69 | rest.data() != NULL && (MaxSplit < 0 || splits < MaxSplit); |
Rafael Espindola | 5ccac24 | 2009-11-13 01:24:40 +0000 | [diff] [blame] | 70 | ++splits) { |
| 71 | std::pair<llvm::StringRef, llvm::StringRef> p = rest.split(Separators); |
| 72 | |
| 73 | if (p.first.size() != 0 || KeepEmpty) |
| 74 | A.push_back(p.first); |
| 75 | rest = p.second; |
| 76 | } |
Rafael Espindola | 20fd4ec | 2009-11-13 04:55:09 +0000 | [diff] [blame^] | 77 | // If we have a tail left, add it. |
| 78 | if (rest.data() != NULL && (rest.size() != 0 || KeepEmpty)) |
Rafael Espindola | 5ccac24 | 2009-11-13 01:24:40 +0000 | [diff] [blame] | 79 | A.push_back(rest); |
| 80 | } |