net/base/net_util.cc - Issue 220253002: Add net/base/filename_util.h.

Unified Diff: net/base/net_util.cc

Issue 220253002: Add net/base/filename_util.h. (Closed) Base URL: svn://svn.chromium.org/chrome/trunk/src

Patch Set: Created 6 years, 9 months ago

Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.

Jump to:

View side-by-side diff with in-line comments

Index: net/base/net_util.cc

diff --git a/net/base/net_util.cc b/net/base/net_util.cc

index 0174c653f83065c3ef48fd90a31254d0d4d4b03c..bafc8354583f64b4381abbaa0168f6801a57763d 100644

--- a/net/base/net_util.cc

+++ b/net/base/net_util.cc

@@ -4,9 +4,12 @@

#include "net/base/net_util.h"

+#include <errno.h>

#include <algorithm>

#include <iterator>

#include <map>

+#include <set>

Ryan Sleevi 2014/04/07 16:38:39 What prompted these additions? Linter? I agree the

Sergey Ulanov 2014/04/07 19:44:36 It wasn't compiling without these - they were some

#include "build/build_config.h"

@@ -26,10 +29,6 @@

#endif

#include "base/basictypes.h"

-#include "base/file_util.h"

-#include "base/files/file_path.h"

-#include "base/i18n/file_util_icu.h"

-#include "base/i18n/icu_string_conversions.h"

#include "base/i18n/time_formatting.h"

#include "base/json/string_escape.h"

#include "base/lazy_instance.h"

@@ -37,7 +36,6 @@

#include "base/memory/singleton.h"

#include "base/message_loop/message_loop.h"

#include "base/metrics/histogram.h"

-#include "base/path_service.h"

#include "base/stl_util.h"

#include "base/strings/string_number_conversions.h"

#include "base/strings/string_piece.h"

@@ -62,7 +60,6 @@

#endif

#include "net/base/dns_util.h"

#include "net/base/escape.h"

-#include "net/base/mime_util.h"

#include "net/base/net_module.h"

#include "net/base/registry_controlled_domains/registry_controlled_domain.h"

#if defined(OS_WIN)

@@ -85,10 +82,6 @@ namespace {

typedef std::vector<size_t> Offsets;

-// what we prepend to get a file URL

-static const base::FilePath::CharType kFileURLPrefix[] =

- FILE_PATH_LITERAL("file:///");

// The general list of blocked ports. Will be blocked unless a specific

// protocol overrides it. (Ex: ftp can use ports 20 and 21)

static const int kRestrictedPorts[] = {

@@ -756,205 +749,6 @@ void AppendFormattedComponent(const std::string& spec,

}

-void SanitizeGeneratedFileName(base::FilePath::StringType* filename,

- bool replace_trailing) {

- const base::FilePath::CharType kReplace[] = FILE_PATH_LITERAL("-");

- if (filename->empty())

- return;

- if (replace_trailing) {

- // Handle CreateFile() stripping trailing dots and spaces on filenames

- // http://support.microsoft.com/kb/115827

- size_t length = filename->size();

- size_t pos = filename->find_last_not_of(FILE_PATH_LITERAL(" ."));

- filename->resize((pos == std::string::npos) ? 0 : (pos + 1));

- base::TrimWhitespace(*filename, base::TRIM_TRAILING, filename);

- if (filename->empty())

- return;

- size_t trimmed = length - filename->size();

- if (trimmed)

- filename->insert(filename->end(), trimmed, kReplace[0]);

- }

- base::TrimString(*filename, FILE_PATH_LITERAL("."), filename);

- if (filename->empty())

- return;

- // Replace any path information by changing path separators.

- ReplaceSubstringsAfterOffset(filename, 0, FILE_PATH_LITERAL("/"), kReplace);

- ReplaceSubstringsAfterOffset(filename, 0, FILE_PATH_LITERAL("\\"), kReplace);

-// Returns the filename determined from the last component of the path portion

-// of the URL. Returns an empty string if the URL doesn't have a path or is

-// invalid. If the generated filename is not reliable,

-// |should_overwrite_extension| will be set to true, in which case a better

-// extension should be determined based on the content type.

-std::string GetFileNameFromURL(const GURL& url,

- const std::string& referrer_charset,

- bool* should_overwrite_extension) {

- // about: and data: URLs don't have file names, but esp. data: URLs may

- // contain parts that look like ones (i.e., contain a slash). Therefore we

- // don't attempt to divine a file name out of them.

- if (!url.is_valid() || url.SchemeIs("about") || url.SchemeIs("data"))

- return std::string();

- const std::string unescaped_url_filename = UnescapeURLComponent(

- url.ExtractFileName(),

- UnescapeRule::SPACES | UnescapeRule::URL_SPECIAL_CHARS);

- // The URL's path should be escaped UTF-8, but may not be.

- std::string decoded_filename = unescaped_url_filename;

- if (!IsStringUTF8(decoded_filename)) {

- // TODO(jshin): this is probably not robust enough. To be sure, we need

- // encoding detection.

- base::string16 utf16_output;

- if (!referrer_charset.empty() &&

- base::CodepageToUTF16(unescaped_url_filename,

- referrer_charset.c_str(),

- base::OnStringConversionError::FAIL,

- &utf16_output)) {

- decoded_filename = base::UTF16ToUTF8(utf16_output);

- } else {

- decoded_filename = base::WideToUTF8(

- base::SysNativeMBToWide(unescaped_url_filename));

- }

- // If the URL contains a (possibly empty) query, assume it is a generator, and

- // allow the determined extension to be overwritten.

- *should_overwrite_extension = !decoded_filename.empty() && url.has_query();

- return decoded_filename;

-// Returns whether the specified extension is automatically integrated into the

-// windows shell.

-bool IsShellIntegratedExtension(const base::FilePath::StringType& extension) {

- base::FilePath::StringType extension_lower = StringToLowerASCII(extension);

- // http://msdn.microsoft.com/en-us/library/ms811694.aspx

- // Right-clicking on shortcuts can be magical.

- if ((extension_lower == FILE_PATH_LITERAL("local")) ||

- (extension_lower == FILE_PATH_LITERAL("lnk")))

- return true;

- // http://www.juniper.net/security/auto/vulnerabilities/vuln2612.html

- // Files become magical if they end in a CLSID, so block such extensions.

- if (!extension_lower.empty() &&

- (extension_lower[0] == FILE_PATH_LITERAL('{')) &&

- (extension_lower[extension_lower.length() - 1] == FILE_PATH_LITERAL('}')))

- return true;

- return false;

-// Returns whether the specified file name is a reserved name on windows.

-// This includes names like "com2.zip" (which correspond to devices) and

-// desktop.ini and thumbs.db which have special meaning to the windows shell.

-bool IsReservedName(const base::FilePath::StringType& filename) {

- // This list is taken from the MSDN article "Naming a file"

- // http://msdn2.microsoft.com/en-us/library/aa365247(VS.85).aspx

- // I also added clock$ because GetSaveFileName seems to consider it as a

- // reserved name too.

- static const char* const known_devices[] = {

- "con", "prn", "aux", "nul", "com1", "com2", "com3", "com4", "com5",

- "com6", "com7", "com8", "com9", "lpt1", "lpt2", "lpt3", "lpt4",

- "lpt5", "lpt6", "lpt7", "lpt8", "lpt9", "clock$"

- };

-#if defined(OS_WIN)

- std::string filename_lower = StringToLowerASCII(base::WideToUTF8(filename));

-#elif defined(OS_POSIX)

- std::string filename_lower = StringToLowerASCII(filename);

-#endif

- for (size_t i = 0; i < arraysize(known_devices); ++i) {

- // Exact match.

- if (filename_lower == known_devices[i])

- return true;

- // Starts with "DEVICE.".

- if (filename_lower.find(std::string(known_devices[i]) + ".") == 0)

- return true;

- }

- static const char* const magic_names[] = {

- // These file names are used by the "Customize folder" feature of the shell.

- "desktop.ini",

- "thumbs.db",

- };

- for (size_t i = 0; i < arraysize(magic_names); ++i) {

- if (filename_lower == magic_names[i])

- return true;

- }

- return false;

-// Examines the current extension in |file_name| and modifies it if necessary in

-// order to ensure the filename is safe. If |file_name| doesn't contain an

-// extension or if |ignore_extension| is true, then a new extension will be

-// constructed based on the |mime_type|.

-//

-// We're addressing two things here:

-//

-// 1) Usability. If there is no reliable file extension, we want to guess a

-// reasonable file extension based on the content type.

-//

-// 2) Shell integration. Some file extensions automatically integrate with the

-// shell. We block these extensions to prevent a malicious web site from

-// integrating with the user's shell.

-void EnsureSafeExtension(const std::string& mime_type,

- bool ignore_extension,

- base::FilePath* file_name) {

- // See if our file name already contains an extension.

- base::FilePath::StringType extension = file_name->Extension();

- if (!extension.empty())

- extension.erase(extension.begin()); // Erase preceding '.'.

- if ((ignore_extension || extension.empty()) && !mime_type.empty()) {

- base::FilePath::StringType preferred_mime_extension;

- std::vector<base::FilePath::StringType> all_mime_extensions;

- // The GetPreferredExtensionForMimeType call will end up going to disk. Do

- // this on another thread to avoid slowing the IO thread.

- // http://crbug.com/61827

- // TODO(asanka): Remove this ScopedAllowIO once all callers have switched

- // over to IO safe threads.

- base::ThreadRestrictions::ScopedAllowIO allow_io;

- net::GetPreferredExtensionForMimeType(mime_type, &preferred_mime_extension);

- net::GetExtensionsForMimeType(mime_type, &all_mime_extensions);

- // If the existing extension is in the list of valid extensions for the

- // given type, use it. This avoids doing things like pointlessly renaming

- // "foo.jpg" to "foo.jpeg".

- if (std::find(all_mime_extensions.begin(),

- all_mime_extensions.end(),

- extension) != all_mime_extensions.end()) {

- // leave |extension| alone

- } else if (!preferred_mime_extension.empty()) {

- extension = preferred_mime_extension;

- }

-#if defined(OS_WIN)

- static const base::FilePath::CharType default_extension[] =

- FILE_PATH_LITERAL("download");

- // Rename shell-integrated extensions.

- // TODO(asanka): Consider stripping out the bad extension and replacing it

- // with the preferred extension for the MIME type if one is available.

- if (IsShellIntegratedExtension(extension))

- extension.assign(default_extension);

-#endif

- *file_name = file_name->ReplaceExtension(extension);

-bool FilePathToString16(const base::FilePath& path, base::string16* converted) {

-#if defined(OS_WIN)

- return base::WideToUTF16(

- path.value().c_str(), path.value().size(), converted);

-#elif defined(OS_POSIX)

- std::string component8 = path.AsUTF8Unsafe();

- return !component8.empty() &&

- base::UTF8ToUTF16(component8.c_str(), component8.size(), converted);

-#endif

bool IPNumberPrefixCheck(const IPAddressNumber& ip_number,

const unsigned char* ip_prefix,

size_t prefix_length_in_bits) {

@@ -993,46 +787,6 @@ size_t GetCountOfExplicitlyAllowedPorts() {

return g_explicitly_allowed_ports.Get().size();

}

-GURL FilePathToFileURL(const base::FilePath& path) {

- // Produce a URL like "file:///C:/foo" for a regular file, or

- // "file://///server/path" for UNC. The URL canonicalizer will fix up the

- // latter case to be the canonical UNC form: "file://server/path"

- base::FilePath::StringType url_string(kFileURLPrefix);

- if (!path.IsAbsolute()) {

- base::FilePath current_dir;

- PathService::Get(base::DIR_CURRENT, &current_dir);

- url_string.append(current_dir.value());

- url_string.push_back(base::FilePath::kSeparators[0]);

- }

- url_string.append(path.value());

- // Now do replacement of some characters. Since we assume the input is a

- // literal filename, anything the URL parser might consider special should

- // be escaped here.

- // must be the first substitution since others will introduce percents as the

- // escape character