make non-strict match more forgiving (closes #100)

namazso · Apr 18, 2022 · 8de2854 · 8de2854
1 parent 3170ed7
commit 8de2854
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 8 deletions.
diff --git a/OpenHashTab/utl.cpp b/OpenHashTab/utl.cpp
@@ -36,11 +36,11 @@ extern "C" NTSTATUS NTAPI RtlLoadString(
 std::vector<uint8_t> utl::FindHashInString(std::wstring_view wv)
 {
   using wvmatch = std::match_results<std::wstring_view::iterator>;
-  constexpr static wchar_t regex_str[] = LR"([^\u0080-\uFFFFa-zA-Z0-9]*((?:[0-9a-fA-F]{2})+)([^\u0080-\uFFFFa-zA-Z0-9].*)?)";
+  constexpr static wchar_t regex_str[] = LR"(((?:[0-9A-F]{2} ?){2,}|(?:[0-9a-f]{2} ?){2,}))";
   static std::wregex regex{ regex_str };
 
   wvmatch pieces;
-  if (std::regex_match(begin(wv), end(wv), pieces, regex))
+  if (std::regex_search(begin(wv), end(wv), pieces, regex))
     return HashStringToBytes(std::wstring_view{ pieces[1].str() });
   return {};
 }

diff --git a/OpenHashTab/utl.h b/OpenHashTab/utl.h
@@ -79,15 +79,18 @@ namespace utl
   template <typename Char>
   std::vector<uint8_t> HashStringToBytes(std::basic_string_view<Char> str)
   {
-    if (str.size() % 2 != 0)
-      return {}; // odd
-
     std::vector<uint8_t> res;
 
-    for (size_t i = 0u; i < str.size() / 2; ++i)
+    for (size_t i = 0u; i < str.size() - 1; i += 2)
     {
-      const auto a = utl::unhex<Char>(str[i * 2]);
-      const auto b = utl::unhex<Char>(str[i * 2 + 1]);
+      while (str[i] == ' ')
+      {
+        i++;
+        if (!(i < str.size() - 1))
+          break;
+      }
+      const auto a = utl::unhex<Char>(str[i]);
+      const auto b = utl::unhex<Char>(str[i + 1]);
       if (a == 0xFF || b == 0xFF)
         return {}; // invalid
       res.push_back(a << 4 | b);