michael@0: // Copyright (c) 2008 The Chromium Authors. All rights reserved.
michael@0: // Use of this source code is governed by a BSD-style license that can be
michael@0: // found in the LICENSE file.
michael@0: 
michael@0: // FilePath is a container for pathnames stored in a platform's native string
michael@0: // type, providing containers for manipulation in according with the
michael@0: // platform's conventions for pathnames.  It supports the following path
michael@0: // types:
michael@0: //
michael@0: //                   POSIX            Windows
michael@0: //                   ---------------  ----------------------------------
michael@0: // Fundamental type  char[]           wchar_t[]
michael@0: // Encoding          unspecified*     UTF-16
michael@0: // Separator         /                \, tolerant of /
michael@0: // Drive letters     no               case-insensitive A-Z followed by :
michael@0: // Alternate root    // (surprise!)   \\, for UNC paths
michael@0: //
michael@0: // * The encoding need not be specified on POSIX systems, although some
michael@0: //   POSIX-compliant systems do specify an encoding.  Mac OS X uses UTF-8.
michael@0: //   Linux does not specify an encoding, but in practice, the locale's
michael@0: //   character set may be used.
michael@0: //
michael@0: // FilePath objects are intended to be used anywhere paths are.  An
michael@0: // application may pass FilePath objects around internally, masking the
michael@0: // underlying differences between systems, only differing in implementation
michael@0: // where interfacing directly with the system.  For example, a single
michael@0: // OpenFile(const FilePath &) function may be made available, allowing all
michael@0: // callers to operate without regard to the underlying implementation.  On
michael@0: // POSIX-like platforms, OpenFile might wrap fopen, and on Windows, it might
michael@0: // wrap _wfopen_s, perhaps both by calling file_path.value().c_str().  This
michael@0: // allows each platform to pass pathnames around without requiring conversions
michael@0: // between encodings, which has an impact on performance, but more imporantly,
michael@0: // has an impact on correctness on platforms that do not have well-defined
michael@0: // encodings for pathnames.
michael@0: //
michael@0: // Several methods are available to perform common operations on a FilePath
michael@0: // object, such as determining the parent directory (DirName), isolating the
michael@0: // final path component (BaseName), and appending a relative pathname string
michael@0: // to an existing FilePath object (Append).  These methods are highly
michael@0: // recommended over attempting to split and concatenate strings directly.
michael@0: // These methods are based purely on string manipulation and knowledge of
michael@0: // platform-specific pathname conventions, and do not consult the filesystem
michael@0: // at all, making them safe to use without fear of blocking on I/O operations.
michael@0: // These methods do not function as mutators but instead return distinct
michael@0: // instances of FilePath objects, and are therefore safe to use on const
michael@0: // objects.  The objects themselves are safe to share between threads.
michael@0: //
michael@0: // To aid in initialization of FilePath objects from string literals, a
michael@0: // FILE_PATH_LITERAL macro is provided, which accounts for the difference
michael@0: // between char[]-based pathnames on POSIX systems and wchar_t[]-based
michael@0: // pathnames on Windows.
michael@0: //
michael@0: // Because a FilePath object should not be instantiated at the global scope,
michael@0: // instead, use a FilePath::CharType[] and initialize it with
michael@0: // FILE_PATH_LITERAL.  At runtime, a FilePath object can be created from the
michael@0: // character array.  Example:
michael@0: //
michael@0: // | const FilePath::CharType kLogFileName[] = FILE_PATH_LITERAL("log.txt");
michael@0: // |
michael@0: // | void Function() {
michael@0: // |   FilePath log_file_path(kLogFileName);
michael@0: // |   [...]
michael@0: // | }
michael@0: 
michael@0: #ifndef BASE_FILE_PATH_H_
michael@0: #define BASE_FILE_PATH_H_
michael@0: 
michael@0: #include <string>
michael@0: 
michael@0: #include "base/basictypes.h"
michael@0: #include "base/compiler_specific.h"
michael@0: #include "base/hash_tables.h"
michael@0: 
michael@0: // Windows-style drive letter support and pathname separator characters can be
michael@0: // enabled and disabled independently, to aid testing.  These #defines are
michael@0: // here so that the same setting can be used in both the implementation and
michael@0: // in the unit test.
michael@0: #if defined(OS_WIN)
michael@0: #define FILE_PATH_USES_DRIVE_LETTERS
michael@0: #define FILE_PATH_USES_WIN_SEPARATORS
michael@0: #endif  // OS_WIN
michael@0: 
michael@0: // An abstraction to isolate users from the differences between native
michael@0: // pathnames on different platforms.
michael@0: class FilePath {
michael@0:  public:
michael@0: #if defined(OS_POSIX)
michael@0:   // On most platforms, native pathnames are char arrays, and the encoding
michael@0:   // may or may not be specified.  On Mac OS X, native pathnames are encoded
michael@0:   // in UTF-8.
michael@0:   typedef std::string StringType;
michael@0: #elif defined(OS_WIN)
michael@0:   // On Windows, for Unicode-aware applications, native pathnames are wchar_t
michael@0:   // arrays encoded in UTF-16.
michael@0:   typedef std::wstring StringType;
michael@0: #endif  // OS_WIN
michael@0: 
michael@0:   typedef StringType::value_type CharType;
michael@0: 
michael@0:   // Null-terminated array of separators used to separate components in
michael@0:   // hierarchical paths.  Each character in this array is a valid separator,
michael@0:   // but kSeparators[0] is treated as the canonical separator and will be used
michael@0:   // when composing pathnames.
michael@0:   static const CharType kSeparators[];
michael@0: 
michael@0:   // A special path component meaning "this directory."
michael@0:   static const CharType kCurrentDirectory[];
michael@0: 
michael@0:   // A special path component meaning "the parent directory."
michael@0:   static const CharType kParentDirectory[];
michael@0: 
michael@0:   // The character used to identify a file extension.
michael@0:   static const CharType kExtensionSeparator;
michael@0: 
michael@0:   FilePath() {}
michael@0:   FilePath(const FilePath& that) : path_(that.path_) {}
michael@0:   explicit FilePath(const StringType& path) : path_(path) {}
michael@0: 
michael@0:   FilePath& operator=(const FilePath& that) {
michael@0:     path_ = that.path_;
michael@0:     return *this;
michael@0:   }
michael@0: 
michael@0:   bool operator==(const FilePath& that) const {
michael@0:     return path_ == that.path_;
michael@0:   }
michael@0: 
michael@0:   bool operator!=(const FilePath& that) const {
michael@0:     return path_ != that.path_;
michael@0:   }
michael@0: 
michael@0:   // Required for some STL containers and operations
michael@0:   bool operator<(const FilePath& that) const {
michael@0:     return path_ < that.path_;
michael@0:   }
michael@0: 
michael@0:   const StringType& value() const { return path_; }
michael@0: 
michael@0:   bool empty() const { return path_.empty(); }
michael@0: 
michael@0:   // Returns true if |character| is in kSeparators.
michael@0:   static bool IsSeparator(CharType character);
michael@0: 
michael@0:   // Returns a FilePath corresponding to the directory containing the path
michael@0:   // named by this object, stripping away the file component.  If this object
michael@0:   // only contains one component, returns a FilePath identifying
michael@0:   // kCurrentDirectory.  If this object already refers to the root directory,
michael@0:   // returns a FilePath identifying the root directory.
michael@0:   FilePath DirName() const;
michael@0: 
michael@0:   // Returns a FilePath corresponding to the last path component of this
michael@0:   // object, either a file or a directory.  If this object already refers to
michael@0:   // the root directory, returns a FilePath identifying the root directory;
michael@0:   // this is the only situation in which BaseName will return an absolute path.
michael@0:   FilePath BaseName() const;
michael@0: 
michael@0:   // Returns ".jpg" for path "C:\pics\jojo.jpg", or an empty string if
michael@0:   // the file has no extension.  If non-empty, Extension() will always start
michael@0:   // with precisely one ".".  The following code should always work regardless
michael@0:   // of the value of path.
michael@0:   // new_path = path.RemoveExtension().value().append(path.Extension());
michael@0:   // ASSERT(new_path == path.value());
michael@0:   // NOTE: this is different from the original file_util implementation which
michael@0:   // returned the extension without a leading "." ("jpg" instead of ".jpg")
michael@0:   StringType Extension() const;
michael@0: 
michael@0:   // Returns "C:\pics\jojo" for path "C:\pics\jojo.jpg"
michael@0:   // NOTE: this is slightly different from the similar file_util implementation
michael@0:   // which returned simply 'jojo'.
michael@0:   FilePath RemoveExtension() const;
michael@0: 
michael@0:   // Inserts |suffix| after the file name portion of |path| but before the
michael@0:   // extension.  Returns "" if BaseName() == "." or "..".
michael@0:   // Examples:
michael@0:   // path == "C:\pics\jojo.jpg" suffix == " (1)", returns "C:\pics\jojo (1).jpg"
michael@0:   // path == "jojo.jpg"         suffix == " (1)", returns "jojo (1).jpg"
michael@0:   // path == "C:\pics\jojo"     suffix == " (1)", returns "C:\pics\jojo (1)"
michael@0:   // path == "C:\pics.old\jojo" suffix == " (1)", returns "C:\pics.old\jojo (1)"
michael@0:   FilePath InsertBeforeExtension(const StringType& suffix) const;
michael@0: 
michael@0:   // Replaces the extension of |file_name| with |extension|.  If |file_name|
michael@0:   // does not have an extension, them |extension| is added.  If |extension| is
michael@0:   // empty, then the extension is removed from |file_name|.
michael@0:   // Returns "" if BaseName() == "." or "..".
michael@0:   FilePath ReplaceExtension(const StringType& extension) const;
michael@0: 
michael@0:   // Returns a FilePath by appending a separator and the supplied path
michael@0:   // component to this object's path.  Append takes care to avoid adding
michael@0:   // excessive separators if this object's path already ends with a separator.
michael@0:   // If this object's path is kCurrentDirectory, a new FilePath corresponding
michael@0:   // only to |component| is returned.  |component| must be a relative path;
michael@0:   // it is an error to pass an absolute path.
michael@0:   FilePath Append(const StringType& component) const WARN_UNUSED_RESULT;
michael@0:   FilePath Append(const FilePath& component) const WARN_UNUSED_RESULT;
michael@0: 
michael@0:   // Although Windows StringType is std::wstring, since the encoding it uses for
michael@0:   // paths is well defined, it can handle ASCII path components as well.
michael@0:   // Mac uses UTF8, and since ASCII is a subset of that, it works there as well.
michael@0:   // On Linux, although it can use any 8-bit encoding for paths, we assume that
michael@0:   // ASCII is a valid subset, regardless of the encoding, since many operating
michael@0:   // system paths will always be ASCII.
michael@0:   FilePath AppendASCII(const std::string& component) const WARN_UNUSED_RESULT;
michael@0: 
michael@0:   // Returns true if this FilePath contains an absolute path.  On Windows, an
michael@0:   // absolute path begins with either a drive letter specification followed by
michael@0:   // a separator character, or with two separator characters.  On POSIX
michael@0:   // platforms, an absolute path begins with a separator character.
michael@0:   bool IsAbsolute() const;
michael@0: 
michael@0:   // Returns a copy of this FilePath that does not end with a trailing
michael@0:   // separator.
michael@0:   FilePath StripTrailingSeparators() const;
michael@0: 
michael@0:   // Calls open on given ifstream instance
michael@0:   void OpenInputStream(std::ifstream &stream) const;
michael@0: 
michael@0:   // Older Chromium code assumes that paths are always wstrings.
michael@0:   // This function converts a wstring to a FilePath, and is useful to smooth
michael@0:   // porting that old code to the FilePath API.
michael@0:   // It has "Hack" in its name so people feel bad about using it.
michael@0:   // TODO(port): remove these functions.
michael@0:   static FilePath FromWStringHack(const std::wstring& wstring);
michael@0: 
michael@0:   // Older Chromium code assumes that paths are always wstrings.
michael@0:   // This function produces a wstring from a FilePath, and is useful to smooth
michael@0:   // porting that old code to the FilePath API.
michael@0:   // It has "Hack" in its name so people feel bad about using it.
michael@0:   // TODO(port): remove these functions.
michael@0:   std::wstring ToWStringHack() const;
michael@0: 
michael@0:  private:
michael@0:   // Remove trailing separators from this object.  If the path is absolute, it
michael@0:   // will never be stripped any more than to refer to the absolute root
michael@0:   // directory, so "////" will become "/", not "".  A leading pair of
michael@0:   // separators is never stripped, to support alternate roots.  This is used to
michael@0:   // support UNC paths on Windows.
michael@0:   void StripTrailingSeparatorsInternal();
michael@0: 
michael@0:   StringType path_;
michael@0: };
michael@0: 
michael@0: // Macros for string literal initialization of FilePath::CharType[].
michael@0: #if defined(OS_POSIX)
michael@0: #define FILE_PATH_LITERAL(x) x
michael@0: #elif defined(OS_WIN)
michael@0: #define FILE_PATH_LITERAL(x) L ## x
michael@0: #endif  // OS_WIN
michael@0: 
michael@0: // Implement hash function so that we can use FilePaths in hashsets and maps.
michael@0: #if defined(COMPILER_GCC) && !defined(ANDROID)
michael@0: namespace __gnu_cxx {
michael@0: 
michael@0: template<>
michael@0: struct hash<FilePath> {
michael@0:   size_t operator()(const FilePath& f) const {
michael@0:     return hash<FilePath::StringType>()(f.value());
michael@0:   }
michael@0: };
michael@0: 
michael@0: }  // namespace __gnu_cxx
michael@0: #elif defined(COMPILER_MSVC)
michael@0: namespace stdext {
michael@0: 
michael@0: inline size_t hash_value(const FilePath& f) {
michael@0:   return hash_value(f.value());
michael@0: }
michael@0: 
michael@0: }  // namespace stdext
michael@0: #endif  // COMPILER
michael@0: 
michael@0: #endif  // BASE_FILE_PATH_H_