Move FileURLToFilePath into platform specific files. Most of the logic in this function is specific to windows' file system.

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@1211 0039d316-1c4b-4281-b951-d872f2087c98

4 files changed, 219 insertions, 76 deletions

diff --git a/net/base/net_util.cc b/net/base/net_util.cc
index 906c16a..4e7fab7 100644
--- a/net/base/net_util.cc
+++ b/net/base/net_util.cc
@@ -675,77 +675,6 @@ GURL FilePathToFileURL(const std::wstring& file_path) {
 #endif
 }
 
-bool FileURLToFilePath(const GURL& url, std::wstring* file_path) {
-  file_path->clear();
-
-  if (!url.is_valid())
-    return false;
-
-  std::string path;
-  std::string host = url.host();
-  if (host.empty()) {
-    // URL contains no host, the path is the filename. In this case, the path
-    // will probably be preceeded with a slash, as in "/C:/foo.txt", so we
-    // trim out that here.
-    path = url.path();
-    size_t first_non_slash = path.find_first_not_of("/\\");
-    if (first_non_slash != std::string::npos && first_non_slash > 0)
-      path.erase(0, first_non_slash);
-  } else {
-    // URL contains a host: this means it's UNC. We keep the preceeding slash
-    // on the path.
-    path = "\\\\";
-    path.append(host);
-    path.append(url.path());
-  }
-
-  if (path.empty())
-    return false;
-  std::replace(path.begin(), path.end(), '/', '\\');
-
-  // GURL stores strings as percent-encoded UTF-8, this will undo if possible.
-  path = UnescapeURLComponent(path,
-      UnescapeRule::SPACES | UnescapeRule::URL_SPECIAL_CHARS);
-
-  if (!IsStringUTF8(path.c_str())) {
-    // Not UTF-8, assume encoding is native codepage and we're done. We know we
-    // are giving the conversion function a nonempty string, and it may fail if
-    // the given string is not in the current encoding and give us an empty
-    // string back. We detect this and report failure.
-    *file_path = base::SysNativeMBToWide(path);
-    return !file_path->empty();
-  }
-  file_path->assign(UTF8ToWide(path));
-
-  // Now we have an unescaped filename, but are still not sure about its
-  // encoding. For example, each character could be part of a UTF-8 string.
-  if (file_path->empty() || !IsString8Bit(*file_path)) {
-    // assume our 16-bit encoding is correct if it won't fit into an 8-bit
-    // string
-    return true;
-  }
-
-  // Convert our narrow string into the native wide path.
-  std::string narrow;
-  if (!WideToLatin1(*file_path, &narrow)) {
-    NOTREACHED() << "Should have filtered out non-8-bit strings above.";
-    return false;
-  }
-  if (IsStringUTF8(narrow.c_str())) {
-    // Our string actually looks like it could be UTF-8, convert to 8-bit
-    // UTF-8 and then to the corresponding wide string.
-    *file_path = UTF8ToWide(narrow);
-  } else {
-    // Our wide string contains only 8-bit characters and it's not UTF-8, so
-    // we assume it's in the native codepage.
-    *file_path = base::SysNativeMBToWide(narrow);
-  }
-
-  // Fail if 8-bit -> wide conversion failed and gave us an empty string back
-  // (we already filtered out empty strings above).
-  return !file_path->empty();
-}
-
 std::wstring GetSpecificHeader(const std::wstring& headers,
                                const std::wstring& name) {
   return GetSpecificHeaderT(headers, name);
diff --git a/net/base/net_util_posix.cc b/net/base/net_util_posix.cc
new file mode 100644
index 0000000..03a006a
--- /dev/null
+++ b/net/base/net_util_posix.cc
@@ -0,0 +1,70 @@
+// Copyright 2008, Google Inc.
+// All rights reserved.
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//    * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//    * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following disclaimer
+// in the documentation and/or other materials provided with the
+// distribution.
+//    * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+#include "net/base/net_util.h"
+
+#include "base/string_util.h"
+#include "googleurl/src/gurl.h"
+#include "net/base/escape.h"
+
+namespace net {
+
+bool FileURLToFilePath(const GURL& url, std::wstring* file_path) {
+  file_path->clear();
+
+  if (!url.is_valid())
+    return false;
+
+  // Firefox seems to ignore the "host" of a file url if there is one. That is,
+  // file://foo/bar.txt maps to /bar.txt.
+  std::string path = url.path();
+
+  if (path.empty())
+    return false;
+
+  // GURL stores strings as percent-encoded 8-bit, this will undo if possible.
+  path = UnescapeURLComponent(path,
+      UnescapeRule::SPACES | UnescapeRule::URL_SPECIAL_CHARS);
+
+  // Collapse multiple path slashes into a single path slash.
+  std::string new_path;
+  do {
+    new_path = path;
+    ReplaceSubstringsAfterOffset(&new_path, 0, "//", "/");
+    path.swap(new_path);
+  } while (new_path != path);
+
+  // TODO(tc): This should actually be 8-bit to wide.  We may lose data if the
+  // string isn't UTF-8.
+  file_path->assign(UTF8ToWide(path));
+
+  return !file_path->empty();
+}
+
+}  // namespace net
diff --git a/net/base/net_util_unittest.cc b/net/base/net_util_unittest.cc
index fcf99ba..f18c642 100644
--- a/net/base/net_util_unittest.cc
+++ b/net/base/net_util_unittest.cc
@@ -83,16 +83,22 @@ struct SuggestedFilenameCase {
 
 }  // anonymous namespace
 
-#ifdef OS_WIN
-// TODO: Determine how we want windows file paths and unc paths to be
-// normalized on posix systems.
 TEST(NetUtilTest, FileURLConversion) {
   // a list of test file names and the corresponding URLs
   const FileCase round_trip_cases[] = {
+#if defined(OS_WIN)
     {L"C:\\foo\\bar.txt", L"file:///C:/foo/bar.txt"},
     {L"\\\\some computer\\foo\\bar.txt", L"file://some%20computer/foo/bar.txt"}, // UNC
     {L"D:\\Name;with%some symbols*#", L"file:///D:/Name%3Bwith%25some%20symbols*%23"},
     {L"D:\\Chinese\\\x6240\x6709\x4e2d\x6587\x7f51\x9875.doc", L"file:///D:/Chinese/%E6%89%80%E6%9C%89%E4%B8%AD%E6%96%87%E7%BD%91%E9%A1%B5.doc"},
+#elif defined(OS_POSIX)
+    {L"/foo/bar.txt", L"file:///foo/bar.txt"},
+    {L"/foo/BAR.txt", L"file:///foo/BAR.txt"},
+    {L"/C:/foo/bar.txt", L"file:///C:/foo/bar.txt"},
+    {L"/some computer/foo/bar.txt", L"file:///some%20computer/foo/bar.txt"},
+    {L"/Name;with%some symbols*#", L"file:///Name%3Bwith%25some%20symbols*%23"},
+    {L"/Chinese/\x6240\x6709\x4e2d\x6587\x7f51\x9875.doc", L"file:///Chinese/%E6%89%80%E6%9C%89%E4%B8%AD%E6%96%87%E7%BD%91%E9%A1%B5.doc"},
+#endif
   };
 
   // First, we'll test that we can round-trip all of the above cases of URLs
@@ -110,6 +116,7 @@ TEST(NetUtilTest, FileURLConversion) {
 
   // Test that various file: URLs get decoded into the correct file type
   FileCase url_cases[] = {
+#if defined(OS_WIN)
     {L"C:\\foo\\bar.txt", L"file:c|/foo\\bar.txt"},
     {L"C:\\foo\\bar.txt", L"file:/c:/foo/bar.txt"},
     {L"\\\\foo\\bar.txt", L"file://foo\\bar.txt"},
@@ -118,6 +125,28 @@ TEST(NetUtilTest, FileURLConversion) {
     {L"\\\\foo\\bar.txt", L"file:/foo/bar.txt"},
     {L"\\\\foo\\bar.txt", L"file://foo\\bar.txt"},
     {L"C:\\foo\\bar.txt", L"file:\\\\\\c:/foo/bar.txt"},
+#elif defined(OS_POSIX)
+    {L"/c:/foo/bar.txt", L"file:/c:/foo/bar.txt"},
+    {L"/c:/foo/bar.txt", L"file:///c:/foo/bar.txt"},
+    {L"/foo/bar.txt", L"file:/foo/bar.txt"},
+    {L"/c:/foo/bar.txt", L"file:\\\\\\c:/foo/bar.txt"},
+    {L"/foo/bar.txt", L"file:foo/bar.txt"},
+    {L"/foo/bar.txt", L"file://foo/bar.txt"},
+    {L"/foo/bar.txt", L"file:///foo/bar.txt"},
+    {L"/foo/bar.txt", L"file:////foo/bar.txt"},
+    {L"/foo/bar.txt", L"file:////foo//bar.txt"},
+    {L"/foo/bar.txt", L"file:////foo///bar.txt"},
+    {L"/foo/bar.txt", L"file:////foo////bar.txt"},
+    {L"/c:/foo/bar.txt", L"file:\\\\\\c:/foo/bar.txt"},
+    {L"/c:/foo/bar.txt", L"file:c:/foo/bar.txt"},
+    // We get these wrong because GURL turns back slashes into forward
+    // slashes.
+    //{L"/foo%5Cbar.txt", L"file://foo\\bar.txt"},
+    //{L"/c|/foo%5Cbar.txt", L"file:c|/foo\\bar.txt"},
+    //{L"/foo%5Cbar.txt", L"file://foo\\bar.txt"},
+    //{L"/foo%5Cbar.txt", L"file:////foo\\bar.txt"},
+    //{L"/foo%5Cbar.txt", L"file://foo\\bar.txt"},
+#endif
   };
   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(url_cases); i++) {
     net::FileURLToFilePath(GURL(WideToUTF16(url_cases[i].url)), &output);
@@ -125,9 +154,14 @@ TEST(NetUtilTest, FileURLConversion) {
   }
 
   // Here, we test that UTF-8 encoded strings get decoded properly, even when
-  // they might be stored with wide characters
+  // they might be stored with wide characters.  On posix systems, just treat
+  // this as a stream of bytes.
   const wchar_t utf8[] = L"file:///d:/Chinese/\xe6\x89\x80\xe6\x9c\x89\xe4\xb8\xad\xe6\x96\x87\xe7\xbd\x91\xe9\xa1\xb5.doc";
+#if defined(OS_WIN)
   const wchar_t wide[] = L"D:\\Chinese\\\x6240\x6709\x4e2d\x6587\x7f51\x9875.doc";
+#elif defined(OS_POSIX)
+  const wchar_t wide[] = L"/d:/Chinese/\xe6\x89\x80\xe6\x9c\x89\xe4\xb8\xad\xe6\x96\x87\xe7\xbd\x91\xe9\xa1\xb5.doc";
+#endif
   EXPECT_TRUE(net::FileURLToFilePath(GURL(WideToUTF16(utf8)), &output));
   EXPECT_EQ(std::wstring(wide), output);
 
@@ -145,7 +179,6 @@ TEST(NetUtilTest, FileURLConversion) {
   // Test that if a file URL is malformed, we get a failure
   EXPECT_FALSE(net::FileURLToFilePath(GURL("filefoobar"), &output));
 }
-#endif
 
 // Just a bunch of fake headers.
 const wchar_t* google_headers =
diff --git a/net/base/net_util_win.cc b/net/base/net_util_win.cc
new file mode 100644
index 0000000..9295b7e
--- /dev/null
+++ b/net/base/net_util_win.cc
@@ -0,0 +1,111 @@
+// Copyright 2008, Google Inc.
+// All rights reserved.
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//    * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//    * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following disclaimer
+// in the documentation and/or other materials provided with the
+// distribution.
+//    * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+#include "net/base/net_util.h"
+
+#include "base/string_piece.h"
+#include "base/string_util.h"
+#include "base/sys_string_conversions.h"
+#include "googleurl/src/gurl.h"
+#include "net/base/escape.h"
+
+namespace net {
+
+bool FileURLToFilePath(const GURL& url, std::wstring* file_path) {
+  file_path->clear();
+
+  if (!url.is_valid())
+    return false;
+
+  std::string path;
+  std::string host = url.host();
+  if (host.empty()) {
+    // URL contains no host, the path is the filename. In this case, the path
+    // will probably be preceeded with a slash, as in "/C:/foo.txt", so we
+    // trim out that here.
+    path = url.path();
+    size_t first_non_slash = path.find_first_not_of("/\\");
+    if (first_non_slash != std::string::npos && first_non_slash > 0)
+      path.erase(0, first_non_slash);
+  } else {
+    // URL contains a host: this means it's UNC. We keep the preceeding slash
+    // on the path.
+    path = "\\\\";
+    path.append(host);
+    path.append(url.path());
+  }
+
+  if (path.empty())
+    return false;
+  std::replace(path.begin(), path.end(), '/', '\\');
+
+  // GURL stores strings as percent-encoded UTF-8, this will undo if possible.
+  path = UnescapeURLComponent(path,
+      UnescapeRule::SPACES | UnescapeRule::URL_SPECIAL_CHARS);
+
+  if (!IsStringUTF8(path.c_str())) {
+    // Not UTF-8, assume encoding is native codepage and we're done. We know we
+    // are giving the conversion function a nonempty string, and it may fail if
+    // the given string is not in the current encoding and give us an empty
+    // string back. We detect this and report failure.
+    *file_path = base::SysNativeMBToWide(path);
+    return !file_path->empty();
+  }
+  file_path->assign(UTF8ToWide(path));
+
+  // Now we have an unescaped filename, but are still not sure about its
+  // encoding. For example, each character could be part of a UTF-8 string.
+  if (file_path->empty() || !IsString8Bit(*file_path)) {
+    // assume our 16-bit encoding is correct if it won't fit into an 8-bit
+    // string
+    return true;
+  }
+
+  // Convert our narrow string into the native wide path.
+  std::string narrow;
+  if (!WideToLatin1(*file_path, &narrow)) {
+    NOTREACHED() << "Should have filtered out non-8-bit strings above.";
+    return false;
+  }
+  if (IsStringUTF8(narrow.c_str())) {
+    // Our string actually looks like it could be UTF-8, convert to 8-bit
+    // UTF-8 and then to the corresponding wide string.
+    *file_path = UTF8ToWide(narrow);
+  } else {
+    // Our wide string contains only 8-bit characters and it's not UTF-8, so
+    // we assume it's in the native codepage.
+    *file_path = base::SysNativeMBToWide(narrow);
+  }
+
+  // Fail if 8-bit -> wide conversion failed and gave us an empty string back
+  // (we already filtered out empty strings above).
+  return !file_path->empty();
+}
+
+}  // namespace net