Merge 4992ec031d into c0c83236b6

Mask azure connection string sensitive info

contrib/libpqxx

@@ -1 +1 @@
-Subproject commit c995193a3a14d71f4711f1f421f65a1a1db64640
+Subproject commit 41e4c331564167cca97ad6eccbd5b8879c2ca044

contrib/libpqxx-cmake/CMakeLists.txt

@@ -1,9 +1,9 @@
 set (LIBRARY_DIR "${ClickHouse_SOURCE_DIR}/contrib/libpqxx")
 
 set (SRCS
-    "${LIBRARY_DIR}/src/strconv.cxx"
     "${LIBRARY_DIR}/src/array.cxx"
     "${LIBRARY_DIR}/src/binarystring.cxx"
+    "${LIBRARY_DIR}/src/blob.cxx"
     "${LIBRARY_DIR}/src/connection.cxx"
     "${LIBRARY_DIR}/src/cursor.cxx"
     "${LIBRARY_DIR}/src/encodings.cxx"
@@ -12,59 +12,25 @@ set (SRCS
     "${LIBRARY_DIR}/src/field.cxx"
     "${LIBRARY_DIR}/src/largeobject.cxx"
     "${LIBRARY_DIR}/src/notification.cxx"
+    "${LIBRARY_DIR}/src/params.cxx"
     "${LIBRARY_DIR}/src/pipeline.cxx"
     "${LIBRARY_DIR}/src/result.cxx"
     "${LIBRARY_DIR}/src/robusttransaction.cxx"
+    "${LIBRARY_DIR}/src/row.cxx"
     "${LIBRARY_DIR}/src/sql_cursor.cxx"
+    "${LIBRARY_DIR}/src/strconv.cxx"
     "${LIBRARY_DIR}/src/stream_from.cxx"
     "${LIBRARY_DIR}/src/stream_to.cxx"
     "${LIBRARY_DIR}/src/subtransaction.cxx"
+    "${LIBRARY_DIR}/src/time.cxx"
     "${LIBRARY_DIR}/src/transaction.cxx"
     "${LIBRARY_DIR}/src/transaction_base.cxx"
-    "${LIBRARY_DIR}/src/row.cxx"
-    "${LIBRARY_DIR}/src/params.cxx"
     "${LIBRARY_DIR}/src/util.cxx"
     "${LIBRARY_DIR}/src/version.cxx"
+    "${LIBRARY_DIR}/src/wait.cxx"
 )
 
-# Need to explicitly include each header file, because in the directory include/pqxx there are also files
-# like just 'array'. So if including the whole directory with `target_include_directories`, it will make
-# conflicts with all includes of <array>.
-set (HDRS
-    "${LIBRARY_DIR}/include/pqxx/array.hxx"
-    "${LIBRARY_DIR}/include/pqxx/params.hxx"
-    "${LIBRARY_DIR}/include/pqxx/binarystring.hxx"
-    "${LIBRARY_DIR}/include/pqxx/composite.hxx"
-    "${LIBRARY_DIR}/include/pqxx/connection.hxx"
-    "${LIBRARY_DIR}/include/pqxx/cursor.hxx"
-    "${LIBRARY_DIR}/include/pqxx/dbtransaction.hxx"
-    "${LIBRARY_DIR}/include/pqxx/errorhandler.hxx"
-    "${LIBRARY_DIR}/include/pqxx/except.hxx"
-    "${LIBRARY_DIR}/include/pqxx/field.hxx"
-    "${LIBRARY_DIR}/include/pqxx/isolation.hxx"
-    "${LIBRARY_DIR}/include/pqxx/largeobject.hxx"
-    "${LIBRARY_DIR}/include/pqxx/nontransaction.hxx"
-    "${LIBRARY_DIR}/include/pqxx/notification.hxx"
-    "${LIBRARY_DIR}/include/pqxx/pipeline.hxx"
-    "${LIBRARY_DIR}/include/pqxx/prepared_statement.hxx"
-    "${LIBRARY_DIR}/include/pqxx/result.hxx"
-    "${LIBRARY_DIR}/include/pqxx/robusttransaction.hxx"
-    "${LIBRARY_DIR}/include/pqxx/row.hxx"
-    "${LIBRARY_DIR}/include/pqxx/separated_list.hxx"
-    "${LIBRARY_DIR}/include/pqxx/strconv.hxx"
-    "${LIBRARY_DIR}/include/pqxx/stream_from.hxx"
-    "${LIBRARY_DIR}/include/pqxx/stream_to.hxx"
-    "${LIBRARY_DIR}/include/pqxx/subtransaction.hxx"
-    "${LIBRARY_DIR}/include/pqxx/transaction.hxx"
-    "${LIBRARY_DIR}/include/pqxx/transaction_base.hxx"
-    "${LIBRARY_DIR}/include/pqxx/types.hxx"
-    "${LIBRARY_DIR}/include/pqxx/util.hxx"
-    "${LIBRARY_DIR}/include/pqxx/version.hxx"
-    "${LIBRARY_DIR}/include/pqxx/zview.hxx"
-)
-
-add_library(_libpqxx ${SRCS} ${HDRS})
-
+add_library(_libpqxx ${SRCS})
 target_link_libraries(_libpqxx PUBLIC ch_contrib::libpq)
 target_include_directories (_libpqxx SYSTEM BEFORE PUBLIC "${LIBRARY_DIR}/include")
 

src/Core/NamesAndTypes.cpp

@@ -151,6 +151,15 @@ Names NamesAndTypesList::getNames() const
     return res;
 }
 
+NameSet NamesAndTypesList::getNameSet() const
+{
+    NameSet res;
+    res.reserve(size());
+    for (const NameAndTypePair & column : *this)
+        res.insert(column.name);
+    return res;
+}
+
 DataTypes NamesAndTypesList::getTypes() const
 {
     DataTypes res;

src/Core/NamesAndTypes.h

@@ -100,6 +100,7 @@ public:
     void getDifference(const NamesAndTypesList & rhs, NamesAndTypesList & deleted, NamesAndTypesList & added) const;
 
     Names getNames() const;
+    NameSet getNameSet() const;
     DataTypes getTypes() const;
 
     /// Remove columns which names are not in the `names`.

src/Interpreters/HashJoin/HashJoin.cpp

@@ -338,11 +338,8 @@ size_t HashJoin::getTotalRowCount() const
     return res;
 }
 
-size_t HashJoin::getTotalByteCount() const
+void HashJoin::doDebugAsserts() const
 {
-    if (!data)
-        return 0;
-
 #ifndef NDEBUG
     size_t debug_blocks_allocated_size = 0;
     for (const auto & block : data->blocks)
@@ -360,6 +357,14 @@ size_t HashJoin::getTotalByteCount() const
         throw Exception(ErrorCodes::LOGICAL_ERROR, "data->blocks_nullmaps_allocated_size != debug_blocks_nullmaps_allocated_size ({} != {})",
                         data->blocks_nullmaps_allocated_size, debug_blocks_nullmaps_allocated_size);
 #endif
+}
+
+size_t HashJoin::getTotalByteCount() const
+{
+    if (!data)
+        return 0;
+
+    doDebugAsserts();
 
     size_t res = 0;
 
@@ -544,9 +549,11 @@ bool HashJoin::addBlockToJoin(const Block & source_block_, bool check_limits)
             have_compressed = true;
         }
 
+        doDebugAsserts();
         data->blocks_allocated_size += block_to_save.allocatedBytes();
         data->blocks.emplace_back(std::move(block_to_save));
         Block * stored_block = &data->blocks.back();
+        doDebugAsserts();
 
         if (rows)
             data->empty = false;
@@ -634,9 +641,11 @@ bool HashJoin::addBlockToJoin(const Block & source_block_, bool check_limits)
 
             if (!flag_per_row && !is_inserted)
             {
+                doDebugAsserts();
                 LOG_TRACE(log, "Skipping inserting block with {} rows", rows);
                 data->blocks_allocated_size -= stored_block->allocatedBytes();
                 data->blocks.pop_back();
+                doDebugAsserts();
             }
 
             if (!check_limits)
@@ -683,6 +692,8 @@ void HashJoin::shrinkStoredBlocksToFit(size_t & total_bytes_in_join, bool force_
 
     for (auto & stored_block : data->blocks)
     {
+        doDebugAsserts();
+
         size_t old_size = stored_block.allocatedBytes();
         stored_block = stored_block.shrinkToFit();
         size_t new_size = stored_block.allocatedBytes();
@@ -700,6 +711,8 @@ void HashJoin::shrinkStoredBlocksToFit(size_t & total_bytes_in_join, bool force_
         else
             /// Sometimes after clone resized block can be bigger than original
             data->blocks_allocated_size += new_size - old_size;
+
+        doDebugAsserts();
     }
 
     auto new_total_bytes_in_join = getTotalByteCount();
@@ -1416,7 +1429,13 @@ void HashJoin::tryRerangeRightTableDataImpl(Map & map [[maybe_unused]])
         };
         BlocksList sorted_blocks;
         visit_rows_map(sorted_blocks, map);
+        doDebugAsserts();
         data->blocks.swap(sorted_blocks);
+        size_t new_blocks_allocated_size = 0;
+        for (const auto & block : data->blocks)
+            new_blocks_allocated_size += block.allocatedBytes();
+        data->blocks_allocated_size = new_blocks_allocated_size;
+        doDebugAsserts();
     }
 }
 

src/Interpreters/HashJoin/HashJoin.h

@@ -470,6 +470,7 @@ private:
     void tryRerangeRightTableData() override;
     template <JoinKind KIND, typename Map, JoinStrictness STRICTNESS>
     void tryRerangeRightTableDataImpl(Map & map);
+    void doDebugAsserts() const;
 };
 
 }

src/Interpreters/MutationsInterpreter.cpp

@@ -857,7 +857,10 @@ void MutationsInterpreter::prepare(bool dry_run)
         else if (command.type == MutationCommand::MATERIALIZE_TTL)
         {
             mutation_kind.set(MutationKind::MUTATE_OTHER);
-            if (materialize_ttl_recalculate_only)
+            bool suitable_for_ttl_optimization = source.getMergeTreeData()->getSettings()->ttl_only_drop_parts
+                && metadata_snapshot->hasOnlyRowsTTL();
+
+            if (materialize_ttl_recalculate_only || suitable_for_ttl_optimization)
             {
                 // just recalculate ttl_infos without remove expired data
                 auto all_columns_vec = all_columns.getNames();

src/Parsers/ASTFunction.cpp

@@ -722,7 +722,14 @@ void ASTFunction::formatImplWithoutAlias(const FormatSettings & settings, Format
                         assert_cast<const ASTFunction *>(argument.get())->arguments->children[0]->formatImpl(settings, state, nested_dont_need_parens);
                         settings.ostr << (settings.hilite ? hilite_operator : "") << " = " << (settings.hilite ? hilite_none : "");
                     }
-                    settings.ostr << "'[HIDDEN]'";
+                    if (!secret_arguments.replacement.empty())
+                    {
+                        settings.ostr << "'" << secret_arguments.replacement << "'";
+                    }
+                    else
+                    {
+                        settings.ostr << "'[HIDDEN]'";
+                    }
                     if (size <= secret_arguments.start + secret_arguments.count && !secret_arguments.are_named)
                         break; /// All other arguments should also be hidden.
                     continue;

src/Parsers/FunctionSecretArgumentsFinder.h

@@ -1,6 +1,7 @@
 #pragma once
 
 #include <Common/KnownObjectNames.h>
+#include <Common/re2.h>
 #include <Core/QualifiedTableName.h>
 #include <base/defines.h>
 #include <boost/algorithm/string/predicate.hpp>
@@ -49,6 +50,11 @@ public:
         bool are_named = false; /// Arguments like `password = 'password'` are considered as named arguments.
         /// E.g. "headers" in `url('..', headers('foo' = '[HIDDEN]'))`
         std::vector<std::string> nested_maps;
+        /// Full replacement of an argument. Only supported when count is 1, otherwise all arguments will be replaced with this string.
+        /// It's needed in cases when we don't want to hide the entire parameter, but some part of it, e.g. "connection_string" in
+        /// `azureBlobStorage('DefaultEndpointsProtocol=https;AccountKey=secretkey;...', ...)` should be replaced with
+        /// `azureBlobStorage('DefaultEndpointsProtocol=https;AccountKey=[HIDDEN];...', ...)`.
+        std::string replacement;
 
         bool hasSecrets() const
         {
@@ -74,6 +80,7 @@ protected:
             result.are_named = argument_is_named;
         }
         chassert(index >= result.start); /// We always check arguments consecutively
+        chassert(result.replacement.empty()); /// We shouldn't use replacement with masking other arguments
         result.count = index + 1 - result.start;
         if (!argument_is_named)
             result.are_named = false;
@@ -199,32 +206,39 @@ protected:
 
     void findAzureBlobStorageFunctionSecretArguments(bool is_cluster_function)
     {
-        /// azureBlobStorage('cluster_name', 'conn_string/storage_account_url', ...) has 'conn_string/storage_account_url' as its second argument.
+        /// azureBlobStorageCluster('cluster_name', 'conn_string/storage_account_url', ...) has 'conn_string/storage_account_url' as its second argument.
         size_t url_arg_idx = is_cluster_function ? 1 : 0;
 
         if (!is_cluster_function && isNamedCollectionName(0))
         {
             /// azureBlobStorage(named_collection, ..., account_key = 'account_key', ...)
+            if (maskAzureConnectionString(-1, true, 1))
+                return;
             findSecretNamedArgument("account_key", 1);
             return;
         }
         else if (is_cluster_function && isNamedCollectionName(1))
         {
             /// azureBlobStorageCluster(cluster, named_collection, ..., account_key = 'account_key', ...)
+            if (maskAzureConnectionString(-1, true, 2))
+                return;
             findSecretNamedArgument("account_key", 2);
             return;
         }
 
-        /// We should check other arguments first because we don't need to do any replacement in case storage_account_url is not used
-        /// azureBlobStorage(connection_string|storage_account_url, container_name, blobpath, account_name, account_key, format, compression, structure)
-        /// azureBlobStorageCluster(cluster, connection_string|storage_account_url, container_name, blobpath, [account_name, account_key, format, compression, structure])
+        if (maskAzureConnectionString(url_arg_idx))
+            return;
+
+        /// We should check other arguments first because we don't need to do any replacement in case of
+        /// azureBlobStorage(connection_string|storage_account_url, container_name, blobpath, format, [account_name, account_key, ...])
+        /// azureBlobStorageCluster(cluster, connection_string|storage_account_url, container_name, blobpath, format, [account_name, account_key, ...])
         size_t count = function->arguments->size();
         if ((url_arg_idx + 4 <= count) && (count <= url_arg_idx + 7))
         {
-            String second_arg;
-            if (tryGetStringFromArgument(url_arg_idx + 3, &second_arg))
+            String fourth_arg;
+            if (tryGetStringFromArgument(url_arg_idx + 3, &fourth_arg))
             {
-                if (second_arg == "auto" || KnownFormatNames::instance().exists(second_arg))
+                if (fourth_arg == "auto" || KnownFormatNames::instance().exists(fourth_arg))
                     return; /// The argument after 'url' is a format: s3('url', 'format', ...)
             }
         }
@@ -234,6 +248,40 @@ protected:
             markSecretArgument(url_arg_idx + 4);
     }
 
+    bool maskAzureConnectionString(ssize_t url_arg_idx, bool argument_is_named = false, size_t start = 0)
+    {
+        String url_arg;
+        if (argument_is_named)
+        {
+            url_arg_idx = findNamedArgument(&url_arg, "connection_string", start);
+            if (url_arg_idx == -1 || url_arg.empty())
+                url_arg_idx = findNamedArgument(&url_arg, "storage_account_url", start);
+            if (url_arg_idx == -1 || url_arg.empty())
+                return false;
+        }
+        else
+        {
+            if (!tryGetStringFromArgument(url_arg_idx, &url_arg))
+                return false;
+        }
+
+        if (!url_arg.starts_with("http"))
+        {
+            static re2::RE2 account_key_pattern = "AccountKey=.*?(;|$)";
+            if (RE2::Replace(&url_arg, account_key_pattern, "AccountKey=[HIDDEN]\\1"))
+            {
+                chassert(result.count == 0); /// We shouldn't use replacement with masking other arguments
+                result.start = url_arg_idx;
+                result.are_named = argument_is_named;
+                result.count = 1;
+                result.replacement = url_arg;
+                return true;
+            }
+        }
+
+        return false;
+    }
+
     void findURLSecretArguments()
     {
         if (!isNamedCollectionName(0))
@@ -513,8 +561,9 @@ protected:
         return function->arguments->at(arg_idx)->isIdentifier();
     }
 
-    /// Looks for a secret argument with a specified name. This function looks for arguments in format `key=value` where the key is specified.
-    void findSecretNamedArgument(const std::string_view & key, size_t start = 0)
+    /// Looks for an argument with a specified name. This function looks for arguments in format `key=value` where the key is specified.
+    /// Returns -1 if no argument was found.
+    ssize_t findNamedArgument(String * res, const std::string_view & key, size_t start = 0)
     {
         for (size_t i = start; i < function->arguments->size(); ++i)
         {
@@ -531,8 +580,22 @@ protected:
                 continue;
 
             if (found_key == key)
-                markSecretArgument(i, /* argument_is_named= */ true);
+            {
+                tryGetStringFromArgument(*equals_func->arguments->at(1), res);
+                return i;
+            }
         }
+
+        return -1;
+    }
+
+    /// Looks for a secret argument with a specified name. This function looks for arguments in format `key=value` where the key is specified.
+    /// If the argument is found, it is marked as a secret.
+    void findSecretNamedArgument(const std::string_view & key, size_t start = 0)
+    {
+        ssize_t arg_idx = findNamedArgument(nullptr, key, start);
+        if (arg_idx >= 0)
+            markSecretArgument(arg_idx, /* argument_is_named= */ true);
     }
 };
 

src/Storages/MergeTree/MutateTask.cpp

@@ -119,6 +119,7 @@ static void splitAndModifyMutationCommands(
     const MutationCommands & commands,
     MutationCommands & for_interpreter,
     MutationCommands & for_file_renames,
+    bool suitable_for_ttl_optimization,
     LoggerPtr log)
 {
     auto part_columns = part->getColumnsDescription();
@@ -128,6 +129,7 @@ static void splitAndModifyMutationCommands(
     {
         NameSet mutated_columns;
         NameSet dropped_columns;
+        NameSet ignored_columns;
 
         for (const auto & command : commands)
         {
@@ -153,6 +155,15 @@ static void splitAndModifyMutationCommands(
                 for_interpreter.push_back(command);
                 for (const auto & [column_name, expr] : command.column_to_update_expression)
                     mutated_columns.emplace(column_name);
+
+                if (command.type == MutationCommand::Type::MATERIALIZE_TTL && suitable_for_ttl_optimization)
+                {
+                    for (const auto & col : part_columns)
+                    {
+                        if (!mutated_columns.contains(col.name))
+                            ignored_columns.emplace(col.name);
+                    }
+                }
             }
             else if (command.type == MutationCommand::Type::DROP_INDEX
                      || command.type == MutationCommand::Type::DROP_PROJECTION
@@ -213,7 +224,7 @@ static void splitAndModifyMutationCommands(
         /// from disk we just don't read dropped columns
         for (const auto & column : part_columns)
         {
-            if (!mutated_columns.contains(column.name))
+            if (!mutated_columns.contains(column.name) && !ignored_columns.contains(column.name))
             {
                 if (!metadata_snapshot->getColumns().has(column.name) && !part->storage.getVirtualsPtr()->has(column.name))
                 {
@@ -1884,6 +1895,82 @@ private:
     std::unique_ptr<PartMergerWriter> part_merger_writer_task{nullptr};
 };
 
+/*
+ * Decorator that'll drop expired parts by replacing them with empty ones.
+ * Main use case (only use case for now) is to decorate `MutateSomePartColumnsTask`,
+ * which is used to recalculate TTL. If the part is expired, this class will replace it with
+ * an empty one.
+ *
+ * Triggered when `ttl_only_drop_parts` is set and the only TTL is rows TTL.
+ * */
+class ExecutableTaskDropTTLExpiredPartsDecorator : public IExecutableTask
+{
+public:
+    explicit ExecutableTaskDropTTLExpiredPartsDecorator(
+        std::unique_ptr<IExecutableTask> executable_task_,
+        MutationContextPtr ctx_
+        )
+        : executable_task(std::move(executable_task_)), ctx(ctx_) {}
+
+    void onCompleted() override { throw Exception(ErrorCodes::LOGICAL_ERROR, "Not implemented"); }
+    StorageID getStorageID() const override { throw Exception(ErrorCodes::LOGICAL_ERROR, "Not implemented"); }
+    Priority getPriority() const override { throw Exception(ErrorCodes::LOGICAL_ERROR, "Not implemented"); }
+    String getQueryId() const override { throw Exception(ErrorCodes::LOGICAL_ERROR, "Not implemented"); }
+
+    bool executeStep() override
+    {
+        switch (state)
+        {
+            case State::NEED_EXECUTE:
+            {
+                if (executable_task->executeStep())
+                    return true;
+
+                if (isRowsMaxTTLExpired())
+                    replacePartWithEmpty();
+
+                state = State::SUCCESS;
+                return true;
+            }
+            case State::SUCCESS:
+            {
+                return false;
+            }
+        }
+        return false;
+    }
+
+private:
+    enum class State
+    {
+        NEED_EXECUTE,
+
+        SUCCESS
+    };
+
+    State state{State::NEED_EXECUTE};
+
+    std::unique_ptr<IExecutableTask> executable_task;
+    MutationContextPtr ctx;
+
+    bool isRowsMaxTTLExpired() const
+    {
+        const auto ttl = ctx->new_data_part->ttl_infos.table_ttl;
+        return ttl.max && ttl.max <= ctx->time_of_mutation;
+    }
+
+    void replacePartWithEmpty()
+    {
+        MergeTreePartInfo part_info = ctx->new_data_part->info;
+        part_info.level += 1;
+
+        MergeTreePartition partition = ctx->new_data_part->partition;
+        std::string part_name = ctx->new_data_part->getNewName(part_info);
+
+        auto [mutable_empty_part, _] = ctx->data->createEmptyPart(part_info, partition, part_name, ctx->txn);
+        ctx->new_data_part = std::move(mutable_empty_part);
+    }
+};
 
 MutateTask::MutateTask(
     FutureMergedMutatedPartPtr future_part_,
@@ -2122,6 +2209,7 @@ bool MutateTask::prepare()
     context_for_reading->setSetting("max_streams_for_merge_tree_reading", Field(0));
     context_for_reading->setSetting("read_from_filesystem_cache_if_exists_otherwise_bypass_cache", 1);
 
+    bool suitable_for_ttl_optimization = ctx->metadata_snapshot->hasOnlyRowsTTL() && ctx->data->getSettings()->ttl_only_drop_parts;
     MutationHelpers::splitAndModifyMutationCommands(
         ctx->source_part,
         ctx->metadata_snapshot,
@@ -2129,6 +2217,7 @@ bool MutateTask::prepare()
         ctx->commands_for_part,
         ctx->for_interpreter,
         ctx->for_file_renames,
+        suitable_for_ttl_optimization,
         ctx->log);
 
     ctx->stage_progress = std::make_unique<MergeStageProgress>(1.0);
@@ -2235,7 +2324,12 @@ bool MutateTask::prepare()
         /// The blobs have to be removed along with the part, this temporary part owns them and does not share them yet.
         ctx->new_data_part->remove_tmp_policy = IMergeTreeDataPart::BlobsRemovalPolicyForTemporaryParts::REMOVE_BLOBS;
 
-        task = std::make_unique<MutateAllPartColumnsTask>(ctx);
+        bool drop_expired_parts = suitable_for_ttl_optimization && !ctx->data->getSettings()->materialize_ttl_recalculate_only;
+        if (drop_expired_parts)
+            task = std::make_unique<ExecutableTaskDropTTLExpiredPartsDecorator>(std::make_unique<MutateAllPartColumnsTask>(ctx), ctx);
+        else
+            task = std::make_unique<MutateAllPartColumnsTask>(ctx);
+
         ProfileEvents::increment(ProfileEvents::MutationAllPartColumns);
     }
     else /// TODO: check that we modify only non-key columns in this case.
@@ -2295,7 +2389,12 @@ bool MutateTask::prepare()
         /// Keeper has to be asked with unlock request to release the references to the blobs
         ctx->new_data_part->remove_tmp_policy = IMergeTreeDataPart::BlobsRemovalPolicyForTemporaryParts::ASK_KEEPER;
 
-        task = std::make_unique<MutateSomePartColumnsTask>(ctx);
+        bool drop_expired_parts = suitable_for_ttl_optimization && !ctx->data->getSettings()->materialize_ttl_recalculate_only;
+        if (drop_expired_parts)
+            task = std::make_unique<ExecutableTaskDropTTLExpiredPartsDecorator>(std::make_unique<MutateSomePartColumnsTask>(ctx), ctx);
+        else
+            task = std::make_unique<MutateSomePartColumnsTask>(ctx);
+
         ProfileEvents::increment(ProfileEvents::MutationSomePartColumns);
     }
 

src/Storages/ObjectStorage/Azure/Configuration.cpp

@@ -223,7 +223,7 @@ void StorageAzureConfiguration::fromAST(ASTs & engine_args, ContextPtr context,
         {
             account_name = fourth_arg;
             account_key = checkAndGetLiteralArgument<String>(engine_args[4], "account_key");
-            auto sixth_arg = checkAndGetLiteralArgument<String>(engine_args[5], "format/account_name");
+            auto sixth_arg = checkAndGetLiteralArgument<String>(engine_args[5], "format/structure");
             if (is_format_arg(sixth_arg))
             {
                 format = sixth_arg;
@@ -257,10 +257,10 @@ void StorageAzureConfiguration::fromAST(ASTs & engine_args, ContextPtr context,
     }
     else if (with_structure && engine_args.size() == 8)
     {
-        auto fourth_arg = checkAndGetLiteralArgument<String>(engine_args[3], "format/account_name");
+        auto fourth_arg = checkAndGetLiteralArgument<String>(engine_args[3], "account_name");
         account_name = fourth_arg;
         account_key = checkAndGetLiteralArgument<String>(engine_args[4], "account_key");
-        auto sixth_arg = checkAndGetLiteralArgument<String>(engine_args[5], "format/account_name");
+        auto sixth_arg = checkAndGetLiteralArgument<String>(engine_args[5], "format");
         if (!is_format_arg(sixth_arg))
             throw Exception(ErrorCodes::BAD_ARGUMENTS, "Unknown format {}", sixth_arg);
         format = sixth_arg;

src/Storages/ObjectStorage/StorageObjectStorageSource.cpp

@@ -131,7 +131,7 @@ std::shared_ptr<StorageObjectStorageSource::IIterator> StorageObjectStorageSourc
     else
     {
         ConfigurationPtr copy_configuration = configuration->clone();
-        auto filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns);
+        auto filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns, local_context);
         if (filter_dag)
         {
             auto keys = configuration->getPaths();
@@ -142,7 +142,7 @@ std::shared_ptr<StorageObjectStorageSource::IIterator> StorageObjectStorageSourc
 
             VirtualColumnUtils::buildSetsForDAG(*filter_dag, local_context);
             auto actions = std::make_shared<ExpressionActions>(std::move(*filter_dag));
-            VirtualColumnUtils::filterByPathOrFile(keys, paths, actions, virtual_columns);
+            VirtualColumnUtils::filterByPathOrFile(keys, paths, actions, virtual_columns, local_context);
             copy_configuration->setPaths(keys);
         }
 
@@ -489,6 +489,7 @@ StorageObjectStorageSource::GlobIterator::GlobIterator(
     , virtual_columns(virtual_columns_)
     , throw_on_zero_files_match(throw_on_zero_files_match_)
     , read_keys(read_keys_)
+    , local_context(context_)
     , file_progress_callback(file_progress_callback_)
 {
     if (configuration->isNamespaceWithGlobs())
@@ -510,7 +511,7 @@ StorageObjectStorageSource::GlobIterator::GlobIterator(
         }
 
         recursive = key_with_globs == "/**";
-        if (auto filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns))
+        if (auto filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns, local_context))
         {
             VirtualColumnUtils::buildSetsForDAG(*filter_dag, getContext());
             filter_expr = std::make_shared<ExpressionActions>(std::move(*filter_dag));
@@ -585,7 +586,7 @@ StorageObjectStorage::ObjectInfoPtr StorageObjectStorageSource::GlobIterator::ne
                 for (const auto & object_info : new_batch)
                     paths.push_back(getUniqueStoragePathIdentifier(*configuration, *object_info, false));
 
-                VirtualColumnUtils::filterByPathOrFile(new_batch, paths, filter_expr, virtual_columns);
+                VirtualColumnUtils::filterByPathOrFile(new_batch, paths, filter_expr, virtual_columns, local_context);
 
                 LOG_TEST(logger, "Filtered files: {} -> {}", paths.size(), new_batch.size());
             }

src/Storages/ObjectStorage/StorageObjectStorageSource.h

@@ -220,6 +220,7 @@ private:
     bool is_finished = false;
     bool first_iteration = true;
     std::mutex next_mutex;
+    const ContextPtr local_context;
 
     std::function<void(FileProgress)> file_progress_callback;
 };

src/Storages/StorageFile.cpp

@@ -1141,13 +1141,13 @@ StorageFileSource::FilesIterator::FilesIterator(
 {
     std::optional<ActionsDAG> filter_dag;
     if (!distributed_processing && !archive_info && !files.empty())
-        filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns);
+        filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns, context_);
 
     if (filter_dag)
     {
         VirtualColumnUtils::buildSetsForDAG(*filter_dag, context_);
         auto actions = std::make_shared<ExpressionActions>(std::move(*filter_dag));
-        VirtualColumnUtils::filterByPathOrFile(files, files, actions, virtual_columns);
+        VirtualColumnUtils::filterByPathOrFile(files, files, actions, virtual_columns, context_);
     }
 }
 

src/Storages/StorageInMemoryMetadata.cpp

@@ -260,6 +260,12 @@ bool StorageInMemoryMetadata::hasAnyTableTTL() const
     return hasAnyMoveTTL() || hasRowsTTL() || hasAnyRecompressionTTL() || hasAnyGroupByTTL() || hasAnyRowsWhereTTL();
 }
 
+bool StorageInMemoryMetadata::hasOnlyRowsTTL() const
+{
+    bool has_any_other_ttl = hasAnyMoveTTL() || hasAnyRecompressionTTL() || hasAnyGroupByTTL() || hasAnyRowsWhereTTL() || hasAnyColumnTTL();
+    return hasRowsTTL() && !has_any_other_ttl;
+}
+
 TTLColumnsDescription StorageInMemoryMetadata::getColumnTTLs() const
 {
     return column_ttls_by_name;

src/Storages/StorageInMemoryMetadata.h

@@ -144,6 +144,9 @@ struct StorageInMemoryMetadata
     /// Returns true if there is set table TTL, any column TTL or any move TTL.
     bool hasAnyTTL() const { return hasAnyColumnTTL() || hasAnyTableTTL(); }
 
+    /// Returns true if only rows TTL is set, not even rows where.
+    bool hasOnlyRowsTTL() const;
+
     /// Common tables TTLs (for rows and moves).
     TTLTableDescription getTableTTLs() const;
     bool hasAnyTableTTL() const;

src/Storages/StorageURL.cpp

@@ -227,7 +227,7 @@ public:
 
         std::optional<ActionsDAG> filter_dag;
         if (!uris.empty())
-            filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns);
+            filter_dag = VirtualColumnUtils::createPathAndFileFilterDAG(predicate, virtual_columns, context);
 
         if (filter_dag)
         {
@@ -238,7 +238,7 @@ public:
 
             VirtualColumnUtils::buildSetsForDAG(*filter_dag, context);
             auto actions = std::make_shared<ExpressionActions>(std::move(*filter_dag));
-            VirtualColumnUtils::filterByPathOrFile(uris, paths, actions, virtual_columns);
+            VirtualColumnUtils::filterByPathOrFile(uris, paths, actions, virtual_columns, context);
         }
     }
 

src/Storages/VirtualColumnUtils.cpp

@@ -1,5 +1,6 @@
 #include <memory>
 #include <stack>
+#include <unordered_set>
 #include <Core/NamesAndTypes.h>
 #include <Core/TypeId.h>
 
@@ -46,6 +47,7 @@
 #include "Functions/IFunction.h"
 #include "Functions/IFunctionAdaptors.h"
 #include "Functions/indexHint.h"
+#include <IO/ReadBufferFromString.h>
 #include <Interpreters/convertFieldToType.h>
 #include <Parsers/makeASTForLogicalFunction.h>
 #include <Columns/ColumnSet.h>
@@ -124,9 +126,18 @@ void filterBlockWithExpression(const ExpressionActionsPtr & actions, Block & blo
     }
 }
 
+NamesAndTypesList getCommonVirtualsForFileLikeStorage()
+{
+    return {{"_path", std::make_shared<DataTypeLowCardinality>(std::make_shared<DataTypeString>())},
+            {"_file", std::make_shared<DataTypeLowCardinality>(std::make_shared<DataTypeString>())},
+            {"_size", makeNullable(std::make_shared<DataTypeUInt64>())},
+            {"_time", makeNullable(std::make_shared<DataTypeDateTime>())},
+            {"_etag", std::make_shared<DataTypeLowCardinality>(std::make_shared<DataTypeString>())}};
+}
+
 NameSet getVirtualNamesForFileLikeStorage()
 {
-    return {"_path", "_file", "_size", "_time", "_etag"};
+    return getCommonVirtualsForFileLikeStorage().getNameSet();
 }
 
 std::unordered_map<std::string, std::string> parseHivePartitioningKeysAndValues(const String & path)
@@ -154,8 +165,10 @@ VirtualColumnsDescription getVirtualsForFileLikeStorage(ColumnsDescription & sto
 {
     VirtualColumnsDescription desc;
 
-    auto add_virtual = [&](const auto & name, const auto & type)
+    auto add_virtual = [&](const NameAndTypePair & pair)
     {
+        const auto & name = pair.getNameInStorage();
+        const auto & type = pair.getTypeInStorage();
         if (storage_columns.has(name))
         {
             if (!context->getSettingsRef().use_hive_partitioning)
@@ -172,11 +185,8 @@ VirtualColumnsDescription getVirtualsForFileLikeStorage(ColumnsDescription & sto
         desc.addEphemeral(name, type, "");
     };
 
-    add_virtual("_path", std::make_shared<DataTypeLowCardinality>(std::make_shared<DataTypeString>()));
-    add_virtual("_file", std::make_shared<DataTypeLowCardinality>(std::make_shared<DataTypeString>()));
-    add_virtual("_size", makeNullable(std::make_shared<DataTypeUInt64>()));
-    add_virtual("_time", makeNullable(std::make_shared<DataTypeDateTime>()));
-    add_virtual("_etag", std::make_shared<DataTypeLowCardinality>(std::make_shared<DataTypeString>()));
+    for (const auto & item : getCommonVirtualsForFileLikeStorage())
+        add_virtual(item);
 
     if (context->getSettingsRef().use_hive_partitioning)
     {
@@ -188,16 +198,16 @@ VirtualColumnsDescription getVirtualsForFileLikeStorage(ColumnsDescription & sto
             if (type == nullptr)
                 type = std::make_shared<DataTypeString>();
             if (type->canBeInsideLowCardinality())
-                add_virtual(item.first, std::make_shared<DataTypeLowCardinality>(type));
+                add_virtual({item.first, std::make_shared<DataTypeLowCardinality>(type)});
             else
-                add_virtual(item.first, type);
+                add_virtual({item.first, type});
         }
     }
 
     return desc;
 }
 
-static void addPathAndFileToVirtualColumns(Block & block, const String & path, size_t idx)
+static void addPathAndFileToVirtualColumns(Block & block, const String & path, size_t idx, const FormatSettings & format_settings, bool use_hive_partitioning)
 {
     if (block.has("_path"))
         block.getByName("_path").column->assumeMutableRef().insert(path);
@@ -214,18 +224,34 @@ static void addPathAndFileToVirtualColumns(Block & block, const String & path, s
         block.getByName("_file").column->assumeMutableRef().insert(file);
     }
 
+    if (use_hive_partitioning)
+    {
+        auto keys_and_values = parseHivePartitioningKeysAndValues(path);
+        for (const auto & [key, value] : keys_and_values)
+        {
+            if (const auto * column = block.findByName(key))
+            {
+                ReadBufferFromString buf(value);
+                column->type->getDefaultSerialization()->deserializeWholeText(column->column->assumeMutableRef(), buf, format_settings);
+            }
+        }
+    }
+
     block.getByName("_idx").column->assumeMutableRef().insert(idx);
 }
 
-std::optional<ActionsDAG> createPathAndFileFilterDAG(const ActionsDAG::Node * predicate, const NamesAndTypesList & virtual_columns)
+std::optional<ActionsDAG> createPathAndFileFilterDAG(const ActionsDAG::Node * predicate, const NamesAndTypesList & virtual_columns, const ContextPtr & context)
 {
     if (!predicate || virtual_columns.empty())
         return {};
 
     Block block;
+    NameSet common_virtuals;
+    if (context->getSettingsRef().use_hive_partitioning)
+        common_virtuals = getVirtualNamesForFileLikeStorage();
     for (const auto & column : virtual_columns)
     {
-        if (column.name == "_file" || column.name == "_path")
+        if (column.name == "_file" || column.name == "_path" || !common_virtuals.contains(column.name))
             block.insert({column.type->createColumn(), column.type, column.name});
     }
 
@@ -233,18 +259,19 @@ std::optional<ActionsDAG> createPathAndFileFilterDAG(const ActionsDAG::Node * pr
     return splitFilterDagForAllowedInputs(predicate, &block);
 }
 
-ColumnPtr getFilterByPathAndFileIndexes(const std::vector<String> & paths, const ExpressionActionsPtr & actions, const NamesAndTypesList & virtual_columns)
+ColumnPtr getFilterByPathAndFileIndexes(const std::vector<String> & paths, const ExpressionActionsPtr & actions, const NamesAndTypesList & virtual_columns, const ContextPtr & context)
 {
     Block block;
+    NameSet common_virtuals = getVirtualNamesForFileLikeStorage();
     for (const auto & column : virtual_columns)
     {
-        if (column.name == "_file" || column.name == "_path")
+        if (column.name == "_file" || column.name == "_path" || !common_virtuals.contains(column.name))
             block.insert({column.type->createColumn(), column.type, column.name});
     }
     block.insert({ColumnUInt64::create(), std::make_shared<DataTypeUInt64>(), "_idx"});
 
     for (size_t i = 0; i != paths.size(); ++i)
-        addPathAndFileToVirtualColumns(block, paths[i], i);
+        addPathAndFileToVirtualColumns(block, paths[i], i, getFormatSettings(context), context->getSettingsRef().use_hive_partitioning);
 
     filterBlockWithExpression(actions, block);
 

src/Storages/VirtualColumnUtils.h

@@ -75,14 +75,14 @@ VirtualColumnsDescription getVirtualsForFileLikeStorage(
     const std::string & sample_path = "",
     std::optional<FormatSettings> format_settings_ = std::nullopt);
 
-std::optional<ActionsDAG> createPathAndFileFilterDAG(const ActionsDAG::Node * predicate, const NamesAndTypesList & virtual_columns);
+std::optional<ActionsDAG> createPathAndFileFilterDAG(const ActionsDAG::Node * predicate, const NamesAndTypesList & virtual_columns, const ContextPtr & context);
 
-ColumnPtr getFilterByPathAndFileIndexes(const std::vector<String> & paths, const ExpressionActionsPtr & actions, const NamesAndTypesList & virtual_columns);
+ColumnPtr getFilterByPathAndFileIndexes(const std::vector<String> & paths, const ExpressionActionsPtr & actions, const NamesAndTypesList & virtual_columns, const ContextPtr & context);
 
 template <typename T>
-void filterByPathOrFile(std::vector<T> & sources, const std::vector<String> & paths, const ExpressionActionsPtr & actions, const NamesAndTypesList & virtual_columns)
+void filterByPathOrFile(std::vector<T> & sources, const std::vector<String> & paths, const ExpressionActionsPtr & actions, const NamesAndTypesList & virtual_columns, const ContextPtr & context)
 {
-    auto indexes_column = getFilterByPathAndFileIndexes(paths, actions, virtual_columns);
+    auto indexes_column = getFilterByPathAndFileIndexes(paths, actions, virtual_columns, context);
     const auto & indexes = typeid_cast<const ColumnUInt64 &>(*indexes_column).getData();
     if (indexes.size() == sources.size())
         return;

tests/integration/test_mask_sensitive_info/test.py

@@ -1,5 +1,6 @@
 import pytest
 import random, string
+import re
 from helpers.cluster import ClickHouseCluster
 from helpers.test_tools import TSV
 
@@ -336,6 +337,10 @@ def test_create_database():
 def test_table_functions():
     password = new_password()
     azure_conn_string = cluster.env_variables["AZURITE_CONNECTION_STRING"]
+    account_key_pattern = re.compile("AccountKey=.*?(;|$)")
+    masked_azure_conn_string = re.sub(
+        account_key_pattern, "AccountKey=[HIDDEN]\\1", azure_conn_string
+    )
     azure_storage_account_url = cluster.env_variables["AZURITE_STORAGE_ACCOUNT_URL"]
     azure_account_name = "devstoreaccount1"
     azure_account_key = "Eby8vdM02xNOcqFlqUwJPLlmEtlCDXJ1OUzFT50uSRZ6IFsuFq2UVErCz4I6tq/K1SZFPTOtr/KBHBeksoGMGw=="
@@ -467,23 +472,23 @@ def test_table_functions():
             "CREATE TABLE tablefunc30 (x int) AS s3('http://minio1:9001/root/data/test9.csv.gz', 'NOSIGN', 'CSV')",
             "CREATE TABLE tablefunc31 (`x` int) AS s3('http://minio1:9001/root/data/test10.csv.gz', 'minio', '[HIDDEN]')",
             "CREATE TABLE tablefunc32 (`x` int) AS deltaLake('http://minio1:9001/root/data/test11.csv.gz', 'minio', '[HIDDEN]')",
-            f"CREATE TABLE tablefunc33 (x int) AS azureBlobStorage('{azure_conn_string}', 'cont', 'test_simple.csv', 'CSV')",
-            f"CREATE TABLE tablefunc34 (x int) AS azureBlobStorage('{azure_conn_string}', 'cont', 'test_simple_1.csv', 'CSV', 'none')",
-            f"CREATE TABLE tablefunc35 (x int) AS azureBlobStorage('{azure_conn_string}', 'cont', 'test_simple_2.csv', 'CSV', 'none', 'auto')",
+            f"CREATE TABLE tablefunc33 (`x` int) AS azureBlobStorage('{masked_azure_conn_string}', 'cont', 'test_simple.csv', 'CSV')",
+            f"CREATE TABLE tablefunc34 (`x` int) AS azureBlobStorage('{masked_azure_conn_string}', 'cont', 'test_simple_1.csv', 'CSV', 'none')",
+            f"CREATE TABLE tablefunc35 (`x` int) AS azureBlobStorage('{masked_azure_conn_string}', 'cont', 'test_simple_2.csv', 'CSV', 'none', 'auto')",
             f"CREATE TABLE tablefunc36 (`x` int) AS azureBlobStorage('{azure_storage_account_url}', 'cont', 'test_simple_3.csv', '{azure_account_name}', '[HIDDEN]')",
             f"CREATE TABLE tablefunc37 (`x` int) AS azureBlobStorage('{azure_storage_account_url}', 'cont', 'test_simple_4.csv', '{azure_account_name}', '[HIDDEN]', 'CSV')",
             f"CREATE TABLE tablefunc38 (`x` int) AS azureBlobStorage('{azure_storage_account_url}', 'cont', 'test_simple_5.csv', '{azure_account_name}', '[HIDDEN]', 'CSV', 'none')",
             f"CREATE TABLE tablefunc39 (`x` int) AS azureBlobStorage('{azure_storage_account_url}', 'cont', 'test_simple_6.csv', '{azure_account_name}', '[HIDDEN]', 'CSV', 'none', 'auto')",
-            f"CREATE TABLE tablefunc40 (x int) AS azureBlobStorage(named_collection_2, connection_string = '{azure_conn_string}', container = 'cont', blob_path = 'test_simple_7.csv', format = 'CSV')",
+            f"CREATE TABLE tablefunc40 (`x` int) AS azureBlobStorage(named_collection_2, connection_string = '{masked_azure_conn_string}', container = 'cont', blob_path = 'test_simple_7.csv', format = 'CSV')",
             f"CREATE TABLE tablefunc41 (`x` int) AS azureBlobStorage(named_collection_2, storage_account_url = '{azure_storage_account_url}', container = 'cont', blob_path = 'test_simple_8.csv', account_name = '{azure_account_name}', account_key = '[HIDDEN]')",
-            f"CREATE TABLE tablefunc42 (x int) AS azureBlobStorageCluster('test_shard_localhost', '{azure_conn_string}', 'cont', 'test_simple_9.csv', 'CSV')",
-            f"CREATE TABLE tablefunc43 (x int) AS azureBlobStorageCluster('test_shard_localhost', '{azure_conn_string}', 'cont', 'test_simple_10.csv', 'CSV', 'none')",
-            f"CREATE TABLE tablefunc44 (x int) AS azureBlobStorageCluster('test_shard_localhost', '{azure_conn_string}', 'cont', 'test_simple_11.csv', 'CSV', 'none', 'auto')",
+            f"CREATE TABLE tablefunc42 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', '{masked_azure_conn_string}', 'cont', 'test_simple_9.csv', 'CSV')",
+            f"CREATE TABLE tablefunc43 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', '{masked_azure_conn_string}', 'cont', 'test_simple_10.csv', 'CSV', 'none')",
+            f"CREATE TABLE tablefunc44 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', '{masked_azure_conn_string}', 'cont', 'test_simple_11.csv', 'CSV', 'none', 'auto')",
             f"CREATE TABLE tablefunc45 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', '{azure_storage_account_url}', 'cont', 'test_simple_12.csv', '{azure_account_name}', '[HIDDEN]')",
             f"CREATE TABLE tablefunc46 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', '{azure_storage_account_url}', 'cont', 'test_simple_13.csv', '{azure_account_name}', '[HIDDEN]', 'CSV')",
             f"CREATE TABLE tablefunc47 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', '{azure_storage_account_url}', 'cont', 'test_simple_14.csv', '{azure_account_name}', '[HIDDEN]', 'CSV', 'none')",
             f"CREATE TABLE tablefunc48 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', '{azure_storage_account_url}', 'cont', 'test_simple_15.csv', '{azure_account_name}', '[HIDDEN]', 'CSV', 'none', 'auto')",
-            f"CREATE TABLE tablefunc49 (x int) AS azureBlobStorageCluster('test_shard_localhost', named_collection_2, connection_string = '{azure_conn_string}', container = 'cont', blob_path = 'test_simple_16.csv', format = 'CSV')",
+            f"CREATE TABLE tablefunc49 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', named_collection_2, connection_string = '{masked_azure_conn_string}', container = 'cont', blob_path = 'test_simple_16.csv', format = 'CSV')",
             f"CREATE TABLE tablefunc50 (`x` int) AS azureBlobStorageCluster('test_shard_localhost', named_collection_2, storage_account_url = '{azure_storage_account_url}', container = 'cont', blob_path = 'test_simple_17.csv', account_name = '{azure_account_name}', account_key = '[HIDDEN]')",
             "CREATE TABLE tablefunc51 (`x` int) AS iceberg('http://minio1:9001/root/data/test11.csv.gz', 'minio', '[HIDDEN]')",
         ],

tests/queries/0_stateless/03203_function_printf.reference

@@ -1,21 +1,31 @@
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
+%d: 123
+%d: -123
+%d: 0
+%d: 9223372036854775807
+%i: 123
+%u: 123
+%o: 173
+%x: 7b
+%X: 7B
+%f: 0.000000
+%f: 123.456000
+%f: -123.456000
+%F: 123.456000
+%e: 1.234560e+02
+%E: 1.234560E+02
+%g: 123.456
+%G: 123.456
+%a: 0x1.edd2f1a9fbe77p+6
+%A: 0X1.EDD2F1A9FBE77P+6
+%s: abc
+   ┌─printf('%%s: %s', '\n\t')─┐
+1. │ %s: 
+	                    │
+   └───────────────────────────┘
+%s: 
+%%: %
+%.5d: 00123
+%.2f: 123.46
+%.2e: 1.23e+02
+%.2g: 1.2e+02
+%.2s: ab

tests/queries/0_stateless/03203_function_printf.sql

@@ -1,39 +1,47 @@
 -- Testing integer formats
-select printf('%%d: %d', 123) = '%d: 123';
-select printf('%%i: %i', 123) = '%i: 123';
-select printf('%%u: %u', 123) = '%u: 123';
-select printf('%%o: %o', 123) = '%o: 173';
-select printf('%%x: %x', 123) = '%x: 7b';
-select printf('%%X: %X', 123) = '%X: 7B';
+select printf('%%d: %d', 123);
+select printf('%%d: %d', -123);
+select printf('%%d: %d', 0);
+select printf('%%d: %d', 9223372036854775807);
+select printf('%%i: %i', 123);
+select printf('%%u: %u', 123);
+select printf('%%o: %o', 123);
+select printf('%%x: %x', 123);
+select printf('%%X: %X', 123);
 
 -- Testing floating point formats
-select printf('%%f: %f', 123.456) = '%f: 123.456000';
-select printf('%%F: %F', 123.456) = '%F: 123.456000';
-select printf('%%e: %e', 123.456) = '%e: 1.234560e+02';
-select printf('%%E: %E', 123.456) = '%E: 1.234560E+02';
-select printf('%%g: %g', 123.456) = '%g: 123.456';
-select printf('%%G: %G', 123.456) = '%G: 123.456';
-select printf('%%a: %a', 123.456) = '%a: 0x1.edd2f1a9fbe77p+6';
-select printf('%%A: %A', 123.456) = '%A: 0X1.EDD2F1A9FBE77P+6';
+select printf('%%f: %f', 0.0);
+select printf('%%f: %f', 123.456);
+select printf('%%f: %f', -123.456);
+select printf('%%F: %F', 123.456);
+select printf('%%e: %e', 123.456);
+select printf('%%E: %E', 123.456);
+select printf('%%g: %g', 123.456);
+select printf('%%G: %G', 123.456);
+select printf('%%a: %a', 123.456);
+select printf('%%A: %A', 123.456);
 
 -- Testing character formats
-select printf('%%s: %s', 'abc') = '%s: abc';
+select printf('%%s: %s', 'abc');
+SELECT printf('%%s: %s', '\n\t') FORMAT PrettyCompact;
+select printf('%%s: %s', '');
 
 -- Testing the %% specifier
-select printf('%%%%: %%') = '%%: %';
+select printf('%%%%: %%');
 
 -- Testing integer formats with precision
-select printf('%%.5d: %.5d', 123) = '%.5d: 00123';
+select printf('%%.5d: %.5d', 123);
 
 -- Testing floating point formats with precision
-select printf('%%.2f: %.2f', 123.456) = '%.2f: 123.46';
-select printf('%%.2e: %.2e', 123.456) = '%.2e: 1.23e+02';
-select printf('%%.2g: %.2g', 123.456) = '%.2g: 1.2e+02';
+select printf('%%.2f: %.2f', 123.456);
+select printf('%%.2e: %.2e', 123.456);
+select printf('%%.2g: %.2g', 123.456);
 
 -- Testing character formats with precision
-select printf('%%.2s: %.2s', 'abc') = '%.2s: ab';
+select printf('%%.2s: %.2s', 'abc');
 
 select printf('%%X: %X', 123.123); -- { serverError BAD_ARGUMENTS }
 select printf('%%A: %A', 'abc'); -- { serverError BAD_ARGUMENTS }
 select printf('%%s: %s', 100); -- { serverError BAD_ARGUMENTS }
 select printf('%%n: %n', 100); -- { serverError BAD_ARGUMENTS }
+select printf('%%f: %f', 0); -- { serverError BAD_ARGUMENTS }

tests/queries/0_stateless/03203_hive_style_partitioning.reference

@@ -33,8 +33,8 @@ Cross	Elizabeth
 [1,2,3]	42.42
 Array(Int64)	LowCardinality(Float64)
 101
-2070
-2070
+2071
+2071
 b
 1
 1

tests/queries/0_stateless/03231_hive_partitioning_filtering.reference

@@ -0,0 +1,6 @@
+1
+1
+1
+1
+1
+1

tests/queries/0_stateless/03231_hive_partitioning_filtering.sh

@@ -0,0 +1,72 @@
+#!/usr/bin/env bash
+# Tags: no-fasttest
+
+CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
+# shellcheck source=../shell_config.sh
+. "$CURDIR"/../shell_config.sh
+
+DATA_DIR=$USER_FILES_PATH/$CLICKHOUSE_TEST_UNIQUE_NAME
+mkdir -p $DATA_DIR
+cp -r $CURDIR/data_hive/ $DATA_DIR
+
+$CLICKHOUSE_CLIENT --query_id="test_03231_1_$CLICKHOUSE_TEST_UNIQUE_NAME" --query "
+    SELECT countDistinct(_path) FROM file('$DATA_DIR/data_hive/partitioning/column0=*/sample.parquet') WHERE column0 = 'Elizabeth' SETTINGS use_hive_partitioning=1, optimize_count_from_files=0;
+"
+
+$CLICKHOUSE_CLIENT --query "
+    SYSTEM FLUSH LOGS;
+"
+
+for _ in {1..5}; do
+    count=$( $CLICKHOUSE_CLIENT --query "
+        SELECT ProfileEvents['EngineFileLikeReadFiles'] FROM system.query_log 
+        WHERE query_id='test_03231_1_$CLICKHOUSE_TEST_UNIQUE_NAME' AND 
+        current_database = currentDatabase() and type='QueryFinish';" )
+    if [[ "$count" == "1" ]]; then
+        echo "1"
+        break
+    fi
+    sleep 1
+done
+
+$CLICKHOUSE_CLIENT --query_id="test_03231_2_$CLICKHOUSE_TEST_UNIQUE_NAME" --query "
+    SELECT countDistinct(_path) FROM file('$DATA_DIR/data_hive/partitioning/identifier=*/email.csv') WHERE identifier = 2070 SETTINGS use_hive_partitioning=1, optimize_count_from_files=0;
+"
+
+$CLICKHOUSE_CLIENT --query "
+    SYSTEM FLUSH LOGS;
+"
+
+for _ in {1..5}; do
+    count=$( $CLICKHOUSE_CLIENT --query "
+        SELECT ProfileEvents['EngineFileLikeReadFiles'] FROM system.query_log 
+        WHERE query_id='test_03231_2_$CLICKHOUSE_TEST_UNIQUE_NAME' AND 
+        current_database = currentDatabase() and type='QueryFinish';" )
+    if [[ "$count" == "1" ]]; then
+        echo "1"
+        break
+    fi
+    sleep 1
+done
+
+$CLICKHOUSE_CLIENT --query_id="test_03231_3_$CLICKHOUSE_TEST_UNIQUE_NAME" --query "
+    SELECT countDistinct(_path) FROM file('$DATA_DIR/data_hive/partitioning/array=*/sample.parquet') WHERE array = [1,2,3] SETTINGS use_hive_partitioning=1, optimize_count_from_files=0;
+"
+
+$CLICKHOUSE_CLIENT --query "
+    SYSTEM FLUSH LOGS;
+"
+
+for _ in {1..5}; do
+    count=$( $CLICKHOUSE_CLIENT --query "
+        SELECT ProfileEvents['EngineFileLikeReadFiles'] FROM system.query_log 
+        WHERE query_id='test_03231_3_$CLICKHOUSE_TEST_UNIQUE_NAME' AND 
+        current_database = currentDatabase() and type='QueryFinish';" )
+    if [[ "$count" == "1" ]]; then
+        echo "1"
+        break
+    fi
+    sleep 1
+done
+
+rm -rf $DATA_DIR

tests/queries/0_stateless/03232_file_path_normalizing.reference

@@ -1 +1 @@
-data_hive/partitioning/column0=Elizabeth/sample.parquet
+data_hive/partitioning/non_existing_column=Elizabeth/sample.parquet

tests/queries/0_stateless/03232_file_path_normalizing.sh

@@ -5,4 +5,4 @@ CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
 # shellcheck source=../shell_config.sh
 . "$CURDIR"/../shell_config.sh
 
-$CLICKHOUSE_LOCAL -q "SELECT substring(_path, position(_path, 'data_hive')) FROM file('$CURDIR/data_hive/partitioning/column0=*/sample.parquet') LIMIT 1;"
+$CLICKHOUSE_LOCAL -q "SELECT substring(_path, position(_path, 'data_hive')) FROM file('$CURDIR/data_hive/partitioning/non_existing_column=*/sample.parquet') LIMIT 1;"

tests/queries/0_stateless/data_hive/partitioning/identifier=2071/email.csv

@@ -0,0 +1,5 @@
+_login_email,_identifier,_first_name,_last_name
+laura@example.com,2070,Laura,Grey
+craig@example.com,4081,Craig,Johnson
+mary@example.com,9346,Mary,Jenkins
+jamie@example.com,5079,Jamie,Smith