setting regexp_max_matches_per_row instead of 3rd argument

src/Core/Settings.h

@@ -475,6 +475,7 @@ class IColumn;
     M(Bool, query_plan_enable_optimizations, true, "Apply optimizations to query plan", 0) \
     M(UInt64, query_plan_max_optimizations_to_apply, 10000, "Limit the total number of optimizations applied to query plan. If zero, ignored. If limit reached, throw exception", 0) \
     M(Bool, query_plan_filter_push_down, true, "Allow to push down filter by predicate query plan step", 0) \
+    M(UInt64, regexp_max_matches_per_row, 1000, "Max matches of any single regexp per row, used to safeguard extractAllGroupsHorizontal() against consuming too much memory with greedy RE.", 0) \
     \
     M(UInt64, limit, 0, "Limit on read rows from the most 'end' result for select query, default 0 means no limit length", 0) \
     M(UInt64, offset, 0, "Offset on read rows from the most 'end' result for select query", 0) \

src/Functions/extractAllGroups.h

@@ -7,6 +7,8 @@
 #include <Functions/FunctionHelpers.h>
 #include <Functions/IFunction.h>
 #include <Functions/Regexps.h>
+#include <Interpreters/Context.h>
+#include <Core/Settings.h>
 
 #include <memory>
 #include <string>
@@ -47,16 +49,21 @@ enum class ExtractAllGroupsResultKind
 template <typename Impl>
 class FunctionExtractAllGroups : public IFunction
 {
+    ContextPtr context;
+
 public:
     static constexpr auto Kind = Impl::Kind;
     static constexpr auto name = Impl::Name;
 
-    static FunctionPtr create(ContextPtr) { return std::make_shared<FunctionExtractAllGroups>(); }
+    FunctionExtractAllGroups(ContextPtr context_)
+        : context(context_)
+    {}
+
+    static FunctionPtr create(ContextPtr context) { return std::make_shared<FunctionExtractAllGroups>(context); }
 
     String getName() const override { return name; }
 
-    size_t getNumberOfArguments() const override { return Kind == ExtractAllGroupsResultKind::HORIZONTAL ? 0 : 2; }
-    bool isVariadic() const override { return Kind == ExtractAllGroupsResultKind::HORIZONTAL; }
+    size_t getNumberOfArguments() const override { return 2; }
 
     bool useDefaultImplementationForConstants() const override { return true; }
     ColumnNumbers getArgumentsThatAreAlwaysConstant() const override { return {1}; }
@@ -67,13 +74,7 @@ public:
             {"haystack", isStringOrFixedString, nullptr, "const String or const FixedString"},
             {"needle", isStringOrFixedString, isColumnConst, "const String or const FixedString"},
         };
-        FunctionArgumentDescriptors optional_args;
-        if constexpr (Kind == ExtractAllGroupsResultKind::HORIZONTAL)
-        {
-            optional_args.push_back(FunctionArgumentDescriptor{"max_matches_per_row", isUnsignedInteger, isColumnConst, "const Unsigned Int"});
-        }
-
-        validateFunctionArgumentTypes(*this, arguments, args, optional_args);
+        validateFunctionArgumentTypes(*this, arguments, args);
 
         /// Two-dimensional array of strings, each `row` of top array represents matching groups.
         return std::make_shared<DataTypeArray>(std::make_shared<DataTypeArray>(std::make_shared<DataTypeString>()));
@@ -154,9 +155,8 @@ public:
         }
         else
         {
-            /// Additional limit to fail fast on supposedly incorrect usage, arbitrary value.
-            static constexpr size_t MAX_MATCHES_PER_ROW = 1000;
-            const auto max_matches_per_row = arguments.size() >= 3 ? arguments[2].column->getUInt(0) : MAX_MATCHES_PER_ROW;
+            /// Additional limit to fail fast on supposedly incorrect usage.
+            const auto max_matches_per_row = context->getSettingsRef().regexp_max_matches_per_row;
 
             PODArray<StringPiece, 0> all_matches;
             /// Number of times RE matched on each row of haystack column.

tests/queries/0_stateless/01246_extractAllGroupsHorizontal.sql

@@ -5,13 +5,11 @@ SELECT extractAllGroupsHorizontal('hello', 123);  --{serverError 43} invalid arg
 SELECT extractAllGroupsHorizontal(123, 'world');  --{serverError 43}  invalid argument type
 SELECT extractAllGroupsHorizontal('hello world', '((('); --{serverError 427}  invalid re
 SELECT extractAllGroupsHorizontal('hello world', materialize('\\w+')); --{serverError 44} non-cons needle
-SELECT extractAllGroupsHorizontal('hello world', '(\\w+)', 'foobar'); --{serverError 43} invalid argument type
-SELECT extractAllGroupsHorizontal('hello world', '(\\w+)', materialize(10)); --{serverError 44} non-const max_matches_per_row
 SELECT extractAllGroupsHorizontal('hello world', '\\w+');  -- { serverError 36 } 0 groups
-SELECT extractAllGroupsHorizontal('hello world', '(\\w+)', 0);  -- { serverError 128 } to many groups matched per row
-SELECT extractAllGroupsHorizontal('hello world', '(\\w+)', 1);  -- { serverError 128 } to many groups matched per row
+SELECT extractAllGroupsHorizontal('hello world', '(\\w+)') SETTINGS regexp_max_matches_per_row = 0;  -- { serverError 128 } to many groups matched per row
+SELECT extractAllGroupsHorizontal('hello world', '(\\w+)') SETTINGS regexp_max_matches_per_row = 1;  -- { serverError 128 } to many groups matched per row
 
-SELECT extractAllGroupsHorizontal('hello world', '(\\w+)', 1000000000) FORMAT Null; -- users now can set limit bigger than previous 1000 matches per row
+SELECT extractAllGroupsHorizontal('hello world', '(\\w+)') SETTINGS regexp_max_matches_per_row = 1000000 FORMAT Null; -- users now can set limit bigger than previous 1000 matches per row
 
 SELECT '1 group, multiple matches, String and FixedString';
 SELECT extractAllGroupsHorizontal('hello world', '(\\w+)');

tests/queries/0_stateless/01246_extractAllGroupsVertical.sql

@@ -5,7 +5,6 @@ SELECT extractAllGroupsVertical('hello', 123);  --{serverError 43} invalid argum
 SELECT extractAllGroupsVertical(123, 'world');  --{serverError 43}  invalid argument type
 SELECT extractAllGroupsVertical('hello world', '((('); --{serverError 427}  invalid re
 SELECT extractAllGroupsVertical('hello world', materialize('\\w+')); --{serverError 44} non-const needle
-SELECT extractAllGroupsVertical('hello world', '(\\w+)', 123); --{serverError 42} only 2 arguments
 SELECT extractAllGroupsVertical('hello world', '\\w+'); -- { serverError 36 } 0 groups
 
 SELECT '1 group, multiple matches, String and FixedString';