zip: Add ZipReader and rework Unzip() using the new class.

chrome/common/zip_reader.h

Index: chrome/common/zip_reader.h
diff --git a/chrome/common/zip_reader.h b/chrome/common/zip_reader.h
new file mode 100644
index 0000000000000000000000000000000000000000..613c5f7380196f0a1888ce0ef95bad3bb65f4155
--- /dev/null
+++ b/chrome/common/zip_reader.h
@@ -0,0 +1,152 @@
+// Copyright (c) 2011 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef CHROME_COMMON_ZIP_READER_H_
+#define CHROME_COMMON_ZIP_READER_H_
+#pragma once
+
+#include <string>
+
+#include "base/basictypes.h"
+#include "base/file_path.h"
+#include "base/memory/scoped_ptr.h"
+#include "base/time.h"
+#include "third_party/zlib/contrib/minizip/unzip.h"
+
+namespace zip {
+
+// This class is used for reading zip files. A typical use case of this
+// class is to scan entries in a zip file and extract them. The code will
+// look like:
+//
+//   ZipReader reader;
+//   reader.Open(zip_file_path);
+//   while (reader.HasMore()) {
+//     reader.OpenCurrentEntryInZip();
+//     reader.ExtractCurrentEntryToDirectory(output_directory_path);
+//     reader.AdvanceToNextEntry();
+//   }
+//
+// For simplicty, error checking is omitted in the example code above. The
+// production code should check return values from all of these functions.
+//
+// This calls can also be used for random access of contents in a zip file
+// using LocateAndOpenEntry().
+//
+class ZipReader {
+ public:
+  // This class represents information of an entry (file or directory) in
+  // a zip file.
+  class EntryInfo {
+   public:
+    EntryInfo(const std::string& filename_in_zip,
+              const unz_file_info& raw_file_info);
+
+    // Returns the file path. The path is usually relative like
+    // "foo/bar.txt", but if it's absolute, is_unsafe() returns true.

[Aaron Boodman, 2011/11/10 08:13:05]

Is it legit for zip files to have paths starting with '/'? If so, can you just
strip the leading '/' if it is present so that callers don't have a chance to
mess it up?

[satorux1, 2011/11/10 17:42:26]

Good question. Per the spec, it seems to legitimate (at least not prohibited).
unzip command on Linux strips the leading '/' automatically. 

I thought about stripping the leading '/' but chose not to for now, as it would
lose the original file names.

Maybe we could return the sanitized path name from file_path(), and provide
original_file_path() that returns the original unmodified file path.

I'll rethink about the file path API, when adding support for non-UTF-8 file
names, and Unicode Path Extra field.

+    const FilePath& file_path() const { return file_path_; }
+
+    // Returns the size of the original file (i.e. after uncompressed).
+    // Returns 0 if the entry is a directory.
+    const int64 original_size() const { return original_size_; }
+
+    // Returns the last modified time.
+    const base::Time last_modified() const { return last_modified_; }
+
+    // Returns true if the entry is a directory.
+    bool is_directory() const { return is_directory_; }
+
+    // Returns true if the entry is unsafe, like having ".." or invalid
+    // UTF-8 characters in its file name, or the file path is absolute.
+    bool is_unsafe() const { return is_unsafe_; }
+
+   private:
+    const FilePath file_path_;
+    int64 original_size_;
+    base::Time last_modified_;
+    bool is_directory_;
+    bool is_unsafe_;
+    DISALLOW_COPY_AND_ASSIGN(EntryInfo);
+  };
+
+  ZipReader();
+  ~ZipReader();
+
+  // Opens the zip file specified by |zip_file_path|. Returns true on
+  // success.
+  bool Open(const FilePath& zip_file_path);
+
+  // Closes the currently opened zip file. This function is called in the
+  // destructor of the class, so you usually don't need to call this.
+  void Close();
+
+  // Returns true if there is at least one entry to read. This function is
+  // used to scan entries with AdvanceToNextEntry(), like:
+  //
+  // while (reader.HasMore()) {
+  //   // Do something with the current file here.
+  //   reader.AdvanceToNextEntry();
+  // }
+  bool HasMore();
+
+  // Advances the next entry. Returns true on success.
+  bool AdvanceToNextEntry();
+
+  // Opens the current entry in the zip file. On success, returns true and
+  // updates the the current entry state (i.e. current_entry_info() is
+  // updated). This function should be called before operations over the
+  // current entry like ExtractCurrentEntryToFile().
+  //
+  // Note that there is no CloseCurrentEntryInZip(). The the current entry
+  // state is reset automatically as needed.
+  bool OpenCurrentEntryInZip();
+
+  // Locates an entry in the zip file and opens it. Returns true on
+  // success. This function internally calls OpenCurrentEntryInZip() on
+  // success. On failure, current_entry_info() becomes NULL.
+  bool LocateAndOpenEntry(const FilePath& path_in_zip);
+
+  // Extracts the current entry to the given output file path. If the
+  // current file is a directory, just creates a directory
+  // instead. Returns true on success. OpenCurrentEntryInZip() must be
+  // called beforehand.
+  //
+  // This function does not preserve the timestamp of the original entry.
+  bool ExtractCurrentEntryToFilePath(const FilePath& output_file_path);
+
+  // Extracts the current entry to the given output directory path using
+  // ExtractCurrentEntryToFilePath(). Sub directories are created as needed
+  // based on the file path of the current entry. For example, if the file
+  // path in zip is "foo/bar.txt", and the output directory is "output",
+  // "output/foo/bar.txt" will be created.
+  //
+  // Returns true on success. OpenCurrentEntryInZip() must be called
+  // beforehand.
+  bool ExtractCurrentEntryIntoDirectory(const FilePath& output_directory_path);
+
+  // Returns the current entry info. Returns NULL if the current entry is
+  // not yet opened. OpenCurrentEntryInZip() must be called beforehand.
+  EntryInfo* current_entry_info() const {
+    return current_entry_info_.get();
+  }
+
+  // Returns the number of entries in the zip file.
+  // Open() must be called beforehand.
+  int num_entries() const { return num_entries_; }
+
+ private:
+  // Resets the internal state.
+  void Reset();
+
+  unzFile zip_file_;
+  int num_entries_;
+  bool reached_end_;
+  scoped_ptr<EntryInfo> current_entry_info_;
+
+  DISALLOW_COPY_AND_ASSIGN(ZipReader);
+};
+
+}  // namespace zip
+
+#endif  // CHROME_COMMON_ZIP_READER_H_