michael@0: // Copyright (C) 2009-2013, International Business Machines
michael@0: // Corporation and others. All Rights Reserved.
michael@0: //
michael@0: // Copyright 2001 and onwards Google Inc.
michael@0: // Author: Sanjay Ghemawat
michael@0: 
michael@0: // This code is a contribution of Google code, and the style used here is
michael@0: // a compromise between the original Google code and the ICU coding guidelines.
michael@0: // For example, data types are ICU-ified (size_t,int->int32_t),
michael@0: // and API comments doxygen-ified, but function names and behavior are
michael@0: // as in the original, if possible.
michael@0: // Assertion-style error handling, not available in ICU, was changed to
michael@0: // parameter "pinning" similar to UnicodeString.
michael@0: //
michael@0: // In addition, this is only a partial port of the original Google code,
michael@0: // limited to what was needed so far. The (nearly) complete original code
michael@0: // is in the ICU svn repository at icuhtml/trunk/design/strings/contrib
michael@0: // (see ICU ticket 6765, r25517).
michael@0: 
michael@0: #ifndef __STRINGPIECE_H__
michael@0: #define __STRINGPIECE_H__
michael@0: 
michael@0: /**
michael@0:  * \file 
michael@0:  * \brief C++ API: StringPiece: Read-only byte string wrapper class.
michael@0:  */
michael@0: 
michael@0: #include "unicode/utypes.h"
michael@0: #include "unicode/uobject.h"
michael@0: #include "unicode/std_string.h"
michael@0: 
michael@0: // Arghh!  I wish C++ literals were "string".
michael@0: 
michael@0: U_NAMESPACE_BEGIN
michael@0: 
michael@0: /**
michael@0:  * A string-like object that points to a sized piece of memory.
michael@0:  *
michael@0:  * We provide non-explicit singleton constructors so users can pass
michael@0:  * in a "const char*" or a "string" wherever a "StringPiece" is
michael@0:  * expected.
michael@0:  *
michael@0:  * Functions or methods may use const StringPiece& parameters to accept either
michael@0:  * a "const char*" or a "string" value that will be implicitly converted to
michael@0:  * a StringPiece.
michael@0:  *
michael@0:  * Systematic usage of StringPiece is encouraged as it will reduce unnecessary
michael@0:  * conversions from "const char*" to "string" and back again.
michael@0:  *
michael@0:  * @stable ICU 4.2
michael@0:  */
michael@0: class U_COMMON_API StringPiece : public UMemory {
michael@0:  private:
michael@0:   const char*   ptr_;
michael@0:   int32_t       length_;
michael@0: 
michael@0:  public:
michael@0:   /**
michael@0:    * Default constructor, creates an empty StringPiece.
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   StringPiece() : ptr_(NULL), length_(0) { }
michael@0:   /**
michael@0:    * Constructs from a NUL-terminated const char * pointer.
michael@0:    * @param str a NUL-terminated const char * pointer
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   StringPiece(const char* str);
michael@0: #if U_HAVE_STD_STRING
michael@0:   /**
michael@0:    * Constructs from a std::string.
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   StringPiece(const std::string& str)
michael@0:     : ptr_(str.data()), length_(static_cast<int32_t>(str.size())) { }
michael@0: #endif
michael@0:   /**
michael@0:    * Constructs from a const char * pointer and a specified length.
michael@0:    * @param offset a const char * pointer (need not be terminated)
michael@0:    * @param len the length of the string; must be non-negative
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   StringPiece(const char* offset, int32_t len) : ptr_(offset), length_(len) { }
michael@0:   /**
michael@0:    * Substring of another StringPiece.
michael@0:    * @param x the other StringPiece
michael@0:    * @param pos start position in x; must be non-negative and <= x.length().
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   StringPiece(const StringPiece& x, int32_t pos);
michael@0:   /**
michael@0:    * Substring of another StringPiece.
michael@0:    * @param x the other StringPiece
michael@0:    * @param pos start position in x; must be non-negative and <= x.length().
michael@0:    * @param len length of the substring;
michael@0:    *            must be non-negative and will be pinned to at most x.length() - pos.
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   StringPiece(const StringPiece& x, int32_t pos, int32_t len);
michael@0: 
michael@0:   /**
michael@0:    * Returns the string pointer. May be NULL if it is empty.
michael@0:    *
michael@0:    * data() may return a pointer to a buffer with embedded NULs, and the
michael@0:    * returned buffer may or may not be null terminated.  Therefore it is
michael@0:    * typically a mistake to pass data() to a routine that expects a NUL
michael@0:    * terminated string.
michael@0:    * @return the string pointer
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   const char* data() const { return ptr_; }
michael@0:   /**
michael@0:    * Returns the string length. Same as length().
michael@0:    * @return the string length
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   int32_t size() const { return length_; }
michael@0:   /**
michael@0:    * Returns the string length. Same as size().
michael@0:    * @return the string length
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   int32_t length() const { return length_; }
michael@0:   /**
michael@0:    * Returns whether the string is empty.
michael@0:    * @return TRUE if the string is empty
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   UBool empty() const { return length_ == 0; }
michael@0: 
michael@0:   /**
michael@0:    * Sets to an empty string.
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   void clear() { ptr_ = NULL; length_ = 0; }
michael@0: 
michael@0:   /**
michael@0:    * Reset the stringpiece to refer to new data.
michael@0:    * @param xdata pointer the new string data.  Need not be nul terminated.
michael@0:    * @param len the length of the new data
michael@0:    * @stable ICU 4.8
michael@0:    */
michael@0:   void set(const char* xdata, int32_t len) { ptr_ = xdata; length_ = len; }
michael@0: 
michael@0:   /**
michael@0:    * Reset the stringpiece to refer to new data.
michael@0:    * @param str a pointer to a NUL-terminated string. 
michael@0:    * @stable ICU 4.8
michael@0:    */
michael@0:   void set(const char* str);
michael@0: 
michael@0:   /**
michael@0:    * Removes the first n string units.
michael@0:    * @param n prefix length, must be non-negative and <=length()
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   void remove_prefix(int32_t n) {
michael@0:     if (n >= 0) {
michael@0:       if (n > length_) {
michael@0:         n = length_;
michael@0:       }
michael@0:       ptr_ += n;
michael@0:       length_ -= n;
michael@0:     }
michael@0:   }
michael@0: 
michael@0:   /**
michael@0:    * Removes the last n string units.
michael@0:    * @param n suffix length, must be non-negative and <=length()
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   void remove_suffix(int32_t n) {
michael@0:     if (n >= 0) {
michael@0:       if (n <= length_) {
michael@0:         length_ -= n;
michael@0:       } else {
michael@0:         length_ = 0;
michael@0:       }
michael@0:     }
michael@0:   }
michael@0: 
michael@0:   /**
michael@0:    * Maximum integer, used as a default value for substring methods.
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   static const int32_t npos; // = 0x7fffffff;
michael@0: 
michael@0:   /**
michael@0:    * Returns a substring of this StringPiece.
michael@0:    * @param pos start position; must be non-negative and <= length().
michael@0:    * @param len length of the substring;
michael@0:    *            must be non-negative and will be pinned to at most length() - pos.
michael@0:    * @return the substring StringPiece
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   StringPiece substr(int32_t pos, int32_t len = npos) const {
michael@0:     return StringPiece(*this, pos, len);
michael@0:   }
michael@0: };
michael@0: 
michael@0: /**
michael@0:  * Global operator == for StringPiece
michael@0:  * @param x The first StringPiece to compare.
michael@0:  * @param y The second StringPiece to compare.
michael@0:  * @return TRUE if the string data is equal
michael@0:  * @stable ICU 4.8
michael@0:  */
michael@0: U_EXPORT UBool U_EXPORT2 
michael@0: operator==(const StringPiece& x, const StringPiece& y);
michael@0: 
michael@0: /**
michael@0:  * Global operator != for StringPiece
michael@0:  * @param x The first StringPiece to compare.
michael@0:  * @param y The second StringPiece to compare.
michael@0:  * @return TRUE if the string data is not equal
michael@0:  * @stable ICU 4.8
michael@0:  */
michael@0: inline UBool operator!=(const StringPiece& x, const StringPiece& y) {
michael@0:   return !(x == y);
michael@0: }
michael@0: 
michael@0: U_NAMESPACE_END
michael@0: 
michael@0: #endif  // __STRINGPIECE_H__