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

chrome/common/zip_reader.cc

+// 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.
+
+#include "chrome/common/zip_reader.h"
+
+#include "base/file_util.h"
+#include "base/logging.h"
+#include "base/string_util.h"
+#include "base/utf_string_conversions.h"
+#include "chrome/common/zip_internal.h"
+#include "net/base/file_stream.h"
+#include "third_party/zlib/contrib/minizip/unzip.h"
+#if defined(OS_WIN)
+#include "third_party/zlib/contrib/minizip/iowin32.h"
+#endif
+
+namespace zip {
+
+// TODO(satorux): The implementation assumes that file names in zip files
+// are encoded in UTF-8. This is true for zip files created by Zip()
+// function in zip.h, but not true for user-supplied random zip files.
+ZipReader::FileInfo::FileInfo(const std::string& file_name_in_zip,
+                              const unz_file_info& raw_file_info)
+    : file_path_(FilePath::FromUTF8Unsafe(file_name_in_zip)),
+      is_directory_(false) {
+  original_size_ = raw_file_info.uncompressed_size;
+
+  // Directory entries in zip files end with "/".
+  is_directory_ = EndsWith(file_name_in_zip, "/", false);
+  // Check the file name here for directory traversal issues. In the name of

[Aaron Boodman, 2011/11/09 08:00:40]

Nit: add blank line before this one.

[satorux1, 2011/11/10 08:00:56]

Done.

[satorux1, 2011/11/10 08:00:56]

Done.

+  // simplicity and security, we might reject a valid file name such as "a..b".
+  is_unsafe_ = file_name_in_zip.find("..") != std::string::npos;
+
+  // We also consider that the file name is unsafe, if it's invalid UTF-8.
+  string16 file_name_utf16;
+  if (!UTF8ToUTF16(file_name_in_zip.data(), file_name_in_zip.size(),
+                   &file_name_utf16))

[Aaron Boodman, 2011/11/09 08:00:40]

If a conditional wraps, it should use curly braces.

[satorux1, 2011/11/10 08:00:56]

Done.

[satorux1, 2011/11/10 08:00:56]

Done.

+    is_unsafe_ = true;
+
+  // Construct the last modified time. The timezone info is not present in
+  // zip files, so we construct the time as local time.
+  base::Time::Exploded exploded_time = {};  // Zero-clear.
+  exploded_time.year = raw_file_info.tmu_date.tm_year;
+  // The month in zip file is 0-based, whereas ours is 1-based.
+  exploded_time.month = raw_file_info.tmu_date.tm_mon + 1;
+  exploded_time.day_of_month = raw_file_info.tmu_date.tm_mday;
+  exploded_time.hour = raw_file_info.tmu_date.tm_hour;
+  exploded_time.minute = raw_file_info.tmu_date.tm_min;
+  exploded_time.second = raw_file_info.tmu_date.tm_sec;
+  exploded_time.millisecond = 0;
+  last_modified_ = base::Time::FromLocalExploded(exploded_time);
+}
+
+// Don't forget to update Close() when you add new code in the constructor.

[Aaron Boodman, 2011/11/09 08:00:40]

Instead of comments like this, it is better to have a Reset() private method
that both Close() and the constructor call that centralizes member
initialization.

[satorux1, 2011/11/10 08:00:56]

You are right. Done.

+ZipReader::ZipReader()
+    : zip_file_(NULL),
+      num_files_(0),
+      reached_end_(false),
+      current_file_info_(NULL) {
+}
+
+ZipReader::~ZipReader() {
+  Close();
+}
+
+bool ZipReader::Open(const FilePath& zip_file_path) {
+  DCHECK(!zip_file_);
+
+  // Use of "Unsafe" function does not look good, but there is no way to do
+  // this safely on Linux. See file_util.h for details.
+  zip_file_ = internal::OpenForUnzipping(zip_file_path.AsUTF8Unsafe());
+  if (!zip_file_) {
+    return false;
+  }
+
+  unz_global_info zip_info = {};  // Zero-clear.

[Aaron Boodman, 2011/11/09 08:00:40]

Unnecessary comment.

[satorux1, 2011/11/10 08:00:56]

Done.

+  if (unzGetGlobalInfo(zip_file_, &zip_info) != UNZ_OK) {
+    return false;
+  }
+  num_files_ = zip_info.number_entry;
+  if (num_files_ < 0)
+    return false;
+
+  // We are already at the end if the zip file is empty.
+  reached_end_ = (num_files_ == 0);
+  return true;
+}
+
+void ZipReader::Close() {
+  if (zip_file_) {
+    unzClose(zip_file_);
+    zip_file_ = NULL;
+  }
+  num_files_ = 0;
+  reached_end_ = false;
+  current_file_info_.reset();
+}
+
+bool ZipReader::HasMore() {
+  return !reached_end_;
+}
+
+bool ZipReader::AdvanceToNextFile() {
+  DCHECK(zip_file_);
+
+  // Should not go further if we already reached the end.
+  if (reached_end_)
+    return false;
+
+  unz_file_pos position = {};
+  if (unzGetFilePos(zip_file_, &position) != UNZ_OK)
+    return false;
+  const int current_file_index = position.num_of_file;
+  // If we are currently at the last of the file, then the next position
+  // is the end of the zip file, so mark that we reached the end.
+  if (current_file_index + 1 == num_files_) {
+    reached_end_ = true;
+  } else {
+    DCHECK_LT(current_file_index + 1, num_files_);
+    if (unzGoToNextFile(zip_file_) != UNZ_OK) {
+      return false;
+    }
+  }
+  current_file_info_.reset();
+  return true;
+}
+
+bool ZipReader::OpenCurrentFileInZip() {
+  DCHECK(zip_file_);
+
+  unz_file_info raw_file_info = {};  // Zero-clear.
+  char raw_file_name_in_zip[internal::kZipMaxPath] = {};
+  const int result = unzGetCurrentFileInfo(zip_file_,
+                                           &raw_file_info,
+                                           raw_file_name_in_zip,
+                                           sizeof(raw_file_name_in_zip) - 1,
+                                           NULL,  // extraField.
+                                           0,  // extraFieldBufferSize.
+                                           NULL,  // szComment.
+                                           0);  // commentBufferSize.
+  if (result != UNZ_OK)
+    return NULL;
+  if (raw_file_name_in_zip[0] == '\0')
+    return NULL;
+  current_file_info_.reset(new FileInfo(raw_file_name_in_zip, raw_file_info));
+  return true;
+}
+
+bool ZipReader::LocateAndOpenFile(const FilePath& path_in_zip) {
+  DCHECK(zip_file_);
+
+  current_file_info_.reset();
+  reached_end_ = false;
+  const int kDefaultCaseSensivityOfOS = 0;
+  const int result = unzLocateFile(zip_file_,
+                                   path_in_zip.AsUTF8Unsafe().c_str(),
+                                   kDefaultCaseSensivityOfOS);
+  if (result != UNZ_OK)
+    return false;
+
+  // Then Open the file.
+  return OpenCurrentFileInZip();
+}
+
+bool ZipReader::ExtractCurrentFileToFile(
+    const FilePath& output_file_path) {
+  DCHECK(zip_file_);
+
+  // If this is a directory, just create it and return.
+  if (current_file_info()->is_directory())
+    return file_util::CreateDirectory(output_file_path);
+
+  const int open_result = unzOpenCurrentFile(zip_file_);
+  if (open_result != UNZ_OK)
+    return false;
+
+  // We can't rely on parent directory entries being specified in the
+  // zip, so we make sure they are created.
+  FilePath output_dir_path = output_file_path.DirName();
+  if (!file_util::CreateDirectory(output_dir_path))
+    return false;
+
+  net::FileStream stream;
+  const int flags = (base::PLATFORM_FILE_CREATE_ALWAYS |
+                     base::PLATFORM_FILE_WRITE);
+  if (stream.Open(output_file_path, flags) != 0)
+    return false;
+
+  bool success = true;  // This becomes false when something bad happens.
+  while (true) {
+    char buf[internal::kZipBufSize];
+    const int num_bytes_read = unzReadCurrentFile(zip_file_, buf,
+                                                  internal::kZipBufSize);
+    if (num_bytes_read == 0) {
+      // Reached the end of the file.
+      break;
+    } else if (num_bytes_read < 0) {
+      // If num_bytes_read < 0, then it's a specific UNZ_* error code.
+      success = false;
+      break;
+    } else if (num_bytes_read > 0) {
+      // Some data is read. Write it to the output file.
+      if (num_bytes_read != stream.Write(buf, num_bytes_read,
+                                         net::CompletionCallback())) {
+        success = false;
+        break;
+      }
+    }
+  }
+
+  stream.Close();
+  unzCloseCurrentFile(zip_file_);
+  return success;
+}
+
+bool ZipReader::ExtractCurrentFileToDirectory(
+    const FilePath& output_directory_path) {
+  DCHECK(current_file_info_.get());
+
+  FilePath output_file_path = output_directory_path.Append(
+      current_file_info()->file_path());
+  return ExtractCurrentFileToFile(output_file_path);
+}
+
+}  // namespace zip