Merge pull request #14030 from ClickHouse/bharatnc-ncb/timezones-table

Merging #13880

contrib/cctz-cmake/CMakeLists.txt

@@ -20,12 +20,18 @@ if (NOT USE_INTERNAL_CCTZ_LIBRARY)
         )
 
         if (NOT EXTERNAL_CCTZ_LIBRARY_WORKS)
-            message (${RECONFIGURE_MESSAGE_LEVEL}  "External cctz is not working: ${LIBRARY_CCTZ} ${INCLUDE_CCTZ}")
+            message (${RECONFIGURE_MESSAGE_LEVEL} "External cctz is not working: ${LIBRARY_CCTZ} ${INCLUDE_CCTZ}")
         else()
             add_library (cctz UNKNOWN IMPORTED)
             set_property (TARGET cctz PROPERTY IMPORTED_LOCATION ${LIBRARY_CCTZ})
             set_property (TARGET cctz PROPERTY INTERFACE_INCLUDE_DIRECTORIES ${INCLUDE_CCTZ})
         endif()
+        
+        set(SYSTEM_STORAGE_TZ_FILE "${CMAKE_BINARY_DIR}/src/Storages/System/StorageSystemTimeZones.generated.cpp")
+        file(REMOVE ${SYSTEM_STORAGE_TZ_FILE})
+        file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "// autogenerated by ClickHouse/contrib/cctz-cmake/CMakeLists.txt\n")
+        file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "const char * auto_time_zones[] {nullptr};\n" )
+
     else()
         set (EXTERNAL_CCTZ_LIBRARY_FOUND 0)
         message (${RECONFIGURE_MESSAGE_LEVEL} "Can't find system cctz")
@@ -57,8 +63,14 @@ if (NOT EXTERNAL_CCTZ_LIBRARY_FOUND OR NOT EXTERNAL_CCTZ_LIBRARY_WORKS)
         target_compile_definitions (cctz PRIVATE __USE_BSD linux _XOPEN_SOURCE=600)
     endif ()
 
+    # Related to time_zones table:
+    # StorageSystemTimeZones.generated.cpp is autogenerated each time during a build
+    # data in this file will be used to populate the system.time_zones table, this is specific to OS_LINUX
+    # as the library that's built using embedded tzdata is also specific to OS_LINUX
+    set(SYSTEM_STORAGE_TZ_FILE "${CMAKE_BINARY_DIR}/src/Storages/System/StorageSystemTimeZones.generated.cpp")
+    # remove existing copies so that its generated fresh on each build.
+    file(REMOVE ${SYSTEM_STORAGE_TZ_FILE})
     # Build a libray with embedded tzdata
-
     if (OS_LINUX)
         # get the list of timezones from tzdata shipped with cctz
         set(TZDIR ${LIBRARY_DIR}/testdata/zoneinfo)
@@ -69,9 +81,16 @@ if (NOT EXTERNAL_CCTZ_LIBRARY_FOUND OR NOT EXTERNAL_CCTZ_LIBRARY_WORKS)
         set(TZ_OBJS)
 
         # each file/symlink in that dir (except of tab and localtime) store the info about timezone
-        execute_process(COMMAND bash -c "cd ${TZDIR} && find * -type f,l -and ! -name '*.tab' -and ! -name 'localtime' | sort | paste -sd ';'" OUTPUT_STRIP_TRAILING_WHITESPACE OUTPUT_VARIABLE TIMEZONES )
+        execute_process(COMMAND
+            bash -c "cd ${TZDIR} && find * -type f,l -and ! -name '*.tab' -and ! -name 'localtime' | sort | paste -sd ';'"
+            OUTPUT_STRIP_TRAILING_WHITESPACE
+            OUTPUT_VARIABLE TIMEZONES)
+
+        file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "// autogenerated by ClickHouse/contrib/cctz-cmake/CMakeLists.txt\n")
+        file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "const char * auto_time_zones[] {\n" )
 
         foreach(TIMEZONE ${TIMEZONES})
+            file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "    \"${TIMEZONE}\",\n")
             string(REPLACE "/" "_" TIMEZONE_ID ${TIMEZONE})
             string(REPLACE "+" "_PLUS_" TIMEZONE_ID ${TIMEZONE_ID})
             set(TZ_OBJ ${TIMEZONE_ID}.o)
@@ -87,6 +106,8 @@ if (NOT EXTERNAL_CCTZ_LIBRARY_FOUND OR NOT EXTERNAL_CCTZ_LIBRARY_WORKS)
             set_source_files_properties(${TZ_OBJ} PROPERTIES EXTERNAL_OBJECT true GENERATED true)
         endforeach(TIMEZONE)
 
+        file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "    nullptr};\n")
+
         add_library(tzdata STATIC ${TZ_OBJS})
         set_target_properties(tzdata PROPERTIES LINKER_LANGUAGE C)
         # whole-archive prevents symbols from being discarded for unknown reason
@@ -95,6 +116,9 @@ if (NOT EXTERNAL_CCTZ_LIBRARY_FOUND OR NOT EXTERNAL_CCTZ_LIBRARY_WORKS)
         # library into single string.
         add_dependencies(cctz tzdata)
         target_link_libraries(cctz INTERFACE "-Wl,${WHOLE_ARCHIVE} $<TARGET_FILE:tzdata> -Wl,${NO_WHOLE_ARCHIVE}")
+    else ()
+        file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "// autogenerated by ClickHouse/contrib/cctz-cmake/CMakeLists.txt\n")
+        file(APPEND ${SYSTEM_STORAGE_TZ_FILE} "const char * auto_time_zones[] {nullptr};\n" )
     endif ()
 endif ()
 

docs/en/operations/system-tables/time_zones.md

@@ -0,0 +1,30 @@
+# system.time_zones {#system-time_zones}
+
+Contains a list of time zones that are supported by the ClickHouse server. This list of timezones might vary depending on the version of ClickHouse.
+
+Columns:
+
+-   `time_zone` (String) — List of supported time zones.
+
+**Example**
+
+``` sql
+SELECT * FROM system.time_zones LIMIT 10
+```
+
+``` text
+┌─time_zone──────────┐
+│ Africa/Abidjan     │
+│ Africa/Accra       │
+│ Africa/Addis_Ababa │
+│ Africa/Algiers     │
+│ Africa/Asmara      │
+│ Africa/Asmera      │
+│ Africa/Bamako      │
+│ Africa/Bangui      │
+│ Africa/Banjul      │
+│ Africa/Bissau      │
+└────────────────────┘
+```
+
+[Original article](https://clickhouse.tech/docs/en/operations/system_tables/time_zones) <!--hide-->

src/Storages/System/CMakeLists.txt

@@ -29,6 +29,7 @@ endif()
 add_dependencies(generate-source generate-contributors)
 
 set(GENERATED_LICENSES_SRC ${CMAKE_CURRENT_BINARY_DIR}/StorageSystemLicenses.generated.cpp)
+set(GENERATED_TIMEZONES_SRC ${CMAKE_CURRENT_BINARY_DIR}/StorageSystemTimeZones.generated.cpp)
 
 add_custom_command(
     OUTPUT StorageSystemLicenses.generated.cpp
@@ -36,6 +37,8 @@ add_custom_command(
     WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
 
 list (APPEND storages_system_sources ${GENERATED_LICENSES_SRC})
+list (APPEND storages_system_sources ${GENERATED_TIMEZONES_SRC})
+
 # Overlength strings
 set_source_files_properties(${GENERATED_LICENSES_SRC} PROPERTIES COMPILE_FLAGS -w)
 

src/Storages/System/StorageSystemTimeZones.cpp

@@ -0,0 +1,23 @@
+#include "StorageSystemTimeZones.h"
+
+#include <algorithm>
+#include <DataTypes/DataTypeString.h>
+
+
+extern const char * auto_time_zones[];
+
+namespace DB
+{
+NamesAndTypesList StorageSystemTimeZones::getNamesAndTypes()
+{
+    return {
+        {"time_zone", std::make_shared<DataTypeString>()},
+    };
+}
+
+void StorageSystemTimeZones::fillData(MutableColumns & res_columns, const Context &, const SelectQueryInfo &) const
+{
+    for (auto * it = auto_time_zones; *it; ++it)
+        res_columns[0]->insert(String(*it));
+}
+}

src/Storages/System/StorageSystemTimeZones.h

@@ -0,0 +1,29 @@
+#pragma once
+
+#include <Storages/System/IStorageSystemOneBlock.h>
+#include <ext/shared_ptr_helper.h>
+
+
+namespace DB
+{
+class Context;
+
+
+/** System table "time_zones" with list of timezones pulled from /contrib/cctz/testdata/zoneinfo
+  */
+class StorageSystemTimeZones final : public ext::shared_ptr_helper<StorageSystemTimeZones>,
+                                     public IStorageSystemOneBlock<StorageSystemTimeZones>
+{
+    friend struct ext::shared_ptr_helper<StorageSystemTimeZones>;
+
+protected:
+    void fillData(MutableColumns & res_columns, const Context & context, const SelectQueryInfo & query_info) const override;
+
+    using IStorageSystemOneBlock::IStorageSystemOneBlock;
+
+public:
+    std::string getName() const override { return "SystemTimeZones"; }
+
+    static NamesAndTypesList getNamesAndTypes();
+};
+}

src/Storages/System/attachSystemTables.cpp

@@ -39,6 +39,7 @@
 #include <Storages/System/StorageSystemContributors.h>
 #if !defined(ARCADIA_BUILD)
     #include <Storages/System/StorageSystemLicenses.h>
+    #include <Storages/System/StorageSystemTimeZones.h>
 #endif
 #include <Storages/System/StorageSystemDisks.h>
 #include <Storages/System/StorageSystemStoragePolicies.h>
@@ -105,9 +106,9 @@ void attachSystemTablesLocal(IDatabase & system_database)
     attach<StorageSystemQuotasUsage>(system_database, "quotas_usage");
     attach<StorageSystemUserDirectories>(system_database, "user_directories");
     attach<StorageSystemPrivileges>(system_database, "privileges");
-
 #if !defined(ARCADIA_BUILD)
     attach<StorageSystemLicenses>(system_database, "licenses");
+    attach<StorageSystemTimeZones>(system_database, "time_zones");
 #endif
 #ifdef OS_LINUX
     attach<StorageSystemStackTrace>(system_database, "stack_trace");

src/Storages/ya.make.in

@@ -7,7 +7,7 @@ PEERDIR(
 )
 
 SRCS(
-<? find . -name '*.cpp' | grep -v -F tests | grep -v -P 'Kafka|RabbitMQ|S3|HDFS|Licenses' | sed 's/^\.\//    /' | sort ?>
+<? find . -name '*.cpp' | grep -v -F tests | grep -v -P 'Kafka|RabbitMQ|S3|HDFS|Licenses|TimeZones' | sed 's/^\.\//    /' | sort ?>
 )
 
 END()

tests/queries/0_stateless/01455_time_zones.reference

@@ -0,0 +1 @@
+ok

tests/queries/0_stateless/01455_time_zones.sql

@@ -0,0 +1,9 @@
+-- There are currently 594 timezones which are used from the binary embedded inside the ClickHouse binary
+-- Refer: contrib/cctz-cmake/CMakeLists.txt for the complete list. The count may change but we expect there will be at least 500 timezones.
+-- SELECT count(*)
+-- FROM system.time_zones
+--
+-- ┌─count()─┐
+-- │     594 │
+-- └─────────┘
+SELECT if ((SELECT count(*) FROM system.time_zones) > 500, 'ok', 'fail');

tests/queries/0_stateless/arcadia_skip_list.txt

@@ -135,4 +135,5 @@
 01391_join_on_dict_crash
 01401_FORMAT_SETTINGS
 01411_bayesian_ab_testing
+01455_time_zones
 01456_ast_optimizations_over_distributed

tests/queries/skip_list.json

@@ -90,7 +90,8 @@
         "01300_client_save_history_when_terminated",
         "orc_output",
         "01370_client_autocomplete_word_break_characters",
-        "01193_metadata_loading"
+        "01193_metadata_loading",
+        "01455_time_zones"
     ],
     "release-build": [
         "avx2"