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
  46 /// UnescapeString - Modify the argument string, turning two character sequences
  47 /// like '\\' 'n' into '\n'.  This handles: \e \a \b \f \n \r \t \v \' \\ and
  48 /// \num (where num is a 1-3 byte octal value).
  49 void llvm::UnescapeString(std::string &Str) {
  50   for (unsigned i = 0; i != Str.size(); ++i) {
  51     if (Str[i] == '\\' && i != Str.size()-1) {
  52       switch (Str[i+1]) {
  53       default: continue;  // Don't execute the code after the switch.
  54       case 'a': Str[i] = '\a'; break;
  55       case 'b': Str[i] = '\b'; break;
  56       case 'e': Str[i] = 27; break;
  57       case 'f': Str[i] = '\f'; break;
  58       case 'n': Str[i] = '\n'; break;
  59       case 'r': Str[i] = '\r'; break;
  60       case 't': Str[i] = '\t'; break;
  61       case 'v': Str[i] = '\v'; break;
  62       case '\'': Str[i] = '\''; break;
  63       case '\\': Str[i] = '\\'; break;
  64       }
  65       // Nuke the second character.
  66       Str.erase(Str.begin()+i+1);
  67     }
  68   }
  69 }
  70
  71 /// EscapeString - Modify the argument string, turning '\\' and anything that
  72 /// doesn't satisfy std::isprint into an escape sequence.
  73 void llvm::EscapeString(std::string &Str) {
  74   for (unsigned i = 0; i != Str.size(); ++i) {
  75     if (Str[i] == '\\') {
  76       ++i;
  77       Str.insert(Str.begin()+i, '\\');
  78     } else if (Str[i] == '\t') {
  79       Str[i++] = '\\';
  80       Str.insert(Str.begin()+i, 't');
  81     } else if (Str[i] == '\n') {
  82       Str[i++] = '\\';
  83       Str.insert(Str.begin()+i, 'n');
  84     } else if (!std::isprint(Str[i])) {
  85       // Always expand to a 3-digit octal escape.
  86       unsigned Char = Str[i];
  87       Str[i++] = '\\';
  88       Str.insert(Str.begin()+i++, '0'+((Char/64) & 7));
  89       Str.insert(Str.begin()+i++, '0'+((Char/8)  & 7));
  90       Str.insert(Str.begin()+i  , '0'+( Char     & 7));
  91     }
  92   }
  93 }