LibreOffice/online
1
0
Fork 0
mirror of https://github.com/LibreOffice/online.git synced 2025-08-10 01:34:37 +00:00
Code Issues Packages Projects Releases Wiki Activity

wsd: improved anonymization algorithm

Browse Source 
Better hashing algorithm based on FNV-1a.
Adds support for salting the hash, and
for providing salt via configuration.

More unit-tests added, and better formatting.

Change-Id: I2be42675d0cdbaa73c3d7faed99e07631a9c20fc
Reviewed-on: https://gerrit.libreoffice.org/70034
Reviewed-by: Ashod Nakashian <ashnakash@gmail.com>
Tested-by: Ashod Nakashian <ashnakash@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/71091
This commit is contained in:
Ashod Nakashian
2019-04-14 12:21:19 -04:00
committed by Ashod Nakashian
parent 226c2fe71c
commit 919a93cd4b
7 changed files with 104 additions and 45 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
common/Util.cpp
View File

@ -304,16 +304,16 @@ namespace Util
return true;
}
std::string encodeId(const unsigned number, const int padding)
std::string encodeId(const std::uint64_t number, const int padding)
{
std::ostringstream oss;
oss << std::hex << std::setw(padding) << std::setfill('0') << number;
return oss.str();
}
unsigned decodeId(const std::string& str)
std::uint64_t decodeId(const std::string& str)
{
unsigned id = 0;
std::uint64_t id = 0;
std::stringstream ss;
ss << std::hex << str;
ss >> id;
@ -684,7 +684,7 @@ namespace Util
}
static std::map<std::string, std::string> AnonymizedStrings;
static std::atomic<unsigned> AnonymizationSalt(0);
static std::atomic<unsigned> AnonymizationCounter(0);
static std::mutex AnonymizedMutex;
void mapAnonymized(const std::string& plain, const std::string& anonymized)
@ -701,7 +701,7 @@ namespace Util
AnonymizedStrings[plain] = anonymized;
}
std::string anonymize(const std::string& text)
std::string anonymize(const std::string& text, const std::uint64_t nAnonymizationSalt)
{
{
std::unique_lock<std::mutex> lock(AnonymizedMutex);
@ -716,15 +716,26 @@ namespace Util
}
}
// We just need something irreversible, short, and
// quite simple.
std::size_t hash = 0;
// Modified 64-bit FNV-1a to add salting.
// For the algorithm and the magic numbers, see http://isthe.com/chongo/tech/comp/fnv/
std::uint64_t hash = 0xCBF29CE484222325LL;
hash ^= nAnonymizationSalt;
hash *= 0x100000001b3ULL;
for (const char c : text)
hash += c;
{
hash ^= static_cast<std::uint64_t>(c);
hash *= 0x100000001b3ULL;
}
hash ^= nAnonymizationSalt;
hash *= 0x100000001b3ULL;
// Generate the anonymized string. The '#' is to hint that it's anonymized.
// Prepend with salt to make it unique, in case we get collisions (which we will, eventually).
const std::string res = '#' + Util::encodeId(AnonymizationSalt++, 0) + '#' + Util::encodeId(hash, 0) + '#';
// Prepend with count to make it unique within a single process instance,
// in case we get collisions (which we will, eventually). N.B.: Identical
// strings likely to have different prefixes when logged in WSD process vs. Kit.
const std::string res
= '#' + Util::encodeId(AnonymizationCounter++, 0) + '#' + Util::encodeId(hash, 0) + '#';
mapAnonymized(text, res);
return res;
}
@ -739,7 +750,7 @@ namespace Util
return filename;
}
std::string anonymizeUrl(const std::string& url)
std::string anonymizeUrl(const std::string& url, const std::uint64_t nAnonymizationSalt)
{
std::string base;
std::string filename;
@ -747,7 +758,7 @@ namespace Util
std::string params;
std::tie(base, filename, ext, params) = Util::splitUrl(url);
return base + Util::anonymize(filename) + ext + params;
return base + Util::anonymize(filename, nAnonymizationSalt) + ext + params;
}
std::string getHttpTimeNow()