1b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Copyright 2013 The Chromium Authors. All rights reserved. 2b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Use of this source code is governed by a BSD-style license that can be 3b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// found in the LICENSE file. 4b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 5b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// This file defines utility functions for working with strings. 6b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 7b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#ifndef BASE_STRINGS_STRING_UTIL_H_ 8b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#define BASE_STRINGS_STRING_UTIL_H_ 9b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 10b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include <ctype.h> 11b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include <stdarg.h> // va_list 12cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#include <stddef.h> 13cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#include <stdint.h> 14b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 15b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include <string> 16b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include <vector> 17b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 18b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include "base/base_export.h" 19b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include "base/compiler_specific.h" 20b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include "base/strings/string16.h" 21b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#include "base/strings/string_piece.h" // For implicit conversions. 22cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#include "build/build_config.h" 23b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 24b910a63ff3111067e79c016f40a7c1baac943405Daniel Erat// On Android, bionic's stdio.h defines an snprintf macro when being built with 25b910a63ff3111067e79c016f40a7c1baac943405Daniel Erat// clang. Undefine it here so it won't collide with base::snprintf(). 26b910a63ff3111067e79c016f40a7c1baac943405Daniel Erat#undef snprintf 27b910a63ff3111067e79c016f40a7c1baac943405Daniel Erat 28b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Eratnamespace base { 29b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 30cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// C standard-library functions that aren't cross-platform are provided as 31cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// "base::...", and their prototypes are listed below. These functions are 32cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// then implemented as inline calls to the platform-specific equivalents in the 33cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// platform-specific headers. 34b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 35b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Wrapper for vsnprintf that always null-terminates and always returns the 36b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// number of characters that would be in an untruncated formatted 37b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// string, even when truncation occurs. 38b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Eratint vsnprintf(char* buffer, size_t size, const char* format, va_list arguments) 39b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat PRINTF_FORMAT(3, 0); 40b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 41b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Some of these implementations need to be inlined. 42b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 43b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// We separate the declaration from the implementation of this inline 44b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// function just so the PRINTF_FORMAT works. 45cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenkoinline int snprintf(char* buffer, 46cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko size_t size, 47cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko _Printf_format_string_ const char* format, 48cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko ...) PRINTF_FORMAT(3, 4); 49cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenkoinline int snprintf(char* buffer, 50cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko size_t size, 51cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko _Printf_format_string_ const char* format, 52cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko ...) { 53b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat va_list arguments; 54b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat va_start(arguments, format); 55b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat int result = vsnprintf(buffer, size, format, arguments); 56b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat va_end(arguments); 57b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return result; 58b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat} 59b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 60b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// BSD-style safe and consistent string copy functions. 61b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Copies |src| to |dst|, where |dst_size| is the total allocated size of |dst|. 62b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Copies at most |dst_size|-1 characters, and always NULL terminates |dst|, as 63b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// long as |dst_size| is not 0. Returns the length of |src| in characters. 64b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// If the return value is >= dst_size, then the output was truncated. 65b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// NOTE: All sizes are in number of characters, NOT in bytes. 66b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT size_t strlcpy(char* dst, const char* src, size_t dst_size); 67b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT size_t wcslcpy(wchar_t* dst, const wchar_t* src, size_t dst_size); 68b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 69b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Scan a wprintf format string to determine whether it's portable across a 70b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// variety of systems. This function only checks that the conversion 71b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// specifiers used by the format string are supported and have the same meaning 72b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// on a variety of systems. It doesn't check for other errors that might occur 73b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// within a format string. 74b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 75b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Nonportable conversion specifiers for wprintf are: 76b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// - 's' and 'c' without an 'l' length modifier. %s and %c operate on char 77b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// data on all systems except Windows, which treat them as wchar_t data. 78b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Use %ls and %lc for wchar_t data instead. 79b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// - 'S' and 'C', which operate on wchar_t data on all systems except Windows, 80b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// which treat them as char data. Use %ls and %lc for wchar_t data 81b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// instead. 82b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// - 'F', which is not identified by Windows wprintf documentation. 83b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// - 'D', 'O', and 'U', which are deprecated and not available on all systems. 84b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Use %ld, %lo, and %lu instead. 85b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 86b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Note that there is no portable conversion specifier for char data when 87b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// working with wprintf. 88b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 89b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// This function is intended to be called from base::vswprintf. 90b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool IsWprintfFormatPortable(const wchar_t* format); 91b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 92b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// ASCII-specific tolower. The standard library's tolower is locale sensitive, 93b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// so we don't want to use it here. 94cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenkoinline char ToLowerASCII(char c) { 95cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko return (c >= 'A' && c <= 'Z') ? (c + ('a' - 'A')) : c; 96cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko} 97cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenkoinline char16 ToLowerASCII(char16 c) { 98b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return (c >= 'A' && c <= 'Z') ? (c + ('a' - 'A')) : c; 99b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat} 100b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 101b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// ASCII-specific toupper. The standard library's toupper is locale sensitive, 102b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// so we don't want to use it here. 103cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenkoinline char ToUpperASCII(char c) { 104cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko return (c >= 'a' && c <= 'z') ? (c + ('A' - 'a')) : c; 105cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko} 106cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenkoinline char16 ToUpperASCII(char16 c) { 107b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return (c >= 'a' && c <= 'z') ? (c + ('A' - 'a')) : c; 108b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat} 109b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 110cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Converts the given string to it's ASCII-lowercase equivalent. 111cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT std::string ToLowerASCII(StringPiece str); 112cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT string16 ToLowerASCII(StringPiece16 str); 113b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 114cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Converts the given string to it's ASCII-uppercase equivalent. 115cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT std::string ToUpperASCII(StringPiece str); 116cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT string16 ToUpperASCII(StringPiece16 str); 117b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 118cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Functor for case-insensitive ASCII comparisons for STL algorithms like 119cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// std::search. 120cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// 121cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Note that a full Unicode version of this functor is not possible to write 122cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// because case mappings might change the number of characters, depend on 123cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// context (combining accents), and require handling UTF-16. If you need 124cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// proper Unicode support, use base::i18n::ToLower/FoldCase and then just 125cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// use a normal operator== on the result. 126b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erattemplate<typename Char> struct CaseInsensitiveCompareASCII { 127b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat public: 128b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat bool operator()(Char x, Char y) const { 129b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return ToLowerASCII(x) == ToLowerASCII(y); 130b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat } 131b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat}; 132b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 133cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Like strcasecmp for case-insensitive ASCII characters only. Returns: 134cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// -1 (a < b) 135cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// 0 (a == b) 136cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// 1 (a > b) 137cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// (unlike strcasecmp which can return values greater or less than 1/-1). For 138cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// full Unicode support, use base::i18n::ToLower or base::i18h::FoldCase 139cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// and then just call the normal string operators on the result. 140cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT int CompareCaseInsensitiveASCII(StringPiece a, StringPiece b); 141cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT int CompareCaseInsensitiveASCII(StringPiece16 a, StringPiece16 b); 142cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko 143cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Equality for ASCII case-insensitive comparisons. For full Unicode support, 144cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// use base::i18n::ToLower or base::i18h::FoldCase and then compare with either 145cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// == or !=. 146cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool EqualsCaseInsensitiveASCII(StringPiece a, StringPiece b); 147cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool EqualsCaseInsensitiveASCII(StringPiece16 a, StringPiece16 b); 148cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko 149b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// These threadsafe functions return references to globally unique empty 150b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// strings. 151b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 152b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// It is likely faster to construct a new empty string object (just a few 153b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// instructions to set the length to 0) than to get the empty string singleton 154b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// returned by these functions (which requires threadsafe singleton access). 155b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 156b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Therefore, DO NOT USE THESE AS A GENERAL-PURPOSE SUBSTITUTE FOR DEFAULT 157b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// CONSTRUCTORS. There is only one case where you should use these: functions 158b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// which need to return a string by reference (e.g. as a class member 159b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// accessor), and don't have an empty string to use (e.g. in an error case). 160b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// These should not be used as initializers, function arguments, or return 161b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// values for functions which return by value or outparam. 162b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT const std::string& EmptyString(); 163b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT const string16& EmptyString16(); 164b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 165b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Contains the set of characters representing whitespace in the corresponding 166b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// encoding. Null-terminated. The ASCII versions are the whitespaces as defined 167b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// by HTML5, and don't include control characters. 168b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT extern const wchar_t kWhitespaceWide[]; // Includes Unicode. 169b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT extern const char16 kWhitespaceUTF16[]; // Includes Unicode. 170b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT extern const char kWhitespaceASCII[]; 171b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT extern const char16 kWhitespaceASCIIAs16[]; // No unicode. 172b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 173b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Null-terminated string representing the UTF-8 byte order mark. 174b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT extern const char kUtf8ByteOrderMark[]; 175b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 176b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Removes characters in |remove_chars| from anywhere in |input|. Returns true 177b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// if any characters were removed. |remove_chars| must be null-terminated. 178b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// NOTE: Safe to use the same variable for both |input| and |output|. 179b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool RemoveChars(const string16& input, 180cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const StringPiece16& remove_chars, 181b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat string16* output); 182b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool RemoveChars(const std::string& input, 183cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const StringPiece& remove_chars, 184b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::string* output); 185b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 186b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Replaces characters in |replace_chars| from anywhere in |input| with 187b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// |replace_with|. Each character in |replace_chars| will be replaced with 188b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// the |replace_with| string. Returns true if any characters were replaced. 189b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// |replace_chars| must be null-terminated. 190b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// NOTE: Safe to use the same variable for both |input| and |output|. 191b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool ReplaceChars(const string16& input, 192cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const StringPiece16& replace_chars, 193b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const string16& replace_with, 194b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat string16* output); 195b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool ReplaceChars(const std::string& input, 196cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const StringPiece& replace_chars, 197b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const std::string& replace_with, 198b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::string* output); 199b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 200b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Eratenum TrimPositions { 201b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TRIM_NONE = 0, 202b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TRIM_LEADING = 1 << 0, 203b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TRIM_TRAILING = 1 << 1, 204b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TRIM_ALL = TRIM_LEADING | TRIM_TRAILING, 205b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat}; 206b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 207b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Removes characters in |trim_chars| from the beginning and end of |input|. 208b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// The 8-bit version only works on 8-bit characters, not UTF-8. 209b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 210b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// It is safe to use the same variable for both |input| and |output| (this is 211b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// the normal usage to trim in-place). 212b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool TrimString(const string16& input, 213cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 trim_chars, 214b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat string16* output); 215b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool TrimString(const std::string& input, 216cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece trim_chars, 217b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::string* output); 218b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 219b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// StringPiece versions of the above. The returned pieces refer to the original 220b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// buffer. 221b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT StringPiece16 TrimString(StringPiece16 input, 222cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const StringPiece16& trim_chars, 223b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TrimPositions positions); 224b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT StringPiece TrimString(StringPiece input, 225cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const StringPiece& trim_chars, 226b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TrimPositions positions); 227b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 228b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Truncates a string to the nearest UTF-8 character that will leave 229b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// the string less than or equal to the specified byte size. 230b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT void TruncateUTF8ToByteSize(const std::string& input, 231b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const size_t byte_size, 232b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::string* output); 233b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 234cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Trims any whitespace from either end of the input string. 235cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// 236cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// The StringPiece versions return a substring referencing the input buffer. 237cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// The ASCII versions look only for ASCII whitespace. 238cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// 239cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// The std::string versions return where whitespace was found. 240b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// NOTE: Safe to use the same variable for both input and output. 241b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT TrimPositions TrimWhitespace(const string16& input, 242b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TrimPositions positions, 243cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko string16* output); 244cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT StringPiece16 TrimWhitespace(StringPiece16 input, 245cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko TrimPositions positions); 246b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT TrimPositions TrimWhitespaceASCII(const std::string& input, 247b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat TrimPositions positions, 248b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::string* output); 249cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT StringPiece TrimWhitespaceASCII(StringPiece input, 250cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko TrimPositions positions); 251b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 252b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Searches for CR or LF characters. Removes all contiguous whitespace 253b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// strings that contain them. This is useful when trying to deal with text 254b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// copied from terminals. 255b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Returns |text|, with the following three transformations: 256b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// (1) Leading and trailing whitespace is trimmed. 257b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// (2) If |trim_sequences_with_line_breaks| is true, any other whitespace 258b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// sequences containing a CR or LF are trimmed. 259b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// (3) All other whitespace sequences are converted to single spaces. 260b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT string16 CollapseWhitespace( 261b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const string16& text, 262b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat bool trim_sequences_with_line_breaks); 263b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT std::string CollapseWhitespaceASCII( 264b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const std::string& text, 265b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat bool trim_sequences_with_line_breaks); 266b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 267b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Returns true if |input| is empty or contains only characters found in 268b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// |characters|. 269b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool ContainsOnlyChars(const StringPiece& input, 270b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const StringPiece& characters); 271b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool ContainsOnlyChars(const StringPiece16& input, 272b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const StringPiece16& characters); 273b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 274b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Returns true if the specified string matches the criteria. How can a wide 275b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// string be 8-bit or UTF8? It contains only characters that are < 256 (in the 276b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// first case) or characters that use only 8-bits and whose 8-bit 277b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// representation looks like a UTF-8 string (the second case). 278b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 279b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Note that IsStringUTF8 checks not only if the input is structurally 280b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// valid but also if it doesn't contain any non-character codepoint 281b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// (e.g. U+FFFE). It's done on purpose because all the existing callers want 282b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// to have the maximum 'discriminating' power from other encodings. If 283b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// there's a use case for just checking the structural validity, we have to 284b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// add a new function for that. 285b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 286b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// IsStringASCII assumes the input is likely all ASCII, and does not leave early 287b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// if it is not the case. 288b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool IsStringUTF8(const StringPiece& str); 289b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool IsStringASCII(const StringPiece& str); 290b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool IsStringASCII(const StringPiece16& str); 291b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// A convenience adaptor for WebStrings, as they don't convert into 292b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// StringPieces directly. 293b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool IsStringASCII(const string16& str); 294b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#if defined(WCHAR_T_IS_UTF32) 295b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT bool IsStringASCII(const std::wstring& str); 296b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#endif 297b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 298cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Compare the lower-case form of the given string against the given 299cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// previously-lower-cased ASCII string (typically a constant). 300cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool LowerCaseEqualsASCII(StringPiece str, 301cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece lowecase_ascii); 302cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool LowerCaseEqualsASCII(StringPiece16 str, 303cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece lowecase_ascii); 304cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko 305cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Performs a case-sensitive string compare of the given 16-bit string against 306cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// the given 8-bit ASCII string (typically a constant). The behavior is 307cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// undefined if the |ascii| string is not ASCII. 308cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool EqualsASCII(StringPiece16 str, StringPiece ascii); 309cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko 310cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Indicates case sensitivity of comparisons. Only ASCII case insensitivity 311cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// is supported. Full Unicode case-insensitive conversions would need to go in 312cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// base/i18n so it can use ICU. 313b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 314cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// If you need to do Unicode-aware case-insensitive StartsWith/EndsWith, it's 315cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// best to call base::i18n::ToLower() or base::i18n::FoldCase() (see 316cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// base/i18n/case_conversion.h for usage advice) on the arguments, and then use 317cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// the results to a case-sensitive comparison. 318cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenkoenum class CompareCase { 319cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko SENSITIVE, 320cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko INSENSITIVE_ASCII, 321cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko}; 322b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 323cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool StartsWith(StringPiece str, 324cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece search_for, 325cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko CompareCase case_sensitivity); 326cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool StartsWith(StringPiece16 str, 327cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 search_for, 328cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko CompareCase case_sensitivity); 329cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool EndsWith(StringPiece str, 330cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece search_for, 331cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko CompareCase case_sensitivity); 332cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool EndsWith(StringPiece16 str, 333cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 search_for, 334cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko CompareCase case_sensitivity); 335b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 336b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Determines the type of ASCII character, independent of locale (the C 337b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// library versions will change based on locale). 338b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erattemplate <typename Char> 339b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Eratinline bool IsAsciiWhitespace(Char c) { 340b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return c == ' ' || c == '\r' || c == '\n' || c == '\t'; 341b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat} 342b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erattemplate <typename Char> 343b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Eratinline bool IsAsciiAlpha(Char c) { 344b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return ((c >= 'A') && (c <= 'Z')) || ((c >= 'a') && (c <= 'z')); 345b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat} 346b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erattemplate <typename Char> 347b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Eratinline bool IsAsciiDigit(Char c) { 348b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return c >= '0' && c <= '9'; 349b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat} 350b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 351b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erattemplate <typename Char> 352b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Eratinline bool IsHexDigit(Char c) { 353b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat return (c >= '0' && c <= '9') || 354b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat (c >= 'A' && c <= 'F') || 355b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat (c >= 'a' && c <= 'f'); 356b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat} 357b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 358cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Returns the integer corresponding to the given hex character. For example: 359cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// '4' -> 4 360cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// 'a' -> 10 361cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// 'B' -> 11 362cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Assumes the input is a valid hex character. DCHECKs in debug builds if not. 363cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT char HexDigitToInt(wchar_t c); 364b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 365cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko// Returns true if it's a Unicode whitespace character. 366cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT bool IsUnicodeWhitespace(wchar_t c); 367b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 368b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Return a byte string in human-readable format with a unit suffix. Not 369b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// appropriate for use in any UI; use of FormatBytes and friends in ui/base is 370b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// highly recommended instead. TODO(avi): Figure out how to get callers to use 371b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// FormatBytes instead; remove this. 372cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT string16 FormatBytesUnlocalized(int64_t bytes); 373b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 374b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Starting at |start_offset| (usually 0), replace the first instance of 375b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// |find_this| with |replace_with|. 376b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT void ReplaceFirstSubstringAfterOffset( 377b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat base::string16* str, 378b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat size_t start_offset, 379cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 find_this, 380cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 replace_with); 381b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT void ReplaceFirstSubstringAfterOffset( 382b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::string* str, 383b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat size_t start_offset, 384cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece find_this, 385cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece replace_with); 386b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 387b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Starting at |start_offset| (usually 0), look through |str| and replace all 388b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// instances of |find_this| with |replace_with|. 389b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 390b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// This does entire substrings; use std::replace in <algorithm> for single 391b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// characters, for example: 392b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// std::replace(str.begin(), str.end(), 'a', 'b'); 393b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT void ReplaceSubstringsAfterOffset( 394cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko string16* str, 395b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat size_t start_offset, 396cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 find_this, 397cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 replace_with); 398cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT void ReplaceSubstringsAfterOffset( 399cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko std::string* str, 400cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko size_t start_offset, 401cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece find_this, 402cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece replace_with); 403b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 404b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Reserves enough memory in |str| to accommodate |length_with_null| characters, 405b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// sets the size of |str| to |length_with_null - 1| characters, and returns a 406b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// pointer to the underlying contiguous array of characters. This is typically 407b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// used when calling a function that writes results into a character array, but 408b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// the caller wants the data to be managed by a string-like object. It is 409b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// convenient in that is can be used inline in the call, and fast in that it 410b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// avoids copying the results of the call from a char* into a string. 411b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 412b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// |length_with_null| must be at least 2, since otherwise the underlying string 413b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// would have size 0, and trying to access &((*str)[0]) in that case can result 414b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// in a number of problems. 415b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// 416b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Internally, this takes linear time because the resize() call 0-fills the 417b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// underlying array for potentially all 418b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// (|length_with_null - 1| * sizeof(string_type::value_type)) bytes. Ideally we 419b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// could avoid this aspect of the resize() call, as we expect the caller to 420b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// immediately write over this memory, but there is no other way to set the size 421b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// of the string, and not doing that will mean people who access |str| rather 422b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// than str.c_str() will get back a string of whatever size |str| had on entry 423b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// to this function (probably 0). 424cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT char* WriteInto(std::string* str, size_t length_with_null); 425cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT char16* WriteInto(string16* str, size_t length_with_null); 426cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#ifndef OS_WIN 427cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT wchar_t* WriteInto(std::wstring* str, size_t length_with_null); 428cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#endif 429b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 430b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Does the opposite of SplitString(). 431cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT std::string JoinString(const std::vector<std::string>& parts, 432cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece separator); 433cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT string16 JoinString(const std::vector<string16>& parts, 434cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko StringPiece16 separator); 435b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 436b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Replace $1-$2-$3..$9 in the format string with |a|-|b|-|c|..|i| respectively. 437b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Additionally, any number of consecutive '$' characters is replaced by that 438b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// number less one. Eg $$->$, $$$->$$, etc. The offsets parameter here can be 439b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// NULL. This only allows you to use up to nine replacements. 440cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT string16 ReplaceStringPlaceholders( 441cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const string16& format_string, 442cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const std::vector<string16>& subst, 443b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::vector<size_t>* offsets); 444b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 445b8cf94937c52feb53b55c39e3f82094d27de464cDaniel EratBASE_EXPORT std::string ReplaceStringPlaceholders( 446cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const StringPiece& format_string, 447b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat const std::vector<std::string>& subst, 448b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat std::vector<size_t>* offsets); 449b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 450b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat// Single-string shortcut for ReplaceStringHolders. |offset| may be NULL. 451cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex VakulenkoBASE_EXPORT string16 ReplaceStringPlaceholders(const string16& format_string, 452cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko const string16& a, 453cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko size_t* offset); 454cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko 455cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko} // namespace base 456cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko 457cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#if defined(OS_WIN) 458cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#include "base/strings/string_util_win.h" 459cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#elif defined(OS_POSIX) 460cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#include "base/strings/string_util_posix.h" 461cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#else 462cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#error Define string operations appropriately for your platform 463cce46a0c214b37e8da48c522c83037e8ffa4f9fdAlex Vakulenko#endif 464b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat 465b8cf94937c52feb53b55c39e3f82094d27de464cDaniel Erat#endif // BASE_STRINGS_STRING_UTIL_H_ 466