Merge pull request #40414 from Avogar/improve-schema-inference-cache

Improve schema inference cache, respect format settings that can change the schema

src/DataTypes/Serializations/SerializationArray.cpp

@@ -535,7 +535,7 @@ void SerializationArray::deserializeTextCSV(IColumn & column, ReadBuffer & istr,
     readCSV(s, istr, settings.csv);
     ReadBufferFromString rb(s);
 
-    if (settings.csv.input_format_arrays_as_nested_csv)
+    if (settings.csv.arrays_as_nested_csv)
     {
         deserializeTextImpl(column, rb,
             [&](IColumn & nested_column)

src/DataTypes/Serializations/SerializationEnum.cpp

@@ -24,7 +24,7 @@ void SerializationEnum<Type>::serializeTextEscaped(const IColumn & column, size_
 template <typename Type>
 void SerializationEnum<Type>::deserializeTextEscaped(IColumn & column, ReadBuffer & istr, const FormatSettings & settings) const
 {
-    if (settings.tsv.input_format_enum_as_number)
+    if (settings.tsv.enum_as_number)
         assert_cast<ColumnType &>(column).getData().push_back(readValue(istr));
     else
     {
@@ -52,7 +52,7 @@ void SerializationEnum<Type>::deserializeTextQuoted(IColumn & column, ReadBuffer
 template <typename Type>
 void SerializationEnum<Type>::deserializeWholeText(IColumn & column, ReadBuffer & istr, const FormatSettings & settings) const
 {
-    if (settings.tsv.input_format_enum_as_number)
+    if (settings.tsv.enum_as_number)
     {
         assert_cast<ColumnType &>(column).getData().push_back(readValue(istr));
         if (!istr.eof())
@@ -100,7 +100,7 @@ void SerializationEnum<Type>::serializeTextCSV(const IColumn & column, size_t ro
 template <typename Type>
 void SerializationEnum<Type>::deserializeTextCSV(IColumn & column, ReadBuffer & istr, const FormatSettings & settings) const
 {
-    if (settings.csv.input_format_enum_as_number)
+    if (settings.csv.enum_as_number)
         assert_cast<ColumnType &>(column).getData().push_back(readValue(istr));
     else
     {

src/Formats/EscapingRuleUtils.cpp

@@ -697,7 +697,7 @@ DataTypePtr determineDataTypeByEscapingRule(const String & field, const FormatSe
             return JSONUtils::getDataTypeFromField(field, format_settings);
         case FormatSettings::EscapingRule::CSV:
         {
-            if (!format_settings.csv.input_format_use_best_effort_in_schema_inference)
+            if (!format_settings.csv.use_best_effort_in_schema_inference)
                 return makeNullable(std::make_shared<DataTypeString>());
 
             if (field.empty() || field == format_settings.csv.null_representation)
@@ -745,7 +745,7 @@ DataTypePtr determineDataTypeByEscapingRule(const String & field, const FormatSe
         case FormatSettings::EscapingRule::Raw: [[fallthrough]];
         case FormatSettings::EscapingRule::Escaped:
         {
-            if (!format_settings.tsv.input_format_use_best_effort_in_schema_inference)
+            if (!format_settings.tsv.use_best_effort_in_schema_inference)
                 return makeNullable(std::make_shared<DataTypeString>());
 
             if (field.empty() || field == format_settings.tsv.null_representation)
@@ -799,4 +799,49 @@ DataTypes getDefaultDataTypeForEscapingRules(const std::vector<FormatSettings::E
     return data_types;
 }
 
+String getAdditionalFormatInfoByEscapingRule(const FormatSettings & settings, FormatSettings::EscapingRule escaping_rule)
+{
+    String result;
+    /// First, settings that are common for all text formats:
+    result = fmt::format(
+        "schema_inference_hints={}, try_infer_integers={}, try_infer_dates={}, try_infer_datetimes={}, max_rows_to_read_for_schema_inference={}",
+        settings.schema_inference_hints,
+        settings.try_infer_integers,
+        settings.try_infer_dates,
+        settings.try_infer_datetimes,
+        settings.max_rows_to_read_for_schema_inference);
+
+    /// Second, format-specific settings:
+    switch (escaping_rule)
+    {
+        case FormatSettings::EscapingRule::Escaped:
+        case FormatSettings::EscapingRule::Raw:
+            result += fmt::format(
+                ", use_best_effort_in_schema_inference={}, bool_true_representation={}, bool_false_representation={}, null_representation={}",
+                settings.tsv.use_best_effort_in_schema_inference,
+                settings.bool_true_representation,
+                settings.bool_false_representation,
+                settings.tsv.null_representation);
+            break;
+        case FormatSettings::EscapingRule::CSV:
+            result += fmt::format(
+                ", use_best_effort_in_schema_inference={}, bool_true_representation={}, bool_false_representation={},"
+                " null_representation={}, delimiter={}, tuple_delimiter={}",
+                settings.tsv.use_best_effort_in_schema_inference,
+                settings.bool_true_representation,
+                settings.bool_false_representation,
+                settings.csv.null_representation,
+                settings.csv.delimiter,
+                settings.csv.tuple_delimiter);
+            break;
+        case FormatSettings::EscapingRule::JSON:
+            result += fmt::format(", try_infer_numbers_from_strings={}, read_bools_as_numbers={}", settings.json.try_infer_numbers_from_strings, settings.json.read_bools_as_numbers);
+            break;
+        default:
+            break;
+    }
+
+    return result;
+}
+
 }

src/Formats/EscapingRuleUtils.h

@@ -77,4 +77,6 @@ void transformInferredTypesIfNeeded(DataTypePtr & first, DataTypePtr & second, c
 void transformInferredJSONTypesIfNeeded(DataTypes & types, const FormatSettings & settings, const std::unordered_set<const IDataType *> * numbers_parsed_from_json_strings = nullptr);
 void transformInferredJSONTypesIfNeeded(DataTypePtr & first, DataTypePtr & second, const FormatSettings & settings);
 
+String getAdditionalFormatInfoByEscapingRule(const FormatSettings & settings,FormatSettings::EscapingRule escaping_rule);
+
 }

src/Formats/FormatFactory.cpp

@@ -63,10 +63,10 @@ FormatSettings getFormatSettings(ContextPtr context, const Settings & settings)
     format_settings.csv.delimiter = settings.format_csv_delimiter;
     format_settings.csv.tuple_delimiter = settings.format_csv_delimiter;
     format_settings.csv.empty_as_default = settings.input_format_csv_empty_as_default;
-    format_settings.csv.input_format_enum_as_number = settings.input_format_csv_enum_as_number;
+    format_settings.csv.enum_as_number = settings.input_format_csv_enum_as_number;
     format_settings.csv.null_representation = settings.format_csv_null_representation;
-    format_settings.csv.input_format_arrays_as_nested_csv = settings.input_format_csv_arrays_as_nested_csv;
+    format_settings.csv.arrays_as_nested_csv = settings.input_format_csv_arrays_as_nested_csv;
-    format_settings.csv.input_format_use_best_effort_in_schema_inference = settings.input_format_csv_use_best_effort_in_schema_inference;
+    format_settings.csv.use_best_effort_in_schema_inference = settings.input_format_csv_use_best_effort_in_schema_inference;
     format_settings.csv.skip_first_lines = settings.input_format_csv_skip_first_lines;
     format_settings.hive_text.fields_delimiter = settings.input_format_hive_text_fields_delimiter;
     format_settings.hive_text.collection_items_delimiter = settings.input_format_hive_text_collection_items_delimiter;
@@ -124,9 +124,9 @@ FormatSettings getFormatSettings(ContextPtr context, const Settings & settings)
     format_settings.template_settings.row_format = settings.format_template_row;
     format_settings.tsv.crlf_end_of_line = settings.output_format_tsv_crlf_end_of_line;
     format_settings.tsv.empty_as_default = settings.input_format_tsv_empty_as_default;
-    format_settings.tsv.input_format_enum_as_number = settings.input_format_tsv_enum_as_number;
+    format_settings.tsv.enum_as_number = settings.input_format_tsv_enum_as_number;
     format_settings.tsv.null_representation = settings.format_tsv_null_representation;
-    format_settings.tsv.input_format_use_best_effort_in_schema_inference = settings.input_format_tsv_use_best_effort_in_schema_inference;
+    format_settings.tsv.use_best_effort_in_schema_inference = settings.input_format_tsv_use_best_effort_in_schema_inference;
     format_settings.tsv.skip_first_lines = settings.input_format_tsv_skip_first_lines;
     format_settings.values.accurate_types_of_literals = settings.input_format_values_accurate_types_of_literals;
     format_settings.values.deduce_templates_of_expressions = settings.input_format_values_deduce_templates_of_expressions;

src/Formats/FormatSettings.h

@@ -111,11 +111,11 @@ struct FormatSettings
         bool allow_double_quotes = true;
         bool empty_as_default = false;
         bool crlf_end_of_line = false;
-        bool input_format_enum_as_number = false;
+        bool enum_as_number = false;
-        bool input_format_arrays_as_nested_csv = false;
+        bool arrays_as_nested_csv = false;
         String null_representation = "\\N";
         char tuple_delimiter = ',';
-        bool input_format_use_best_effort_in_schema_inference = true;
+        bool use_best_effort_in_schema_inference = true;
         UInt64 skip_first_lines = 0;
     } csv;
 
@@ -227,8 +227,8 @@ struct FormatSettings
         bool empty_as_default = false;
         bool crlf_end_of_line = false;
         String null_representation = "\\N";
-        bool input_format_enum_as_number = false;
+        bool enum_as_number = false;
-        bool input_format_use_best_effort_in_schema_inference = true;
+        bool use_best_effort_in_schema_inference = true;
         UInt64 skip_first_lines = 0;
     } tsv;
 

src/Formats/ReadSchemaUtils.cpp

@@ -239,26 +239,17 @@ NamesAndTypesList getNamesAndRecursivelyNullableTypes(const Block & header)
     return result;
 }
 
-String getKeyForSchemaCache(const String & source, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context)
+SchemaCache::Key getKeyForSchemaCache(const String & source, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context)
 {
     return getKeysForSchemaCache({source}, format, format_settings, context).front();
 }
 
-static String makeSchemaCacheKey(const String & source, const String & format, const String & additional_format_info)
+static SchemaCache::Key makeSchemaCacheKey(const String & source, const String & format, const String & additional_format_info)
 {
-    return source + "@@" + format + "@@" + additional_format_info;
+    return SchemaCache::Key{source, format, additional_format_info};
 }
 
-void splitSchemaCacheKey(const String & key, String & source, String & format, String & additional_format_info)
+SchemaCache::Keys getKeysForSchemaCache(const Strings & sources, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context)
-{
-    size_t additional_format_info_pos = key.rfind("@@");
-    additional_format_info = key.substr(additional_format_info_pos + 2, key.size() - additional_format_info_pos - 2);
-    size_t format_pos = key.rfind("@@", additional_format_info_pos - 1);
-    format = key.substr(format_pos + 2, additional_format_info_pos - format_pos - 2);
-    source = key.substr(0, format_pos);
-}
-
-Strings getKeysForSchemaCache(const Strings & sources, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context)
 {
     /// For some formats data schema depends on some settings, so it's possible that
     /// two queries to the same source will get two different schemas. To process this
@@ -266,7 +257,7 @@ Strings getKeysForSchemaCache(const Strings & sources, const String & format, co
     /// For example, for Protobuf format additional information is the path to the schema
     /// and message name.
     String additional_format_info = FormatFactory::instance().getAdditionalInfoForSchemaCache(format, context, format_settings);
-    Strings cache_keys;
+    SchemaCache::Keys cache_keys;
     cache_keys.reserve(sources.size());
     std::transform(sources.begin(), sources.end(), std::back_inserter(cache_keys), [&](const auto & source){ return makeSchemaCacheKey(source, format, additional_format_info); });
     return cache_keys;

src/Formats/ReadSchemaUtils.h

@@ -1,6 +1,7 @@
 #pragma once
 
 #include <Storages/ColumnsDescription.h>
+#include <Storages/Cache/SchemaCache.h>
 #include <Formats/FormatFactory.h>
 
 namespace DB
@@ -47,8 +48,8 @@ DataTypePtr makeNullableRecursivelyAndCheckForNothing(DataTypePtr type);
 /// in the block and return names and types.
 NamesAndTypesList getNamesAndRecursivelyNullableTypes(const Block & header);
 
-String getKeyForSchemaCache(const String & source, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context);
+SchemaCache::Key  getKeyForSchemaCache(const String & source, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context);
-Strings getKeysForSchemaCache(const Strings & sources, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context);
+SchemaCache::Keys  getKeysForSchemaCache(const Strings & sources, const String & format, const std::optional<FormatSettings> & format_settings, const ContextPtr & context);
 
 void splitSchemaCacheKey(const String & key, String & source, String & format, String & additional_format_info);
 }

src/Processors/Formats/ISchemaReader.cpp

@@ -89,15 +89,13 @@ void IIRowSchemaReader::setContext(ContextPtr & context)
 }
 
 IRowSchemaReader::IRowSchemaReader(ReadBuffer & in_, const FormatSettings & format_settings_)
-    : IIRowSchemaReader(in_, format_settings_)
+    : IIRowSchemaReader(in_, format_settings_), column_names(splitColumnNames(format_settings.column_names_for_schema_inference))
 {
-    initColumnNames(format_settings.column_names_for_schema_inference);
 }
 
 IRowSchemaReader::IRowSchemaReader(ReadBuffer & in_, const FormatSettings & format_settings_, DataTypePtr default_type_)
-    : IIRowSchemaReader(in_, format_settings_, default_type_)
+    : IIRowSchemaReader(in_, format_settings_, default_type_), column_names(splitColumnNames(format_settings.column_names_for_schema_inference))
 {
-    initColumnNames(format_settings.column_names_for_schema_inference);
 }
 
 IRowSchemaReader::IRowSchemaReader(ReadBuffer & in_, const FormatSettings & format_settings_, const DataTypes & default_types_)
@@ -171,11 +169,12 @@ NamesAndTypesList IRowSchemaReader::readSchema()
     return result;
 }
 
-void IRowSchemaReader::initColumnNames(const String & column_names_str)
+Strings splitColumnNames(const String & column_names_str)
 {
     if (column_names_str.empty())
-        return;
+        return {};
 
+    Strings column_names;
     /// column_names_for_schema_inference is a string in format 'column1,column2,column3,...'
     boost::split(column_names, column_names_str, boost::is_any_of(","));
     for (auto & column_name : column_names)
@@ -184,6 +183,7 @@ void IRowSchemaReader::initColumnNames(const String & column_names_str)
         if (!col_name_trimmed.empty())
             column_name = col_name_trimmed;
     }
+    return column_names;
 }
 
 DataTypePtr IRowSchemaReader::getDefaultType(size_t column) const

src/Processors/Formats/ISchemaReader.h

@@ -136,4 +136,6 @@ void chooseResultColumnType(
 void checkResultColumnTypeAndAppend(
     NamesAndTypesList & result, DataTypePtr & type, const String & name, const DataTypePtr & default_type, size_t rows_read);
 
+Strings splitColumnNames(const String & column_names_str);
+
 }

src/Processors/Formats/Impl/CSVRowInputFormat.cpp

@@ -401,6 +401,13 @@ void registerCSVSchemaReader(FormatFactory & factory)
         {
             return std::make_shared<CSVSchemaReader>(buf, with_names, with_types, settings);
         });
+        factory.registerAdditionalInfoForSchemaCacheGetter(format_name, [with_names](const FormatSettings & settings)
+        {
+            String result = getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::CSV);
+            if (!with_names)
+                result += fmt::format(", column_names_for_schema_inference={}", settings.column_names_for_schema_inference);
+            return result;
+        });
     };
 
     registerWithNamesAndTypes("CSV", register_func);

src/Processors/Formats/Impl/CapnProtoRowInputFormat.cpp

@@ -319,7 +319,7 @@ void registerInputFormatCapnProto(FormatFactory & factory)
     factory.markFormatSupportsSubsetOfColumns("CapnProto");
     factory.registerFileExtension("capnp", "CapnProto");
     factory.registerAdditionalInfoForSchemaCacheGetter(
-        "CapnProto", [](const FormatSettings & settings) { return "Format schema: " + settings.schema.format_schema; });
+        "CapnProto", [](const FormatSettings & settings) { return fmt::format("format_schema={}", settings.schema.format_schema); });
 }
 
 void registerCapnProtoSchemaReader(FormatFactory & factory)

src/Processors/Formats/Impl/CustomSeparatedRowInputFormat.cpp

@@ -353,6 +353,19 @@ void registerCustomSeparatedSchemaReader(FormatFactory & factory)
             {
                 return std::make_shared<CustomSeparatedSchemaReader>(buf, with_names, with_types, ignore_spaces, settings);
             });
+            factory.registerAdditionalInfoForSchemaCacheGetter(format_name, [](const FormatSettings & settings)
+            {
+                String result = getAdditionalFormatInfoByEscapingRule(settings, settings.custom.escaping_rule);
+                return result + fmt::format(
+                        ", result_before_delimiter={}, row_before_delimiter={}, field_delimiter={},"
+                        " row_after_delimiter={}, row_between_delimiter={}, result_after_delimiter={}",
+                        settings.custom.result_before_delimiter,
+                        settings.custom.row_before_delimiter,
+                        settings.custom.field_delimiter,
+                        settings.custom.row_after_delimiter,
+                        settings.custom.row_between_delimiter,
+                        settings.custom.result_after_delimiter);
+            });
         };
 
         registerWithNamesAndTypes(ignore_spaces ? "CustomSeparatedIgnoreSpaces" : "CustomSeparated", register_func);

src/Processors/Formats/Impl/JSONColumnsBlockInputFormat.cpp

@@ -1,6 +1,7 @@
 #include <Processors/Formats/Impl/JSONColumnsBlockInputFormat.h>
 #include <IO/ReadHelpers.h>
 #include <Formats/FormatFactory.h>
+#include <Formats/EscapingRuleUtils.h>
 
 namespace DB
 {
@@ -66,6 +67,10 @@ void registerJSONColumnsSchemaReader(FormatFactory & factory)
             return std::make_shared<JSONColumnsSchemaReaderBase>(buf, settings, std::make_unique<JSONColumnsReader>(buf));
         }
     );
+    factory.registerAdditionalInfoForSchemaCacheGetter("JSONColumns", [](const FormatSettings & settings)
+    {
+        return getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::JSON);
+    });
 }
 
 }

src/Processors/Formats/Impl/JSONColumnsBlockInputFormatBase.cpp

@@ -1,4 +1,5 @@
 #include <Processors/Formats/Impl/JSONColumnsBlockInputFormatBase.h>
+#include <Processors/Formats/ISchemaReader.h>
 #include <Formats/JSONUtils.h>
 #include <Formats/EscapingRuleUtils.h>
 #include <IO/ReadHelpers.h>
@@ -178,7 +179,10 @@ Chunk JSONColumnsBlockInputFormatBase::generate()
 
 JSONColumnsSchemaReaderBase::JSONColumnsSchemaReaderBase(
     ReadBuffer & in_, const FormatSettings & format_settings_, std::unique_ptr<JSONColumnsReaderBase> reader_)
-    : ISchemaReader(in_), format_settings(format_settings_), reader(std::move(reader_))
+    : ISchemaReader(in_)
+    , format_settings(format_settings_)
+    , reader(std::move(reader_))
+    , column_names_from_settings(splitColumnNames(format_settings_.column_names_for_schema_inference))
 {
 }
 
@@ -214,8 +218,15 @@ NamesAndTypesList JSONColumnsSchemaReaderBase::readSchema()
         do
         {
             auto column_name_opt = reader->readColumnStart();
-            /// If format doesn't have named for columns, use default names 'c1', 'c2', ...
+            /// If format doesn't have names for columns, use names from setting column_names_for_schema_inference or default names 'c1', 'c2', ...
-            String column_name = column_name_opt.has_value() ? *column_name_opt : "c" + std::to_string(iteration + 1);
+            String column_name;
+            if (column_name_opt.has_value())
+                column_name = *column_name_opt;
+            else if (iteration < column_names_from_settings.size())
+                column_name = column_names_from_settings[iteration];
+            else
+                column_name = "c" + std::to_string(iteration + 1);
+
             /// Keep order of column names as it is in input data.
             if (!names_to_types.contains(column_name))
                 names_order.push_back(column_name);

src/Processors/Formats/Impl/JSONColumnsBlockInputFormatBase.h

@@ -87,6 +87,7 @@ private:
 
     const FormatSettings format_settings;
     std::unique_ptr<JSONColumnsReaderBase> reader;
+    Names column_names_from_settings;
 };
 
 }

src/Processors/Formats/Impl/JSONCompactColumnsBlockInputFormat.cpp

@@ -1,6 +1,7 @@
 #include <Processors/Formats/Impl/JSONCompactColumnsBlockInputFormat.h>
 #include <IO/ReadHelpers.h>
 #include <Formats/FormatFactory.h>
+#include <Formats/EscapingRuleUtils.h>
 
 namespace DB
 {
@@ -60,6 +61,11 @@ void registerJSONCompactColumnsSchemaReader(FormatFactory & factory)
             return std::make_shared<JSONColumnsSchemaReaderBase>(buf, settings, std::make_unique<JSONCompactColumnsReader>(buf));
         }
     );
+    factory.registerAdditionalInfoForSchemaCacheGetter("JSONCompactColumns", [](const FormatSettings & settings)
+    {
+        auto result = getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::JSON);
+        return result + fmt::format(", column_names_for_schema_inference={}", settings.column_names_for_schema_inference);
+    });
 }
 
 }

src/Processors/Formats/Impl/JSONCompactEachRowRowInputFormat.cpp

@@ -245,6 +245,11 @@ void registerJSONCompactEachRowSchemaReader(FormatFactory & factory)
             {
                 return std::make_shared<JSONCompactEachRowRowSchemaReader>(buf, with_names, with_types, json_strings, settings);
             });
+            factory.registerAdditionalInfoForSchemaCacheGetter(format_name, [](const FormatSettings & settings)
+            {
+                auto result = getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::JSON);
+                return result + fmt::format(", column_names_for_schema_inference={}", settings.column_names_for_schema_inference);
+            });
         };
         registerWithNamesAndTypes(json_strings ? "JSONCompactStringsEachRow" : "JSONCompactEachRow", register_func);
     }

src/Processors/Formats/Impl/JSONEachRowRowInputFormat.cpp

@@ -355,44 +355,26 @@ void JSONEachRowSchemaReader::transformTypesIfNeeded(DataTypePtr & type, DataTyp
 
 void registerInputFormatJSONEachRow(FormatFactory & factory)
 {
-    factory.registerInputFormat("JSONEachRow", [](
+    auto register_format = [&](const String & format_name, bool json_strings)
-        ReadBuffer & buf,
-        const Block & sample,
-        IRowInputFormat::Params params,
-        const FormatSettings & settings)
     {
-        return std::make_shared<JSONEachRowRowInputFormat>(buf, sample, std::move(params), settings, false);
+        factory.registerInputFormat(format_name, [json_strings](
-    });
+            ReadBuffer & buf,
+            const Block & sample,
+            IRowInputFormat::Params params,
+            const FormatSettings & settings)
+        {
+            return std::make_shared<JSONEachRowRowInputFormat>(buf, sample, std::move(params), settings, json_strings);
+        });
+    };
 
-    factory.registerInputFormat("JSONLines", [](
+    register_format("JSONEachRow", false);
-        ReadBuffer & buf,
+    register_format("JSONLines", false);
-        const Block & sample,
+    register_format("NDJSON", false);
-        IRowInputFormat::Params params,
-        const FormatSettings & settings)
-    {
-        return std::make_shared<JSONEachRowRowInputFormat>(buf, sample, std::move(params), settings, false);
-    });
-
-    factory.registerInputFormat("NDJSON", [](
-        ReadBuffer & buf,
-        const Block & sample,
-        IRowInputFormat::Params params,
-        const FormatSettings & settings)
-    {
-        return std::make_shared<JSONEachRowRowInputFormat>(buf, sample, std::move(params), settings, false);
-    });
 
     factory.registerFileExtension("ndjson", "JSONEachRow");
     factory.registerFileExtension("jsonl", "JSONEachRow");
 
-    factory.registerInputFormat("JSONStringsEachRow", [](
+    register_format("JSONStringsEachRow", true);
-        ReadBuffer & buf,
-        const Block & sample,
-        IRowInputFormat::Params params,
-        const FormatSettings & settings)
-    {
-        return std::make_shared<JSONEachRowRowInputFormat>(buf, sample, std::move(params), settings, true);
-    });
 
     factory.markFormatSupportsSubsetOfColumns("JSONEachRow");
     factory.markFormatSupportsSubsetOfColumns("JSONLines");
@@ -418,25 +400,22 @@ void registerNonTrivialPrefixAndSuffixCheckerJSONEachRow(FormatFactory & factory
 
 void registerJSONEachRowSchemaReader(FormatFactory & factory)
 {
-    factory.registerSchemaReader("JSONEachRow", [](ReadBuffer & buf, const FormatSettings & settings)
+    auto register_schema_reader = [&](const String & format_name, bool json_strings)
     {
-        return std::make_unique<JSONEachRowSchemaReader>(buf, false, settings);
+        factory.registerSchemaReader(format_name, [json_strings](ReadBuffer & buf, const FormatSettings & settings)
-    });
+        {
+            return std::make_unique<JSONEachRowSchemaReader>(buf, json_strings, settings);
+        });
+        factory.registerAdditionalInfoForSchemaCacheGetter(format_name, [](const FormatSettings & settings)
+        {
+            return getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::JSON);
+        });
+    };
 
-    factory.registerSchemaReader("JSONStringsEachRow", [](ReadBuffer & buf, const FormatSettings & settings)
+    register_schema_reader("JSONEachRow", false);
-    {
+    register_schema_reader("JSONLines", false);
-        return std::make_unique<JSONEachRowSchemaReader>(buf, true, settings);
+    register_schema_reader("NDJSON", false);
-    });
+    register_schema_reader("JSONStringsEachRow", true);
-
-    factory.registerSchemaReader("JSONLines", [](ReadBuffer & buf, const FormatSettings & settings)
-    {
-        return std::make_unique<JSONEachRowSchemaReader>(buf, false, settings);
-    });
-
-    factory.registerSchemaReader("NDJSON", [](ReadBuffer & buf, const FormatSettings & settings)
-    {
-        return std::make_unique<JSONEachRowSchemaReader>(buf, false, settings);
-    });
 }
 
 }

src/Processors/Formats/Impl/MsgPackRowInputFormat.cpp

@@ -539,6 +539,14 @@ void registerMsgPackSchemaReader(FormatFactory & factory)
     {
         return std::make_shared<MsgPackSchemaReader>(buf, settings);
     });
+    factory.registerAdditionalInfoForSchemaCacheGetter("MsgPack", [](const FormatSettings & settings)
+    {
+            return fmt::format(
+                "number_of_columns={}, schema_inference_hints={}, max_rows_to_read_for_schema_inference={}",
+                settings.msgpack.number_of_columns,
+                settings.schema_inference_hints,
+                settings.max_rows_to_read_for_schema_inference);
+        });
 }
 
 }

src/Processors/Formats/Impl/MySQLDumpRowInputFormat.cpp

@@ -452,9 +452,6 @@ void registerInputFormatMySQLDump(FormatFactory & factory)
     {
         return std::make_shared<MySQLDumpRowInputFormat>(buf, header, params, settings);
     });
-
-    factory.registerAdditionalInfoForSchemaCacheGetter(
-        "MySQLDump", [](const FormatSettings & settings) { return "Table name: " + settings.mysql_dump.table_name; });
 }
 
 void registerMySQLSchemaReader(FormatFactory & factory)
@@ -463,6 +460,12 @@ void registerMySQLSchemaReader(FormatFactory & factory)
     {
         return std::make_shared<MySQLDumpSchemaReader>(buf, settings);
     });
+
+    factory.registerAdditionalInfoForSchemaCacheGetter("MySQLDump", [](const FormatSettings & settings)
+    {
+        auto result = getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::Quoted);
+        return result + fmt::format(", table_name={}", settings.mysql_dump.table_name);
+    });
 }
 
 

src/Processors/Formats/Impl/ProtobufListInputFormat.cpp

@@ -82,7 +82,7 @@ void registerInputFormatProtobufList(FormatFactory & factory)
             });
     factory.markFormatSupportsSubsetOfColumns("ProtobufList");
     factory.registerAdditionalInfoForSchemaCacheGetter(
-        "ProtobufList", [](const FormatSettings & settings) { return "Format schema: " + settings.schema.format_schema; });
+        "ProtobufList", [](const FormatSettings & settings) { return fmt::format("format_schema={}", settings.schema.format_schema); });
 }
 
 void registerProtobufListSchemaReader(FormatFactory & factory)

src/Processors/Formats/Impl/ProtobufRowInputFormat.cpp

@@ -104,7 +104,7 @@ void registerProtobufSchemaReader(FormatFactory & factory)
 
     for (const auto & name : {"Protobuf", "ProtobufSingle"})
         factory.registerAdditionalInfoForSchemaCacheGetter(
-            name, [](const FormatSettings & settings) { return "Format schema: " + settings.schema.format_schema; });
+            name, [](const FormatSettings & settings) { return fmt::format("format_schema={}", settings.schema.format_schema); });
 }
 
 }

src/Processors/Formats/Impl/RegexpRowInputFormat.cpp

@@ -211,6 +211,11 @@ void registerRegexpSchemaReader(FormatFactory & factory)
     {
         return std::make_shared<RegexpSchemaReader>(buf, settings);
     });
+    factory.registerAdditionalInfoForSchemaCacheGetter("Regexp", [](const FormatSettings & settings)
+    {
+        auto result = getAdditionalFormatInfoByEscapingRule(settings, settings.regexp.escaping_rule);
+        return result + fmt::format(", regexp={}", settings.regexp.regexp);
+    });
 }
 
 }

src/Processors/Formats/Impl/TSKVRowInputFormat.cpp

@@ -285,6 +285,10 @@ void registerTSKVSchemaReader(FormatFactory & factory)
     {
         return std::make_shared<TSKVSchemaReader>(buf, settings);
     });
+    factory.registerAdditionalInfoForSchemaCacheGetter("TSKV", [](const FormatSettings & settings)
+    {
+        return getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::Escaped);
+    });
 }
 
 }

src/Processors/Formats/Impl/TabSeparatedRowInputFormat.cpp

@@ -302,6 +302,14 @@ void registerTSVSchemaReader(FormatFactory & factory)
             {
                 return std::make_shared<TabSeparatedSchemaReader>(buf, with_names, with_types, is_raw, settings);
             });
+            factory.registerAdditionalInfoForSchemaCacheGetter(format_name, [with_names, is_raw](const FormatSettings & settings)
+            {
+                String result = getAdditionalFormatInfoByEscapingRule(
+                    settings, is_raw ? FormatSettings::EscapingRule::Raw : FormatSettings::EscapingRule::Escaped);
+                if (!with_names)
+                    result += fmt::format(", column_names_for_schema_inference={}", settings.column_names_for_schema_inference);
+                return result;
+            });
         };
 
         registerWithNamesAndTypes(is_raw ? "TabSeparatedRaw" : "TabSeparated", register_func);

src/Processors/Formats/Impl/TemplateRowInputFormat.cpp

@@ -566,13 +566,32 @@ void registerTemplateSchemaReader(FormatFactory & factory)
 {
     for (bool ignore_spaces : {false, true})
     {
-        factory.registerSchemaReader(ignore_spaces ? "TemplateIgnoreSpaces" : "Template", [ignore_spaces](ReadBuffer & buf, const FormatSettings & settings)
+        String format_name = ignore_spaces ? "TemplateIgnoreSpaces" : "Template";
+        factory.registerSchemaReader(format_name, [ignore_spaces](ReadBuffer & buf, const FormatSettings & settings)
         {
             size_t index = 0;
             auto idx_getter = [&](const String &) -> std::optional<size_t> { return index++; };
             auto row_format = fillRowFormat(settings, idx_getter, false);
             return std::make_shared<TemplateSchemaReader>(buf, ignore_spaces, fillResultSetFormat(settings), row_format, settings.template_settings.row_between_delimiter, settings);
         });
+        factory.registerAdditionalInfoForSchemaCacheGetter(format_name, [](const FormatSettings & settings)
+        {
+            size_t index = 0;
+            auto idx_getter = [&](const String &) -> std::optional<size_t> { return index++; };
+            auto row_format = fillRowFormat(settings, idx_getter, false);
+            std::unordered_set<FormatSettings::EscapingRule> visited_escaping_rules;
+            String result = fmt::format("row_format={}, resultset_format={}, row_between_delimiter={}",
+                settings.template_settings.row_format,
+                settings.template_settings.resultset_format,
+                settings.template_settings.row_between_delimiter);
+            for (auto escaping_rule : row_format.escaping_rules)
+            {
+                if (!visited_escaping_rules.contains(escaping_rule))
+                    result += ", " + getAdditionalFormatInfoByEscapingRule(settings, settings.regexp.escaping_rule);
+                visited_escaping_rules.insert(escaping_rule);
+            }
+            return result;
+        });
     }
 }
 

src/Processors/Formats/Impl/ValuesBlockInputFormat.cpp

@@ -634,6 +634,10 @@ void registerValuesSchemaReader(FormatFactory & factory)
     {
         return std::make_shared<ValuesSchemaReader>(buf, settings);
     });
+    factory.registerAdditionalInfoForSchemaCacheGetter("Values", [](const FormatSettings & settings)
+    {
+        return getAdditionalFormatInfoByEscapingRule(settings, FormatSettings::EscapingRule::Quoted);
+    });
 }
 
 }

src/Storages/Cache/SchemaCache.cpp

@@ -17,21 +17,21 @@ SchemaCache::SchemaCache(size_t max_elements_) : max_elements(max_elements_)
 {
 }
 
-void SchemaCache::add(const String & key, const ColumnsDescription & columns)
+void SchemaCache::add(const Key & key, const ColumnsDescription & columns)
 {
     std::lock_guard lock(mutex);
     addUnlocked(key, columns);
 }
 
 
-void SchemaCache::addMany(const Strings & keys, const ColumnsDescription & columns)
+void SchemaCache::addMany(const Keys & keys, const ColumnsDescription & columns)
 {
     std::lock_guard lock(mutex);
     for (const auto & key : keys)
         addUnlocked(key, columns);
 }
 
-void SchemaCache::addUnlocked(const String & key, const ColumnsDescription & columns)
+void SchemaCache::addUnlocked(const Key & key, const ColumnsDescription & columns)
 {
     /// Do nothing if this key is already in cache;
     if (data.contains(key))
@@ -54,7 +54,7 @@ void SchemaCache::checkOverflow()
     ProfileEvents::increment(ProfileEvents::SchemaInferenceCacheEvictions);
 }
 
-std::optional<ColumnsDescription> SchemaCache::tryGet(const String & key, LastModificationTimeGetter get_last_mod_time)
+std::optional<ColumnsDescription> SchemaCache::tryGet(const Key & key, LastModificationTimeGetter get_last_mod_time)
 {
     std::lock_guard lock(mutex);
     auto it = data.find(key);
@@ -101,10 +101,10 @@ void SchemaCache::clear()
     queue.clear();
 }
 
-std::unordered_map<String, SchemaCache::SchemaInfo> SchemaCache::getAll()
+std::unordered_map<SchemaCache::Key, SchemaCache::SchemaInfo, SchemaCache::KeyHash> SchemaCache::getAll()
 {
     std::lock_guard lock(mutex);
-    std::unordered_map<String, SchemaCache::SchemaInfo> result;
+    std::unordered_map<Key, SchemaCache::SchemaInfo, SchemaCache::KeyHash> result;
     for (const auto & [key, value] : data)
         result[key] = value.schema_info;
 

src/Storages/Cache/SchemaCache.h

@@ -23,6 +23,28 @@ class SchemaCache
 public:
     SchemaCache(size_t max_elements_);
 
+    struct Key
+    {
+        String source;
+        String format;
+        String additional_format_info;
+
+        bool operator==(const Key & other) const
+        {
+            return source == other.source && format == other.format && additional_format_info == other.additional_format_info;
+        }
+    };
+
+    using Keys = std::vector<Key>;
+
+    struct KeyHash
+    {
+        size_t operator()(const Key & key) const
+        {
+            return std::hash<String>()(key.source + key.format + key.additional_format_info);
+        }
+    };
+
     struct SchemaInfo
     {
         ColumnsDescription columns;
@@ -32,22 +54,22 @@ public:
     using LastModificationTimeGetter = std::function<std::optional<time_t>()>;
 
     /// Add new key with a schema
-    void add(const String & key, const ColumnsDescription & columns);
+    void add(const Key & key, const ColumnsDescription & columns);
 
     /// Add many keys with the same schema (usually used for globs)
-    void addMany(const Strings & keys, const ColumnsDescription & columns);
+    void addMany(const Keys & keys, const ColumnsDescription & columns);
 
-    std::optional<ColumnsDescription> tryGet(const String & key, LastModificationTimeGetter get_last_mod_time = {});
+    std::optional<ColumnsDescription> tryGet(const Key & key, LastModificationTimeGetter get_last_mod_time = {});
 
     void clear();
 
-    std::unordered_map<String, SchemaInfo> getAll();
+    std::unordered_map<Key, SchemaInfo, SchemaCache::KeyHash> getAll();
 
 private:
-    void addUnlocked(const String & key, const ColumnsDescription & columns);
+    void addUnlocked(const Key & key, const ColumnsDescription & columns);
     void checkOverflow();
 
-    using Queue = std::list<String>;
+    using Queue = std::list<Key>;
     using QueueIterator = Queue::iterator;
 
     struct Cell
@@ -57,7 +79,7 @@ private:
     };
 
     Queue queue;
-    std::unordered_map<String, Cell> data;
+    std::unordered_map<Key, Cell, KeyHash> data;
 
     size_t max_elements;
     std::mutex mutex;

src/Storages/HDFS/StorageHDFS.cpp

@@ -774,7 +774,7 @@ std::optional<ColumnsDescription> StorageHDFS::tryGetColumnsFromCache(
         };
 
         String url = fs::path(uri_without_path) / path;
-        String cache_key = getKeyForSchemaCache(url, format_name, {}, ctx);
+        auto cache_key = getKeyForSchemaCache(url, format_name, {}, ctx);
         auto columns = schema_cache.tryGet(cache_key, get_last_mod_time);
         if (columns)
             return columns;
@@ -794,7 +794,7 @@ void StorageHDFS::addColumnsToCache(
     Strings sources;
     sources.reserve(paths.size());
     std::transform(paths.begin(), paths.end(), std::back_inserter(sources), [&](const String & path){ return fs::path(uri_without_path) / path; });
-    Strings cache_keys = getKeysForSchemaCache(sources, format_name, {}, ctx);
+    auto cache_keys = getKeysForSchemaCache(sources, format_name, {}, ctx);
     schema_cache.addMany(cache_keys, columns);
 }
 

src/Storages/StorageFile.cpp

@@ -1257,7 +1257,7 @@ std::optional<ColumnsDescription> StorageFile::tryGetColumnsFromCache(
             return file_stat.st_mtime;
         };
 
-        String cache_key = getKeyForSchemaCache(path, format_name, format_settings, context);
+        auto cache_key = getKeyForSchemaCache(path, format_name, format_settings, context);
         auto columns = schema_cache.tryGet(cache_key, get_last_mod_time);
         if (columns)
             return columns;
@@ -1274,7 +1274,7 @@ void StorageFile::addColumnsToCache(
     const ContextPtr & context)
 {
     auto & schema_cache = getSchemaCache(context);
-    Strings cache_keys = getKeysForSchemaCache(paths, format_name, format_settings, context);
+    auto cache_keys = getKeysForSchemaCache(paths, format_name, format_settings, context);
     schema_cache.addMany(cache_keys, columns);
 }
 

src/Storages/StorageS3.cpp

@@ -1389,7 +1389,7 @@ std::optional<ColumnsDescription> StorageS3::tryGetColumnsFromCache(
         };
 
         String source = fs::path(s3_configuration.uri.uri.getHost() + std::to_string(s3_configuration.uri.uri.getPort())) / path;
-        String cache_key = getKeyForSchemaCache(source, format_name, format_settings, ctx);
+        auto cache_key = getKeyForSchemaCache(source, format_name, format_settings, ctx);
         auto columns = schema_cache.tryGet(cache_key, get_last_mod_time);
         if (columns)
             return columns;
@@ -1410,7 +1410,7 @@ void StorageS3::addColumnsToCache(
     Strings sources;
     sources.reserve(keys.size());
     std::transform(keys.begin(), keys.end(), std::back_inserter(sources), [&](const String & key){ return host_and_bucket / key; });
-    Strings cache_keys = getKeysForSchemaCache(sources, format_name, format_settings, ctx);
+    auto cache_keys = getKeysForSchemaCache(sources, format_name, format_settings, ctx);
     auto & schema_cache = getSchemaCache(ctx);
     schema_cache.addMany(cache_keys, columns);
 }

src/Storages/StorageURL.cpp

@@ -836,7 +836,7 @@ std::optional<ColumnsDescription> IStorageURLBase::tryGetColumnsFromCache(
             return last_mod_time;
         };
 
-        String cache_key = getKeyForSchemaCache(url, format_name, format_settings, context);
+        auto cache_key = getKeyForSchemaCache(url, format_name, format_settings, context);
         auto columns = schema_cache.tryGet(cache_key, get_last_mod_time);
         if (columns)
             return columns;
@@ -853,7 +853,7 @@ void IStorageURLBase::addColumnsToCache(
     const ContextPtr & context)
 {
     auto & schema_cache = getSchemaCache(context);
-    Strings cache_keys = getKeysForSchemaCache(urls, format_name, format_settings, context);
+    auto cache_keys = getKeysForSchemaCache(urls, format_name, format_settings, context);
     schema_cache.addMany(cache_keys, columns);
 }
 

src/Storages/System/StorageSystemSchemaInferenceCache.cpp

@@ -44,16 +44,13 @@ NamesAndTypesList StorageSystemSchemaInferenceCache::getNamesAndTypes()
 static void fillDataImpl(MutableColumns & res_columns, SchemaCache & schema_cache, const String & storage_name)
 {
     auto s3_schema_cache_data = schema_cache.getAll();
-    String source;
+
-    String format;
-    String additional_format_info;
     for (const auto & [key, schema_info] : s3_schema_cache_data)
     {
-        splitSchemaCacheKey(key, source, format, additional_format_info);
         res_columns[0]->insert(storage_name);
-        res_columns[1]->insert(source);
+        res_columns[1]->insert(key.source);
-        res_columns[2]->insert(format);
+        res_columns[2]->insert(key.format);
-        res_columns[3]->insert(additional_format_info);
+        res_columns[3]->insert(key.additional_format_info);
         res_columns[4]->insert(schema_info.registration_time);
         res_columns[5]->insert(getSchemaString(schema_info.columns));
     }

tests/queries/0_stateless/02318_template_schema_inference_bug.sql

@@ -1,2 +1,2 @@
 insert into function file(data_02318.tsv) select * from numbers(10);
-desc file('data_02318.tsv', 'Template') SETTINGS format_template_row='nonexist', format_template_resultset='nonexist'; -- {serverError CANNOT_EXTRACT_TABLE_STRUCTURE}
+desc file('data_02318.tsv', 'Template') SETTINGS format_template_row='nonexist', format_template_resultset='nonexist'; -- {serverError FILE_DOESNT_EXIST}

tests/queries/0_stateless/02404_schema_inference_cache_respect_format_settings.reference

@@ -0,0 +1,90 @@
+TSV
+c1	Nullable(Int64)					
+c2	Nullable(Date)					
+c1	Nullable(Float64)					
+c2	Nullable(Date)					
+c1	Nullable(Int64)					
+c2	Nullable(DateTime64(9))					
+c1	UInt8					
+c2	Nullable(Date)					
+4
+TSVWithNames
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Float64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Int64)					
+toDate(number)	Nullable(DateTime64(9))					
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+4
+CSV
+c1	Nullable(Int64)					
+c2	Nullable(Date)					
+c1	Nullable(Float64)					
+c2	Nullable(Date)					
+c1	Nullable(Int64)					
+c2	Nullable(DateTime64(9))					
+c1	UInt8					
+c2	Nullable(Date)					
+4
+CSVWithNames
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Float64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Int64)					
+toDate(number)	Nullable(DateTime64(9))					
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+4
+TSKV
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Float64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Int64)					
+toDate(number)	Nullable(DateTime64(9))					
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+4
+CustomSeparated
+c1	Nullable(Int64)					
+c2	Nullable(Date)					
+c1	Nullable(Float64)					
+c2	Nullable(Date)					
+c1	Nullable(Int64)					
+c2	Nullable(DateTime64(9))					
+c1	UInt8					
+c2	Nullable(Date)					
+4
+JSONEachRow
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Float64)					
+toDate(number)	Nullable(Date)					
+number	Nullable(Int64)					
+toDate(number)	Nullable(DateTime64(9))					
+number	Nullable(Int64)					
+toDate(number)	Nullable(Date)					
+4
+JSONCompactEachRow
+c1	Nullable(Int64)					
+c2	Nullable(Date)					
+c1	Nullable(Float64)					
+c2	Nullable(Date)					
+c1	Nullable(Int64)					
+c2	Nullable(DateTime64(9))					
+c1	UInt8					
+c2	Nullable(Date)					
+4
+Values
+c1	Nullable(Int64)					
+c2	Nullable(Date)					
+c1	Nullable(Float64)					
+c2	Nullable(Date)					
+c1	Nullable(Int64)					
+c2	Nullable(DateTime64(9))					
+c1	UInt8					
+c2	Nullable(Date)					
+4

tests/queries/0_stateless/02404_schema_inference_cache_respect_format_settings.sql.j2

@@ -0,0 +1,16 @@
+-- Tags: no-parallel, no-fasttest
+
+system drop schema cache for file;
+
+{% for format in ['TSV', 'TSVWithNames', 'CSV', 'CSVWithNames', 'TSKV', 'CustomSeparated', 'JSONEachRow', 'JSONCompactEachRow', 'Values'] -%}
+
+select '{{ format }}';
+insert into function file(02404_data.{{ format }}) select number, toDate(number) from numbers(10);
+desc file(02404_data.{{ format }});
+desc file(02404_data.{{ format }}) settings input_format_try_infer_integers=0;
+desc file(02404_data.{{ format }}) settings input_format_try_infer_dates=0;
+desc file(02404_data.{{ format }}) settings schema_inference_hints='c1 UInt8';
+select count() from system.schema_inference_cache where countSubstrings(source, '02404_data.{{ format }}') > 0;
+
+{% endfor -%}
+