lib/Support/StringExtras.cpp

   1 //===-- StringExtras.cpp - Implement the StringExtras header --------------===//
   2 //
   3 //                     The LLVM Compiler Infrastructure
   4 //
   5 // This file was developed by the LLVM research group and is distributed under
   6 // the University of Illinois Open Source License. See LICENSE.TXT for details.
   7 //
   8 //===----------------------------------------------------------------------===//
   9 //
  10 // This file implements the StringExtras.h header
  11 //
  12 //===----------------------------------------------------------------------===//
  13
  14 #include "llvm/ADT/StringExtras.h"
  15 using namespace llvm;
  16
  17 /// getToken - This function extracts one token from source, ignoring any
  18 /// leading characters that appear in the Delimiters string, and ending the
  19 /// token at any of the characters that appear in the Delimiters string.  If
  20 /// there are no tokens in the source string, an empty string is returned.
  21 /// The Source source string is updated in place to remove the returned string
  22 /// and any delimiter prefix from it.
  23 std::string llvm::getToken(std::string &Source, const char *Delimiters) {
  24   unsigned NumDelimiters = std::strlen(Delimiters);
  25
  26   // Figure out where the token starts.
  27   std::string::size_type Start =
  28     Source.find_first_not_of(Delimiters, 0, NumDelimiters);
  29   if (Start == std::string::npos) Start = Source.size();
  30
  31   // Find the next occurance of the delimiter.
  32   std::string::size_type End =
  33     Source.find_first_of(Delimiters, Start, NumDelimiters);
  34   if (End == std::string::npos) End = Source.size();
  35
  36   // Create the return token.
  37   std::string Result = std::string(Source.begin()+Start, Source.begin()+End);
  38
  39   // Erase the token that we read in.
  40   Source.erase(Source.begin(), Source.begin()+End);
  41
  42   return Result;
  43 }
  44
  45 /// SplitString - Split up the specified string according to the specified
  46 /// delimiters, appending the result fragments to the output list.
  47 void llvm::SplitString(const std::string &Source,
  48                        std::vector<std::string> &OutFragments,
  49                        const char *Delimiters) {
  50   std::string S = Source;
  51
  52   std::string S2 = getToken(S, Delimiters);
  53   while (!S2.empty()) {
  54     OutFragments.push_back(S2);
  55     S2 = getToken(S, Delimiters);
  56   }
  57 }
  58
  59
  60
  61 /// UnescapeString - Modify the argument string, turning two character sequences
  62 /// @verbatim
  63 /// like '\\' 'n' into '\n'.  This handles: \e \a \b \f \n \r \t \v \' \ and
  64 /// \num (where num is a 1-3 byte octal value).
  65 /// @endverbatim
  66 void llvm::UnescapeString(std::string &Str) {
  67   for (unsigned i = 0; i != Str.size(); ++i) {
  68     if (Str[i] == '\\' && i != Str.size()-1) {
  69       switch (Str[i+1]) {
  70       default: continue;  // Don't execute the code after the switch.
  71       case 'a': Str[i] = '\a'; break;
  72       case 'b': Str[i] = '\b'; break;
  73       case 'e': Str[i] = 27; break;
  74       case 'f': Str[i] = '\f'; break;
  75       case 'n': Str[i] = '\n'; break;
  76       case 'r': Str[i] = '\r'; break;
  77       case 't': Str[i] = '\t'; break;
  78       case 'v': Str[i] = '\v'; break;
  79       case '\'': Str[i] = '\''; break;
  80       case '\\': Str[i] = '\\'; break;
  81       }
  82       // Nuke the second character.
  83       Str.erase(Str.begin()+i+1);
  84     }
  85   }
  86 }
  87
  88 /// EscapeString - Modify the argument string, turning '\\' and anything that
  89 /// doesn't satisfy std::isprint into an escape sequence.
  90 void llvm::EscapeString(std::string &Str) {
  91   for (unsigned i = 0; i != Str.size(); ++i) {
  92     if (Str[i] == '\\') {
  93       ++i;
  94       Str.insert(Str.begin()+i, '\\');
  95     } else if (Str[i] == '\t') {
  96       Str[i++] = '\\';
  97       Str.insert(Str.begin()+i, 't');
  98     } else if (Str[i] == '\n') {
  99       Str[i++] = '\\';
 100       Str.insert(Str.begin()+i, 'n');
 101     } else if (!std::isprint(Str[i])) {
 102       // Always expand to a 3-digit octal escape.
 103       unsigned Char = Str[i];
 104       Str[i++] = '\\';
 105       Str.insert(Str.begin()+i++, '0'+((Char/64) & 7));
 106       Str.insert(Str.begin()+i++, '0'+((Char/8)  & 7));
 107       Str.insert(Str.begin()+i  , '0'+( Char     & 7));
 108     }
 109   }
 110 }