Hyperscan better usage of scratch space and error handling

dbms/src/Common/ErrorCodes.cpp

@@ -421,6 +421,7 @@ namespace ErrorCodes
     extern const int UNKNOWN_PROTOBUF_FORMAT = 444;
     extern const int CANNOT_MPROTECT = 445;
     extern const int FUNCTION_NOT_ALLOWED = 446;
+    extern const int HYPERSCAN_CANNOT_SCAN_TEXT = 447;
 
     extern const int KEEPER_EXCEPTION = 999;
     extern const int POCO_EXCEPTION = 1000;

dbms/src/Functions/FunctionsStringRegex.cpp

@@ -40,6 +40,7 @@ namespace ErrorCodes
     extern const int ILLEGAL_COLUMN;
     extern const int TOO_MANY_BYTES;
     extern const int NOT_IMPLEMENTED;
+    extern const int HYPERSCAN_CANNOT_SCAN_TEXT;
 }
 
 /// Is the LIKE expression reduced to finding a substring in a string?
@@ -289,10 +290,10 @@ struct MultiMatchAnyImpl
 #if USE_HYPERSCAN
         const auto & hyperscan_regex = MultiRegexps::get<FindAnyIndex, MultiSearchDistance>(needles, edit_distance);
         hs_scratch_t * scratch = nullptr;
-        hs_error_t err = hs_alloc_scratch(hyperscan_regex->get(), &scratch);
+        hs_error_t err = hs_clone_scratch(hyperscan_regex->getScratch(), &scratch);
 
         if (err != HS_SUCCESS)
-            throw Exception("Could not allocate scratch space for hyperscan", ErrorCodes::CANNOT_ALLOCATE_MEMORY);
+            throw Exception("Could not clone scratch space for hyperscan", ErrorCodes::CANNOT_ALLOCATE_MEMORY);
 
         MultiRegexps::ScratchPtr smart_scratch(scratch);
 
@@ -316,14 +317,16 @@ struct MultiMatchAnyImpl
             if (length > std::numeric_limits<UInt32>::max())
                 throw Exception("Too long string to search", ErrorCodes::TOO_MANY_BYTES);
             res[i] = 0;
-            hs_scan(
-                hyperscan_regex->get(),
+            err = hs_scan(
+                hyperscan_regex->getDB(),
                 reinterpret_cast<const char *>(haystack_data.data()) + offset,
                 length,
                 0,
                 smart_scratch.get(),
                 on_match,
                 &res[i]);
+            if (err != HS_SUCCESS)
+                throw Exception("Failed to scan with hyperscan", ErrorCodes::HYPERSCAN_CANNOT_SCAN_TEXT);
             offset = haystack_offsets[i];
         }
 #else

dbms/src/Functions/Regexps.h

@@ -84,8 +84,20 @@ namespace MultiRegexps
     };
 
     using CompilerError = std::unique_ptr<hs_compile_error_t, HyperscanDeleter<decltype(&hs_free_compile_error), &hs_free_compile_error>>;
-    using ScratchPtr = std::unique_ptr<hs_scratch_t, DB::MultiRegexps::HyperscanDeleter<decltype(&hs_free_scratch), &hs_free_scratch>>;
-    using Regexps = std::unique_ptr<hs_database_t, HyperscanDeleter<decltype(&hs_free_database), &hs_free_database>>;
+    using ScratchPtr = std::unique_ptr<hs_scratch_t, HyperscanDeleter<decltype(&hs_free_scratch), &hs_free_scratch>>;
+    using DataBasePtr = std::unique_ptr<hs_database_t, HyperscanDeleter<decltype(&hs_free_database), &hs_free_database>>;
+
+    class Regexps
+    {
+    public:
+        Regexps(hs_database_t * db_, hs_scratch_t * scratch_) : db{db_}, scratch{scratch_} {}
+
+        hs_database_t * getDB() const { return db.get(); };
+        hs_scratch_t * getScratch() const { return scratch.get(); };
+    private:
+        DataBasePtr db;
+        ScratchPtr scratch;
+    };
 
     using Pool = ObjectPoolMap<Regexps, std::pair<std::vector<String>, std::optional<UInt32>>>;
 
@@ -183,7 +195,13 @@ namespace MultiRegexps
 
             ProfileEvents::increment(ProfileEvents::RegexpCreated);
 
-            return new Regexps{db};
+            hs_scratch_t * scratch = nullptr;
+            err = hs_alloc_scratch(db, &scratch);
+
+            if (err != HS_SUCCESS)
+                throw Exception("Could not allocate scratch space for hyperscan", ErrorCodes::CANNOT_ALLOCATE_MEMORY);
+
+            return new Regexps{db, scratch};
         });
     }
 }