Merge pull request #49420 from ClickHouse/fix_00002

Fix some bad error messages

src/AggregateFunctions/AggregateFunctionMinMaxAny.h

@@ -554,7 +554,8 @@ public:
         if (capacity < size_to_reserve)
         {
             if (unlikely(MAX_STRING_SIZE < size_to_reserve))
-                throw Exception(ErrorCodes::TOO_LARGE_STRING_SIZE, "String size is too big ({})", size_to_reserve);
+                throw Exception(ErrorCodes::TOO_LARGE_STRING_SIZE, "String size is too big ({}), maximum: {}",
+                                size_to_reserve, MAX_STRING_SIZE);
 
             size_t rounded_capacity = roundUpToPowerOfTwoOrZero(size_to_reserve);
             chassert(rounded_capacity <= MAX_STRING_SIZE + 1);  /// rounded_capacity <= 2^31
@@ -624,7 +625,8 @@ public:
     void changeImpl(StringRef value, Arena * arena)
     {
         if (unlikely(MAX_STRING_SIZE < value.size))
-            throw Exception(ErrorCodes::TOO_LARGE_STRING_SIZE, "String size is too big ({})", value.size);
+            throw Exception(ErrorCodes::TOO_LARGE_STRING_SIZE, "String size is too big ({}), maximum: {}",
+                            value.size, MAX_STRING_SIZE);
 
         UInt32 value_size = static_cast<UInt32>(value.size);
 

src/Databases/DatabaseReplicated.cpp

@@ -1048,7 +1048,7 @@ void DatabaseReplicated::dropReplica(
     assert(!database || database_zookeeper_path == database->zookeeper_path);
 
     if (full_replica_name.find('/') != std::string::npos)
-        throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid replica name: {}", full_replica_name);
+        throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid replica name, '/' is not allowed: {}", full_replica_name);
 
     auto zookeeper = Context::getGlobalContextInstance()->getZooKeeper();
 

src/Functions/parseDateTime.cpp

@@ -150,7 +150,7 @@ namespace
             if (text == "bc")
                 throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Era BC exceeds the range of DateTime");
             else if (text != "ad")
-                throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Unknown era {}", text);
+                throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Unknown era {} (expected 'ad' or 'bc')", text);
         }
 
         void setCentury(Int32 century)

src/IO/HTTPChunkedReadBuffer.cpp

@@ -33,7 +33,7 @@ size_t HTTPChunkedReadBuffer::readChunkHeader()
     } while (!in->eof() && isHexDigit(*in->position()));
 
     if (res > max_chunk_size)
-        throw Exception(ErrorCodes::ARGUMENT_OUT_OF_BOUND, "Chunk size exceeded the limit");
+        throw Exception(ErrorCodes::ARGUMENT_OUT_OF_BOUND, "Chunk size exceeded the limit (max size: {})", max_chunk_size);
 
     /// NOTE: If we want to read any chunk extensions, it should be done here.
 

src/Processors/Transforms/WindowTransform.cpp

@@ -2012,7 +2012,7 @@ struct WindowFunctionNtile final : public WindowFunction
 
             if (!buckets)
             {
-                throw Exception(ErrorCodes::BAD_ARGUMENTS, "ntile's argument must > 0");
+                throw Exception(ErrorCodes::BAD_ARGUMENTS, "ntile's argument must be greater than 0");
             }
         }
         // new partition

tests/clickhouse-test

@@ -2117,7 +2117,14 @@ def reportLogStats(args):
               'Column ''{}'' already exists', 'No macro {} in config', 'Invalid origin H3 index: {}',
               'Invalid session timeout: ''{}''', 'Tuple cannot be empty', 'Database name is empty',
               'Table {} is not a Dictionary', 'Expected function, got: {}', 'Unknown identifier: ''{}''',
-              'Failed to {} input ''{}''', '{}.{} is not a VIEW', 'Cannot convert NULL to {}', 'Dictionary {} doesn''t exist'
+              'Failed to {} input ''{}''', '{}.{} is not a VIEW', 'Cannot convert NULL to {}', 'Dictionary {} doesn''t exist',
+              'Write file: {}', 'Unable to parse JSONPath', 'Host is empty in S3 URI.', 'Expected end of line',
+              'inflate failed: {}{}', 'Center is not valid', 'Column ''{}'' is ambiguous', 'Cannot parse object', 'Invalid date: {}',
+              'There is no cache by name: {}', 'No part {} in table', '`{}` should be a String', 'There are duplicate id {}',
+              'Invalid replica name: {}', 'Unexpected value {} in enum', 'Unknown BSON type: {}', 'Point is not valid',
+              'Invalid qualified name: {}', 'INTO OUTFILE is not allowed', 'Arguments must not be NaN', 'Cell is not valid',
+              'brotli decode error{}', 'Invalid H3 index: {}', 'Too large node state size', 'No additional keys found.',
+              'Attempt to read after EOF.', 'Replication was stopped', '{}	building file infos', 'Cannot parse uuid {}'
         ) AS known_short_messages
         SELECT count() AS c, message_format_string, substr(any(message), 1, 120)
         FROM system.text_log

tests/queries/0_stateless/00002_log_and_exception_messages_formatting.reference

@@ -2,7 +2,7 @@ runtime messages	0.001
 runtime exceptions	0.05
 messages shorter than 10	1
 messages shorter than 16	3
-exceptions shorter than 30	30
+exceptions shorter than 30	3
 noisy messages	0.3
 noisy Trace messages	0.16
 noisy Debug messages	0.09

tests/queries/0_stateless/00002_log_and_exception_messages_formatting.sql

@@ -49,7 +49,14 @@ create temporary table known_short_messages (s String) as select * from (select
 'Column ''{}'' already exists', 'No macro {} in config', 'Invalid origin H3 index: {}',
 'Invalid session timeout: ''{}''', 'Tuple cannot be empty', 'Database name is empty',
 'Table {} is not a Dictionary', 'Expected function, got: {}', 'Unknown identifier: ''{}''',
-'Failed to {} input ''{}''', '{}.{} is not a VIEW', 'Cannot convert NULL to {}', 'Dictionary {} doesn''t exist'
+'Failed to {} input ''{}''', '{}.{} is not a VIEW', 'Cannot convert NULL to {}', 'Dictionary {} doesn''t exist',
+'Write file: {}', 'Unable to parse JSONPath', 'Host is empty in S3 URI.', 'Expected end of line',
+'inflate failed: {}{}', 'Center is not valid', 'Column ''{}'' is ambiguous', 'Cannot parse object', 'Invalid date: {}',
+'There is no cache by name: {}', 'No part {} in table', '`{}` should be a String', 'There are duplicate id {}',
+'Invalid replica name: {}', 'Unexpected value {} in enum', 'Unknown BSON type: {}', 'Point is not valid',
+'Invalid qualified name: {}', 'INTO OUTFILE is not allowed', 'Arguments must not be NaN', 'Cell is not valid',
+'brotli decode error{}', 'Invalid H3 index: {}', 'Too large node state size', 'No additional keys found.',
+'Attempt to read after EOF.', 'Replication was stopped', '{}	building file infos', 'Cannot parse uuid {}'
 ] as arr) array join arr;
 
 -- Check that we don't have too many short meaningless message patterns.
@@ -59,7 +66,7 @@ select 'messages shorter than 10', max2(countDistinctOrDefault(message_format_st
 select 'messages shorter than 16', max2(countDistinctOrDefault(message_format_string), 3) from logs where length(message_format_string) < 16 and message_format_string not in known_short_messages;
 
 -- Same as above, but exceptions must be more informative. Feel free to update the threshold or remove this query if really necessary
-select 'exceptions shorter than 30', max2(countDistinctOrDefault(message_format_string), 30) from logs where length(message_format_string) < 30 and message ilike '%DB::Exception%' and message_format_string not in known_short_messages;
+select 'exceptions shorter than 30', max2(countDistinctOrDefault(message_format_string), 3) from logs where length(message_format_string) < 30 and message ilike '%DB::Exception%' and message_format_string not in known_short_messages;
 
 
 -- Avoid too noisy messages: top 1 message frequency must be less than 30%. We should reduce the threshold
@@ -98,7 +105,9 @@ select 'incorrect patterns', max2(countDistinct(message_format_string), 15) from
     where ((rand() % 8) = 0)
     and message not like (replaceRegexpAll(message_format_string, '{[:.0-9dfx]*}', '%') as s)
     and message not like (s || ' (skipped % similar messages)')
-    and message not like ('%Exception: '||s||'%') group by message_format_string
+    and message not like ('%Exception: '||s||'%')
+    and message not like ('%(skipped % similar messages)%')
+    group by message_format_string
 ) where any_message not like '%Poco::Exception%';
 
 drop table logs;