thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-26 09:32:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #57053 from Avogar/better-parsing-exceptions

Browse Source 
Better exception messages in input formats
This commit is contained in:
Kruglov Pavel 2024-01-03 14:47:51 +01:00 committed by GitHub
commit 570d1c013b
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
50 changed files with 176 additions and 287 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/Common/CounterInFile.h
View File

@ -88,7 +88,7 @@ public:
{ {
/// A more understandable error message. /// A more understandable error message.
if (e.code() == DB::ErrorCodes::CANNOT_READ_ALL_DATA || e.code() == DB::ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF) if (e.code() == DB::ErrorCodes::CANNOT_READ_ALL_DATA || e.code() == DB::ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF)
throw DB::ParsingException(e.code(), "File {} is empty. You must fill it manually with appropriate value.", path); throw DB::Exception(e.code(), "File {} is empty. You must fill it manually with appropriate value.", path);
else else
throw; throw;
} }

44
src/Common/Exception.cpp
View File

@ -616,48 +616,4 @@ ExecutionStatus ExecutionStatus::fromText(const std::string & data)
return status; return status;
} }
ParsingException::ParsingException() = default;
ParsingException::ParsingException(const std::string & msg, int code)
: Exception(msg, code)
{
}
/// We use additional field formatted_message_ to make this method const.
std::string ParsingException::displayText() const
{
try
{
formatted_message = message();
bool need_newline = false;
if (!file_name.empty())
{
formatted_message += fmt::format(": (in file/uri {})", file_name);
need_newline = true;
}
if (line_number != -1)
{
formatted_message += fmt::format(": (at row {})", line_number);
need_newline = true;
}
if (need_newline)
formatted_message += "\n";
}
catch (...) {} // NOLINT(bugprone-empty-catch)
if (!formatted_message.empty())
{
std::string result = name();
result.append(": ");
result.append(formatted_message);
return result;
}
else
{
return Exception::displayText();
}
}
} }

37
src/Common/Exception.h
View File

@ -235,43 +235,6 @@ private:
const char * className() const noexcept override { return "DB::ErrnoException"; } const char * className() const noexcept override { return "DB::ErrnoException"; }
}; };
/// Special class of exceptions, used mostly in ParallelParsingInputFormat for
/// more convenient calculation of problem line number.
class ParsingException : public Exception
{
ParsingException(const std::string & msg, int code);
public:
ParsingException();
// Format message with fmt::format, like the logging functions.
template <typename... Args>
ParsingException(int code, FormatStringHelper<Args...> fmt, Args &&... args) : Exception(fmt::format(fmt.fmt_str, std::forward<Args>(args)...), code)
{
message_format_string = fmt.message_format_string;
}
std::string displayText() const override;
ssize_t getLineNumber() const { return line_number; }
void setLineNumber(int line_number_) { line_number = line_number_;}
String getFileName() const { return file_name; }
void setFileName(const String & file_name_) { file_name = file_name_; }
Exception * clone() const override { return new ParsingException(*this); }
void rethrow() const override { throw *this; } // NOLINT
private:
ssize_t line_number{-1};
String file_name;
mutable std::string formatted_message;
const char * name() const noexcept override { return "DB::ParsingException"; }
const char * className() const noexcept override { return "DB::ParsingException"; }
};
using Exceptions = std::vector<std::exception_ptr>; using Exceptions = std::vector<std::exception_ptr>;
/** Try to write an exception to the log (and forget about it). /** Try to write an exception to the log (and forget about it).

4
src/DataTypes/Serializations/SerializationArray.cpp
View File

@ -390,7 +390,7 @@ void SerializationArray::deserializeBinaryBulkWithMultipleStreams(
/// Check consistency between offsets and elements subcolumns. /// Check consistency between offsets and elements subcolumns.
/// But if elements column is empty - it's ok for columns of Nested types that was added by ALTER. /// But if elements column is empty - it's ok for columns of Nested types that was added by ALTER.
if (!nested_column->empty() && nested_column->size() != last_offset) if (!nested_column->empty() && nested_column->size() != last_offset)
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Cannot read all array values: read just {} of {}", throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Cannot read all array values: read just {} of {}",
toString(nested_column->size()), toString(last_offset)); toString(nested_column->size()), toString(last_offset));
column = std::move(mutable_column); column = std::move(mutable_column);
@ -445,7 +445,7 @@ static void deserializeTextImpl(IColumn & column, ReadBuffer & istr, Reader && r
if (*istr.position() == ',') if (*istr.position() == ',')
++istr.position(); ++istr.position();
else else
throw ParsingException(ErrorCodes::CANNOT_READ_ARRAY_FROM_TEXT, throw Exception(ErrorCodes::CANNOT_READ_ARRAY_FROM_TEXT,
"Cannot read array from text, expected comma or end of array, found '{}'", "Cannot read array from text, expected comma or end of array, found '{}'",
*istr.position()); *istr.position());
} }

10
src/DataTypes/Serializations/SerializationNullable.cpp
View File

@ -359,7 +359,7 @@ ReturnType SerializationNullable::deserializeTextEscapedAndRawImpl(IColumn & col
nested_column.popBack(1); nested_column.popBack(1);
if (null_representation.find('\t') != std::string::npos || null_representation.find('\n') != std::string::npos) if (null_representation.find('\t') != std::string::npos || null_representation.find('\n') != std::string::npos)
throw DB::ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "TSV custom null representation " throw DB::Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "TSV custom null representation "
"containing '\\t' or '\\n' may not work correctly for large input."); "containing '\\t' or '\\n' may not work correctly for large input.");
WriteBufferFromOwnString parsed_value; WriteBufferFromOwnString parsed_value;
@ -367,7 +367,7 @@ ReturnType SerializationNullable::deserializeTextEscapedAndRawImpl(IColumn & col
nested_serialization->serializeTextEscaped(nested_column, nested_column.size() - 1, parsed_value, settings); nested_serialization->serializeTextEscaped(nested_column, nested_column.size() - 1, parsed_value, settings);
else else
nested_serialization->serializeTextRaw(nested_column, nested_column.size() - 1, parsed_value, settings); nested_serialization->serializeTextRaw(nested_column, nested_column.size() - 1, parsed_value, settings);
throw DB::ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while parsing \"{}{}\" as Nullable" throw DB::Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while parsing \"{}{}\" as Nullable"
" at position {}: got \"{}\", which was deserialized as \"{}\". " " at position {}: got \"{}\", which was deserialized as \"{}\". "
"It seems that input data is ill-formatted.", "It seems that input data is ill-formatted.",
std::string(pos, buf.buffer().end()), std::string(pos, buf.buffer().end()),
@ -452,7 +452,7 @@ ReturnType SerializationNullable::deserializeTextQuotedImpl(IColumn & column, Re
/// It can happen only if there is an unquoted string instead of a number. /// It can happen only if there is an unquoted string instead of a number.
/// We also should delete incorrectly deserialized value from nested column. /// We also should delete incorrectly deserialized value from nested column.
nested_column.popBack(1); nested_column.popBack(1);
throw DB::ParsingException( throw DB::Exception(
ErrorCodes::CANNOT_READ_ALL_DATA, ErrorCodes::CANNOT_READ_ALL_DATA,
"Error while parsing Nullable: got an unquoted string {} instead of a number", "Error while parsing Nullable: got an unquoted string {} instead of a number",
String(buf.position(), std::min(10ul, buf.available()))); String(buf.position(), std::min(10ul, buf.available())));
@ -589,12 +589,12 @@ ReturnType SerializationNullable::deserializeTextCSVImpl(IColumn & column, ReadB
if (null_representation.find(settings.csv.delimiter) != std::string::npos || null_representation.find('\r') != std::string::npos if (null_representation.find(settings.csv.delimiter) != std::string::npos || null_representation.find('\r') != std::string::npos
|| null_representation.find('\n') != std::string::npos) || null_representation.find('\n') != std::string::npos)
throw DB::ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "CSV custom null representation containing " throw DB::Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "CSV custom null representation containing "
"format_csv_delimiter, '\\r' or '\\n' may not work correctly for large input."); "format_csv_delimiter, '\\r' or '\\n' may not work correctly for large input.");
WriteBufferFromOwnString parsed_value; WriteBufferFromOwnString parsed_value;
nested_serialization->serializeTextCSV(nested_column, nested_column.size() - 1, parsed_value, settings); nested_serialization->serializeTextCSV(nested_column, nested_column.size() - 1, parsed_value, settings);
throw DB::ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while parsing \"{}{}\" as Nullable" throw DB::Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while parsing \"{}{}\" as Nullable"
" at position {}: got \"{}\", which was deserialized as \"{}\". " " at position {}: got \"{}\", which was deserialized as \"{}\". "
"It seems that input data is ill-formatted.", "It seems that input data is ill-formatted.",
std::string(pos, buf.buffer().end()), std::string(pos, buf.buffer().end()),

2
src/Formats/JSONUtils.cpp
View File

@ -43,7 +43,7 @@ namespace JSONUtils
{ {
const auto current_object_size = memory.size() + static_cast<size_t>(pos - in.position()); const auto current_object_size = memory.size() + static_cast<size_t>(pos - in.position());
if (min_bytes != 0 && current_object_size > 10 * min_bytes) if (min_bytes != 0 && current_object_size > 10 * min_bytes)
throw ParsingException(ErrorCodes::INCORRECT_DATA, throw Exception(ErrorCodes::INCORRECT_DATA,
"Size of JSON object at position {} is extremely large. Expected not greater than {} bytes, but current is {} bytes per row. " "Size of JSON object at position {} is extremely large. Expected not greater than {} bytes, but current is {} bytes per row. "
"Increase the value setting 'min_chunk_bytes_for_parallel_parsing' or check your data manually, " "Increase the value setting 'min_chunk_bytes_for_parallel_parsing' or check your data manually, "
"most likely JSON is malformed", in.count(), min_bytes, current_object_size); "most likely JSON is malformed", in.count(), min_bytes, current_object_size);

2
src/Formats/NativeReader.cpp
View File

@ -120,7 +120,7 @@ Block NativeReader::read()
if (istr.eof()) if (istr.eof())
{ {
if (use_index) if (use_index)
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Input doesn't contain all data for index."); throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Input doesn't contain all data for index.");
return res; return res;
} }

18
src/IO/ReadHelpers.cpp
View File

@ -89,7 +89,7 @@ void NO_INLINE throwAtAssertionFailed(const char * s, ReadBuffer & buf)
else else
out << " before: " << quote << String(buf.position(), std::min(SHOW_CHARS_ON_SYNTAX_ERROR, buf.buffer().end() - buf.position())); out << " before: " << quote << String(buf.position(), std::min(SHOW_CHARS_ON_SYNTAX_ERROR, buf.buffer().end() - buf.position()));
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "Cannot parse input: expected {}", out.str()); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "Cannot parse input: expected {}", out.str());
} }
@ -562,7 +562,7 @@ static ReturnType readAnyQuotedStringInto(Vector & s, ReadBuffer & buf)
if (buf.eof() || *buf.position() != quote) if (buf.eof() || *buf.position() != quote)
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_QUOTED_STRING, throw Exception(ErrorCodes::CANNOT_PARSE_QUOTED_STRING,
"Cannot parse quoted string: expected opening quote '{}', got '{}'", "Cannot parse quoted string: expected opening quote '{}', got '{}'",
std::string{quote}, buf.eof() ? "EOF" : std::string{*buf.position()}); std::string{quote}, buf.eof() ? "EOF" : std::string{*buf.position()});
else else
@ -608,7 +608,7 @@ static ReturnType readAnyQuotedStringInto(Vector & s, ReadBuffer & buf)
} }
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_QUOTED_STRING, "Cannot parse quoted string: expected closing quote"); throw Exception(ErrorCodes::CANNOT_PARSE_QUOTED_STRING, "Cannot parse quoted string: expected closing quote");
else else
return ReturnType(false); return ReturnType(false);
} }
@ -958,7 +958,7 @@ ReturnType readJSONStringInto(Vector & s, ReadBuffer & buf)
auto error = [](FormatStringHelper<> message [[maybe_unused]], int code [[maybe_unused]]) auto error = [](FormatStringHelper<> message [[maybe_unused]], int code [[maybe_unused]])
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(code, std::move(message)); throw Exception(code, std::move(message));
return ReturnType(false); return ReturnType(false);
}; };
@ -1009,7 +1009,7 @@ ReturnType readJSONObjectOrArrayPossiblyInvalid(Vector & s, ReadBuffer & buf)
auto error = [](FormatStringHelper<> message [[maybe_unused]], int code [[maybe_unused]]) auto error = [](FormatStringHelper<> message [[maybe_unused]], int code [[maybe_unused]])
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(code, std::move(message)); throw Exception(code, std::move(message));
return ReturnType(false); return ReturnType(false);
}; };
@ -1185,7 +1185,7 @@ ReturnType readDateTimeTextFallback(time_t & datetime, ReadBuffer & buf, const D
else else
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime"); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime");
else else
return false; return false;
} }
@ -1212,7 +1212,7 @@ ReturnType readDateTimeTextFallback(time_t & datetime, ReadBuffer & buf, const D
s_pos[size] = 0; s_pos[size] = 0;
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime {}", s); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime {}", s);
else else
return false; return false;
} }
@ -1235,7 +1235,7 @@ ReturnType readDateTimeTextFallback(time_t & datetime, ReadBuffer & buf, const D
s_pos[size] = 0; s_pos[size] = 0;
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse time component of DateTime {}", s); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse time component of DateTime {}", s);
else else
return false; return false;
} }
@ -1266,7 +1266,7 @@ ReturnType readDateTimeTextFallback(time_t & datetime, ReadBuffer & buf, const D
if (too_short && negative_multiplier != -1) if (too_short && negative_multiplier != -1)
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime"); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime");
else else
return false; return false;
} }

32
src/IO/ReadHelpers.h
View File

@ -296,7 +296,7 @@ inline void readBoolTextWord(bool & x, ReadBuffer & buf, bool support_upper_case
[[fallthrough]]; [[fallthrough]];
} }
default: default:
throw ParsingException(ErrorCodes::CANNOT_PARSE_BOOL, "Unexpected Bool value"); throw Exception(ErrorCodes::CANNOT_PARSE_BOOL, "Unexpected Bool value");
} }
} }
@ -340,7 +340,7 @@ ReturnType readIntTextImpl(T & x, ReadBuffer & buf)
if (has_sign) if (has_sign)
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER,
"Cannot parse number with multiple sign (+/-) characters"); "Cannot parse number with multiple sign (+/-) characters");
else else
return ReturnType(false); return ReturnType(false);
@ -357,7 +357,7 @@ ReturnType readIntTextImpl(T & x, ReadBuffer & buf)
if (has_sign) if (has_sign)
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER,
"Cannot parse number with multiple sign (+/-) characters"); "Cannot parse number with multiple sign (+/-) characters");
else else
return ReturnType(false); return ReturnType(false);
@ -368,7 +368,7 @@ ReturnType readIntTextImpl(T & x, ReadBuffer & buf)
else else
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Unsigned type must not contain '-' symbol"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Unsigned type must not contain '-' symbol");
else else
return ReturnType(false); return ReturnType(false);
} }
@ -430,7 +430,7 @@ end:
if (has_sign && !has_number) if (has_sign && !has_number)
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER,
"Cannot parse number with a sign character but without any numeric character"); "Cannot parse number with a sign character but without any numeric character");
else else
return ReturnType(false); return ReturnType(false);
@ -837,7 +837,7 @@ inline ReturnType readUUIDTextImpl(UUID & uuid, ReadBuffer & buf)
if constexpr (throw_exception) if constexpr (throw_exception)
{ {
throw ParsingException(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse uuid {}", s); throw Exception(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse uuid {}", s);
} }
else else
{ {
@ -855,7 +855,7 @@ inline ReturnType readUUIDTextImpl(UUID & uuid, ReadBuffer & buf)
if constexpr (throw_exception) if constexpr (throw_exception)
{ {
throw ParsingException(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse uuid {}", s); throw Exception(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse uuid {}", s);
} }
else else
{ {
@ -881,7 +881,7 @@ inline ReturnType readIPv4TextImpl(IPv4 & ip, ReadBuffer & buf)
return ReturnType(true); return ReturnType(true);
if constexpr (std::is_same_v<ReturnType, void>) if constexpr (std::is_same_v<ReturnType, void>)
throw ParsingException(ErrorCodes::CANNOT_PARSE_IPV4, "Cannot parse IPv4 {}", std::string_view(buf.position(), buf.available())); throw Exception(ErrorCodes::CANNOT_PARSE_IPV4, "Cannot parse IPv4 {}", std::string_view(buf.position(), buf.available()));
else else
return ReturnType(false); return ReturnType(false);
} }
@ -903,7 +903,7 @@ inline ReturnType readIPv6TextImpl(IPv6 & ip, ReadBuffer & buf)
return ReturnType(true); return ReturnType(true);
if constexpr (std::is_same_v<ReturnType, void>) if constexpr (std::is_same_v<ReturnType, void>)
throw ParsingException(ErrorCodes::CANNOT_PARSE_IPV6, "Cannot parse IPv6 {}", std::string_view(buf.position(), buf.available())); throw Exception(ErrorCodes::CANNOT_PARSE_IPV6, "Cannot parse IPv6 {}", std::string_view(buf.position(), buf.available()));
else else
return ReturnType(false); return ReturnType(false);
} }
@ -944,7 +944,7 @@ inline ReturnType readDateTimeTextImpl(time_t & datetime, ReadBuffer & buf, cons
if (!buf.eof() && !isNumericASCII(*buf.position())) if (!buf.eof() && !isNumericASCII(*buf.position()))
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse datetime"); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse datetime");
else else
return false; return false;
} }
@ -1017,7 +1017,7 @@ inline ReturnType readDateTimeTextImpl(DateTime64 & datetime64, UInt32 scale, Re
{ {
readDateTimeTextImpl<ReturnType, true>(whole, buf, date_lut); readDateTimeTextImpl<ReturnType, true>(whole, buf, date_lut);
} }
catch (const DB::ParsingException &) catch (const DB::Exception &)
{ {
if (buf.eof() || *buf.position() != '.') if (buf.eof() || *buf.position() != '.')
throw; throw;
@ -1125,7 +1125,7 @@ inline void readDateTimeText(LocalDateTime & datetime, ReadBuffer & buf)
if (10 != size) if (10 != size)
{ {
s[size] = 0; s[size] = 0;
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime {}", s); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime {}", s);
} }
datetime.year((s[0] - '0') * 1000 + (s[1] - '0') * 100 + (s[2] - '0') * 10 + (s[3] - '0')); datetime.year((s[0] - '0') * 1000 + (s[1] - '0') * 100 + (s[2] - '0') * 10 + (s[3] - '0'));
@ -1141,7 +1141,7 @@ inline void readDateTimeText(LocalDateTime & datetime, ReadBuffer & buf)
if (8 != size) if (8 != size)
{ {
s[size] = 0; s[size] = 0;
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse time component of DateTime {}", s); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse time component of DateTime {}", s);
} }
datetime.hour((s[0] - '0') * 10 + (s[1] - '0')); datetime.hour((s[0] - '0') * 10 + (s[1] - '0'));
@ -1174,7 +1174,7 @@ inline ReturnType readTimeTextImpl(time_t & time, ReadBuffer & buf)
s[size] = 0; s[size] = 0;
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime {}", s); throw Exception(ErrorCodes::CANNOT_PARSE_DATETIME, "Cannot parse DateTime {}", s);
else else
return false; return false;
} }
@ -1482,7 +1482,7 @@ void readQuoted(std::vector<T> & x, ReadBuffer & buf)
if (*buf.position() == ',') if (*buf.position() == ',')
++buf.position(); ++buf.position();
else else
throw ParsingException(ErrorCodes::CANNOT_READ_ARRAY_FROM_TEXT, "Cannot read array from text"); throw Exception(ErrorCodes::CANNOT_READ_ARRAY_FROM_TEXT, "Cannot read array from text");
} }
first = false; first = false;
@ -1505,7 +1505,7 @@ void readDoubleQuoted(std::vector<T> & x, ReadBuffer & buf)
if (*buf.position() == ',') if (*buf.position() == ',')
++buf.position(); ++buf.position();
else else
throw ParsingException(ErrorCodes::CANNOT_READ_ARRAY_FROM_TEXT, "Cannot read array from text"); throw Exception(ErrorCodes::CANNOT_READ_ARRAY_FROM_TEXT, "Cannot read array from text");
} }
first = false; first = false;

2
src/IO/parseDateTimeBestEffort.cpp
View File

@ -95,7 +95,7 @@ ReturnType parseDateTimeBestEffortImpl(
FmtArgs && ...fmt_args [[maybe_unused]]) FmtArgs && ...fmt_args [[maybe_unused]])
{ {
if constexpr (std::is_same_v<ReturnType, void>) if constexpr (std::is_same_v<ReturnType, void>)
throw ParsingException(error_code, std::move(fmt_string), std::forward<FmtArgs>(fmt_args)...); throw Exception(error_code, std::move(fmt_string), std::forward<FmtArgs>(fmt_args)...);
else else
return false; return false;
}; };

4
src/IO/readDecimalText.h
View File

@ -121,7 +121,7 @@ inline bool readDigits(ReadBuffer & buf, T & x, uint32_t & digits, int32_t & exp
if (!tryReadIntText(addition_exp, buf)) if (!tryReadIntText(addition_exp, buf))
{ {
if constexpr (_throw_on_error) if constexpr (_throw_on_error)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot parse exponent while reading decimal"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot parse exponent while reading decimal");
else else
return false; return false;
} }
@ -134,7 +134,7 @@ inline bool readDigits(ReadBuffer & buf, T & x, uint32_t & digits, int32_t & exp
if (digits_only) if (digits_only)
{ {
if constexpr (_throw_on_error) if constexpr (_throw_on_error)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Unexpected symbol while reading decimal"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Unexpected symbol while reading decimal");
return false; return false;
} }
stop = true; stop = true;

14
src/IO/readFloatText.h
View File

@ -160,7 +160,7 @@ ReturnType readFloatTextPreciseImpl(T & x, ReadBuffer & buf)
if (unlikely(res.ec != std::errc())) if (unlikely(res.ec != std::errc()))
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException( throw Exception(
ErrorCodes::CANNOT_PARSE_NUMBER, ErrorCodes::CANNOT_PARSE_NUMBER,
"Cannot read floating point value here: {}", "Cannot read floating point value here: {}",
String(initial_position, buf.buffer().end() - initial_position)); String(initial_position, buf.buffer().end() - initial_position));
@ -253,7 +253,7 @@ ReturnType readFloatTextPreciseImpl(T & x, ReadBuffer & buf)
if (unlikely(res.ec != std::errc() || res.ptr - tmp_buf != num_copied_chars)) if (unlikely(res.ec != std::errc() || res.ptr - tmp_buf != num_copied_chars))
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException( throw Exception(
ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value here: {}", String(tmp_buf, num_copied_chars)); ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value here: {}", String(tmp_buf, num_copied_chars));
else else
return ReturnType(false); return ReturnType(false);
@ -342,7 +342,7 @@ ReturnType readFloatTextFastImpl(T & x, ReadBuffer & in)
if (in.eof()) if (in.eof())
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value");
else else
return false; return false;
} }
@ -400,7 +400,7 @@ ReturnType readFloatTextFastImpl(T & x, ReadBuffer & in)
if (in.eof()) if (in.eof())
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: nothing after exponent"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: nothing after exponent");
else else
return false; return false;
} }
@ -438,7 +438,7 @@ ReturnType readFloatTextFastImpl(T & x, ReadBuffer & in)
if (in.eof()) if (in.eof())
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: no digits read"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: no digits read");
else else
return false; return false;
} }
@ -449,14 +449,14 @@ ReturnType readFloatTextFastImpl(T & x, ReadBuffer & in)
if (in.eof()) if (in.eof())
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: nothing after plus sign"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: nothing after plus sign");
else else
return false; return false;
} }
else if (negative) else if (negative)
{ {
if constexpr (throw_exception) if constexpr (throw_exception)
throw ParsingException(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: plus after minus sign"); throw Exception(ErrorCodes::CANNOT_PARSE_NUMBER, "Cannot read floating point value: plus after minus sign");
else else
return false; return false;
} }

18
src/Processors/Formats/IInputFormat.cpp
View File

@ -1,6 +1,7 @@
#include <Processors/Formats/IInputFormat.h> #include <Processors/Formats/IInputFormat.h>
#include <IO/ReadBuffer.h> #include <IO/ReadBuffer.h>
#include <IO/WithFileName.h>
#include <Common/Exception.h>
namespace DB namespace DB
{ {
@ -11,6 +12,21 @@ IInputFormat::IInputFormat(Block header, ReadBuffer * in_)
column_mapping = std::make_shared<ColumnMapping>(); column_mapping = std::make_shared<ColumnMapping>();
} }
Chunk IInputFormat::generate()
{
try
{
return read();
}
catch (Exception & e)
{
auto file_name = getFileNameFromReadBuffer(getReadBuffer());
if (!file_name.empty())
e.addMessage(fmt::format("(in file/uri {})", file_name));
throw;
}
}
void IInputFormat::resetParser() void IInputFormat::resetParser()
{ {
chassert(in); chassert(in);

13
src/Processors/Formats/IInputFormat.h
View File

@ -27,6 +27,11 @@ public:
/// ReadBuffer can be nullptr for random-access formats. /// ReadBuffer can be nullptr for random-access formats.
IInputFormat(Block header, ReadBuffer * in_); IInputFormat(Block header, ReadBuffer * in_);
Chunk generate() override;
/// All data reading from the read buffer must be performed by this method.
virtual Chunk read() = 0;
/** In some usecase (hello Kafka) we need to read a lot of tiny streams in exactly the same format. /** In some usecase (hello Kafka) we need to read a lot of tiny streams in exactly the same format.
* The recreating of parser for each small stream takes too long, so we introduce a method * The recreating of parser for each small stream takes too long, so we introduce a method
* resetParser() which allow to reset the state of parser to continue reading of * resetParser() which allow to reset the state of parser to continue reading of
@ -49,8 +54,9 @@ public:
/// Must be called from ParallelParsingInputFormat before readPrefix /// Must be called from ParallelParsingInputFormat before readPrefix
void setColumnMapping(ColumnMappingPtr column_mapping_) { column_mapping = column_mapping_; } void setColumnMapping(ColumnMappingPtr column_mapping_) { column_mapping = column_mapping_; }
size_t getCurrentUnitNumber() const { return current_unit_number; } /// Set the number of rows that was already read in
void setCurrentUnitNumber(size_t current_unit_number_) { current_unit_number = current_unit_number_; } /// parallel parsing before creating this parser.
virtual void setRowsReadBefore(size_t /*rows*/) {}
void addBuffer(std::unique_ptr<ReadBuffer> buffer) { owned_buffers.emplace_back(std::move(buffer)); } void addBuffer(std::unique_ptr<ReadBuffer> buffer) { owned_buffers.emplace_back(std::move(buffer)); }
@ -72,9 +78,6 @@ protected:
bool need_only_count = false; bool need_only_count = false;
private: private:
/// Number of currently parsed chunk (if parallel parsing is enabled)
size_t current_unit_number = 0;
std::vector<std::unique_ptr<ReadBuffer>> owned_buffers; std::vector<std::unique_ptr<ReadBuffer>> owned_buffers;
}; };

37
src/Processors/Formats/IRowInputFormat.cpp
View File

@ -83,7 +83,7 @@ void IRowInputFormat::logError()
errors_logger->logError(InputFormatErrorsLogger::ErrorEntry{now_time, total_rows, diagnostic, raw_data}); errors_logger->logError(InputFormatErrorsLogger::ErrorEntry{now_time, total_rows, diagnostic, raw_data});
} }
Chunk IRowInputFormat::generate() Chunk IRowInputFormat::read()
{ {
if (total_rows == 0) if (total_rows == 0)
{ {
@ -93,10 +93,6 @@ Chunk IRowInputFormat::generate()
} }
catch (Exception & e) catch (Exception & e)
{ {
auto file_name = getFileNameFromReadBuffer(getReadBuffer());
if (!file_name.empty())
e.addMessage(fmt::format("(in file/uri {})", file_name));
e.addMessage("(while reading header)"); e.addMessage("(while reading header)");
throw; throw;
} }
@ -132,8 +128,6 @@ Chunk IRowInputFormat::generate()
{ {
try try
{ {
++total_rows;
info.read_columns.clear(); info.read_columns.clear();
continue_reading = readRow(columns, info); continue_reading = readRow(columns, info);
@ -148,6 +142,8 @@ Chunk IRowInputFormat::generate()
} }
} }
++total_rows;
/// Some formats may read row AND say the read is finished. /// Some formats may read row AND say the read is finished.
/// For such a case, get the number or rows from first column. /// For such a case, get the number or rows from first column.
if (!columns.empty()) if (!columns.empty())
@ -162,6 +158,8 @@ Chunk IRowInputFormat::generate()
} }
catch (Exception & e) catch (Exception & e)
{ {
++total_rows;
/// Logic for possible skipping of errors. /// Logic for possible skipping of errors.
if (!isParseError(e.code())) if (!isParseError(e.code()))
@ -204,27 +202,6 @@ Chunk IRowInputFormat::generate()
} }
} }
} }
catch (ParsingException & e)
{
String verbose_diagnostic;
try
{
verbose_diagnostic = getDiagnosticInfo();
}
catch (const Exception & exception)
{
verbose_diagnostic = "Cannot get verbose diagnostic: " + exception.message();
}
catch (...) // NOLINT(bugprone-empty-catch)
{
/// Error while trying to obtain verbose diagnostic. Ok to ignore.
}
e.setFileName(getFileNameFromReadBuffer(getReadBuffer()));
e.setLineNumber(static_cast<int>(total_rows));
e.addMessage(verbose_diagnostic);
throw;
}
catch (Exception & e) catch (Exception & e)
{ {
if (!isParseError(e.code())) if (!isParseError(e.code()))
@ -244,10 +221,6 @@ Chunk IRowInputFormat::generate()
/// Error while trying to obtain verbose diagnostic. Ok to ignore. /// Error while trying to obtain verbose diagnostic. Ok to ignore.
} }
auto file_name = getFileNameFromReadBuffer(getReadBuffer());
if (!file_name.empty())
e.addMessage(fmt::format("(in file/uri {})", file_name));
e.addMessage(fmt::format("(at row {})\n", total_rows)); e.addMessage(fmt::format("(at row {})\n", total_rows));
e.addMessage(verbose_diagnostic); e.addMessage(verbose_diagnostic);
throw; throw;

6
src/Processors/Formats/IRowInputFormat.h
View File

@ -42,7 +42,7 @@ public:
IRowInputFormat(Block header, ReadBuffer & in_, Params params_); IRowInputFormat(Block header, ReadBuffer & in_, Params params_);
Chunk generate() override; Chunk read() override;
void resetParser() override; void resetParser() override;
@ -79,10 +79,12 @@ protected:
const BlockMissingValues & getMissingValues() const override { return block_missing_values; } const BlockMissingValues & getMissingValues() const override { return block_missing_values; }
size_t getTotalRows() const { return total_rows; } size_t getRowNum() const { return total_rows; }
size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; } size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; }
void setRowsReadBefore(size_t rows) override { total_rows = rows; }
Serializations serializations; Serializations serializations;
private: private:

10
src/Processors/Formats/Impl/ArrowBlockInputFormat.cpp
View File

@ -28,7 +28,7 @@ ArrowBlockInputFormat::ArrowBlockInputFormat(ReadBuffer & in_, const Block & hea
{ {
} }
Chunk ArrowBlockInputFormat::generate() Chunk ArrowBlockInputFormat::read()
{ {
Chunk res; Chunk res;
block_missing_values.clear(); block_missing_values.clear();
@ -64,7 +64,7 @@ Chunk ArrowBlockInputFormat::generate()
{ {
auto rows = file_reader->RecordBatchCountRows(record_batch_current++); auto rows = file_reader->RecordBatchCountRows(record_batch_current++);
if (!rows.ok()) if (!rows.ok())
throw ParsingException( throw Exception(
ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading batch of Arrow data: {}", rows.status().ToString()); ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading batch of Arrow data: {}", rows.status().ToString());
return getChunkForCount(*rows); return getChunkForCount(*rows);
} }
@ -73,12 +73,12 @@ Chunk ArrowBlockInputFormat::generate()
} }
if (!batch_result.ok()) if (!batch_result.ok())
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA,
"Error while reading batch of Arrow data: {}", batch_result.status().ToString()); "Error while reading batch of Arrow data: {}", batch_result.status().ToString());
auto table_result = arrow::Table::FromRecordBatches({*batch_result}); auto table_result = arrow::Table::FromRecordBatches({*batch_result});
if (!table_result.ok()) if (!table_result.ok())
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA,
"Error while reading batch of Arrow data: {}", table_result.status().ToString()); "Error while reading batch of Arrow data: {}", table_result.status().ToString());
++record_batch_current; ++record_batch_current;
@ -213,7 +213,7 @@ std::optional<size_t> ArrowSchemaReader::readNumberOrRows()
auto rows = file_reader->CountRows(); auto rows = file_reader->CountRows();
if (!rows.ok()) if (!rows.ok())
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading batch of Arrow data: {}", rows.status().ToString()); throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading batch of Arrow data: {}", rows.status().ToString());
return *rows; return *rows;
} }

2
src/Processors/Formats/Impl/ArrowBlockInputFormat.h
View File

@ -30,7 +30,7 @@ public:
size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; } size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; }
private: private:
Chunk generate() override; Chunk read() override;
void onCancel() override void onCancel() override
{ {

6
src/Processors/Formats/Impl/AvroRowInputFormat.cpp
View File

@ -186,7 +186,7 @@ static AvroDeserializer::DeserializeFn createDecimalDeserializeFn(const avro::No
tmp = decoder.decodeBytes(); tmp = decoder.decodeBytes();
if (tmp.size() > field_type_size || tmp.empty()) if (tmp.size() > field_type_size || tmp.empty())
throw ParsingException( throw Exception(
ErrorCodes::CANNOT_PARSE_UUID, ErrorCodes::CANNOT_PARSE_UUID,
"Cannot parse type {}, expected non-empty binary data with size equal to or less than {}, got {}", "Cannot parse type {}, expected non-empty binary data with size equal to or less than {}, got {}",
target_type->getName(), target_type->getName(),
@ -274,7 +274,7 @@ AvroDeserializer::DeserializeFn AvroDeserializer::createDeserializeFn(const avro
{ {
decoder.decodeString(tmp); decoder.decodeString(tmp);
if (tmp.length() != 36) if (tmp.length() != 36)
throw ParsingException(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse uuid {}", tmp); throw Exception(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse uuid {}", tmp);
const UUID uuid = parseUUID({reinterpret_cast<const UInt8 *>(tmp.data()), tmp.length()}); const UUID uuid = parseUUID({reinterpret_cast<const UInt8 *>(tmp.data()), tmp.length()});
assert_cast<DataTypeUUID::ColumnType &>(column).insertValue(uuid); assert_cast<DataTypeUUID::ColumnType &>(column).insertValue(uuid);
@ -530,7 +530,7 @@ AvroDeserializer::DeserializeFn AvroDeserializer::createDeserializeFn(const avro
{ {
decoder.decodeFixed(fixed_size, tmp); decoder.decodeFixed(fixed_size, tmp);
if (tmp.size() != 36) if (tmp.size() != 36)
throw ParsingException(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse UUID from type Fixed, because it's size ({}) is not equal to the size of UUID (36)", fixed_size); throw Exception(ErrorCodes::CANNOT_PARSE_UUID, "Cannot parse UUID from type Fixed, because it's size ({}) is not equal to the size of UUID (36)", fixed_size);
const UUID uuid = parseUUID({reinterpret_cast<const UInt8 *>(tmp.data()), tmp.size()}); const UUID uuid = parseUUID({reinterpret_cast<const UInt8 *>(tmp.data()), tmp.size()});
assert_cast<DataTypeUUID::ColumnType &>(column).insertValue(uuid); assert_cast<DataTypeUUID::ColumnType &>(column).insertValue(uuid);

6
src/Processors/Formats/Impl/BSONEachRowRowInputFormat.cpp
View File

@ -1031,17 +1031,17 @@ fileSegmentationEngineBSONEachRow(ReadBuffer & in, DB::Memory<> & memory, size_t
readBinaryLittleEndian(document_size, in); readBinaryLittleEndian(document_size, in);
if (document_size < sizeof(document_size)) if (document_size < sizeof(document_size))
throw ParsingException(ErrorCodes::INCORRECT_DATA, "Size of BSON document is invalid"); throw Exception(ErrorCodes::INCORRECT_DATA, "Size of BSON document is invalid");
if (min_bytes != 0 && document_size > 10 * min_bytes) if (min_bytes != 0 && document_size > 10 * min_bytes)
throw ParsingException( throw Exception(
ErrorCodes::INCORRECT_DATA, ErrorCodes::INCORRECT_DATA,
"Size of BSON document is extremely large. Expected not greater than {} bytes, but current is {} bytes per row. Increase " "Size of BSON document is extremely large. Expected not greater than {} bytes, but current is {} bytes per row. Increase "
"the value setting 'min_chunk_bytes_for_parallel_parsing' or check your data manually, most likely BSON is malformed", "the value setting 'min_chunk_bytes_for_parallel_parsing' or check your data manually, most likely BSON is malformed",
min_bytes, document_size); min_bytes, document_size);
if (document_size < sizeof(document_size)) if (document_size < sizeof(document_size))
throw ParsingException(ErrorCodes::INCORRECT_DATA, "Size of BSON document is invalid"); throw Exception(ErrorCodes::INCORRECT_DATA, "Size of BSON document is invalid");
size_t old_size = memory.size(); size_t old_size = memory.size();
memory.resize(old_size + document_size); memory.resize(old_size + document_size);

3
src/Processors/Formats/Impl/BSONEachRowRowInputFormat.h
View File

@ -57,9 +57,6 @@ public:
void resetParser() override; void resetParser() override;
private: private:
void readPrefix() override { }
void readSuffix() override { }
bool readRow(MutableColumns & columns, RowReadExtension & ext) override; bool readRow(MutableColumns & columns, RowReadExtension & ext) override;
bool allowSyncAfterError() const override { return true; } bool allowSyncAfterError() const override { return true; }
void syncAfterError() override; void syncAfterError() override;

2
src/Processors/Formats/Impl/DWARFBlockInputFormat.cpp
View File

@ -888,7 +888,7 @@ void DWARFBlockInputFormat::parseRanges(
} }
} }
Chunk DWARFBlockInputFormat::generate() Chunk DWARFBlockInputFormat::read()
{ {
initializeIfNeeded(); initializeIfNeeded();

2
src/Processors/Formats/Impl/DWARFBlockInputFormat.h
View File

@ -30,7 +30,7 @@ public:
size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; } size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; }
protected: protected:
Chunk generate() override; Chunk read() override;
void onCancel() override void onCancel() override
{ {

2
src/Processors/Formats/Impl/JSONColumnsBlockInputFormatBase.cpp
View File

@ -109,7 +109,7 @@ void JSONColumnsBlockInputFormatBase::setReadBuffer(ReadBuffer & in_)
IInputFormat::setReadBuffer(in_); IInputFormat::setReadBuffer(in_);
} }
Chunk JSONColumnsBlockInputFormatBase::generate() Chunk JSONColumnsBlockInputFormatBase::read()
{ {
MutableColumns columns = getPort().getHeader().cloneEmptyColumns(); MutableColumns columns = getPort().getHeader().cloneEmptyColumns();
block_missing_values.clear(); block_missing_values.clear();

2
src/Processors/Formats/Impl/JSONColumnsBlockInputFormatBase.h
View File

@ -56,7 +56,7 @@ public:
size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; } size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; }
protected: protected:
Chunk generate() override; Chunk read() override;
size_t readColumn(IColumn & column, const DataTypePtr & type, const SerializationPtr & serialization, const String & column_name); size_t readColumn(IColumn & column, const DataTypePtr & type, const SerializationPtr & serialization, const String & column_name);

6
src/Processors/Formats/Impl/JSONEachRowRowInputFormat.cpp
View File

@ -142,7 +142,7 @@ inline bool JSONEachRowRowInputFormat::advanceToNextKey(size_t key_index)
skipWhitespaceIfAny(*in); skipWhitespaceIfAny(*in);
if (in->eof()) if (in->eof())
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected end of stream while parsing JSONEachRow format"); throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected end of stream while parsing JSONEachRow format");
else if (*in->position() == '}') else if (*in->position() == '}')
{ {
++in->position(); ++in->position();
@ -205,7 +205,7 @@ bool JSONEachRowRowInputFormat::readRow(MutableColumns & columns, RowReadExtensi
return false; return false;
skipWhitespaceIfAny(*in); skipWhitespaceIfAny(*in);
bool is_first_row = getCurrentUnitNumber() == 0 && getTotalRows() == 1; bool is_first_row = getRowNum() == 0;
if (checkEndOfData(is_first_row)) if (checkEndOfData(is_first_row))
return false; return false;
@ -308,7 +308,7 @@ size_t JSONEachRowRowInputFormat::countRows(size_t max_block_size)
return 0; return 0;
size_t num_rows = 0; size_t num_rows = 0;
bool is_first_row = getCurrentUnitNumber() == 0 && getTotalRows() == 0; bool is_first_row = getRowNum() == 0;
skipWhitespaceIfAny(*in); skipWhitespaceIfAny(*in);
while (num_rows < max_block_size && !checkEndOfData(is_first_row)) while (num_rows < max_block_size && !checkEndOfData(is_first_row))
{ {

2
src/Processors/Formats/Impl/NativeFormat.cpp
View File

@ -35,7 +35,7 @@ public:
reader->resetParser(); reader->resetParser();
} }
Chunk generate() override Chunk read() override
{ {
block_missing_values.clear(); block_missing_values.clear();
size_t block_start = getDataOffsetMaybeCompressed(*in); size_t block_start = getDataOffsetMaybeCompressed(*in);

2
src/Processors/Formats/Impl/NativeORCBlockInputFormat.cpp
View File

@ -905,7 +905,7 @@ bool NativeORCBlockInputFormat::prepareStripeReader()
return true; return true;
} }
Chunk NativeORCBlockInputFormat::generate() Chunk NativeORCBlockInputFormat::read()
{ {
block_missing_values.clear(); block_missing_values.clear();

2
src/Processors/Formats/Impl/NativeORCBlockInputFormat.h
View File

@ -62,7 +62,7 @@ public:
size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; } size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; }
protected: protected:
Chunk generate() override; Chunk read() override;
void onCancel() override { is_stopped = 1; } void onCancel() override { is_stopped = 1; }

6
src/Processors/Formats/Impl/ORCBlockInputFormat.cpp
View File

@ -27,7 +27,7 @@ ORCBlockInputFormat::ORCBlockInputFormat(ReadBuffer & in_, Block header_, const
{ {
} }
Chunk ORCBlockInputFormat::generate() Chunk ORCBlockInputFormat::read()
{ {
block_missing_values.clear(); block_missing_values.clear();
@ -48,7 +48,7 @@ Chunk ORCBlockInputFormat::generate()
auto batch_result = file_reader->ReadStripe(stripe_current, include_indices); auto batch_result = file_reader->ReadStripe(stripe_current, include_indices);
if (!batch_result.ok()) if (!batch_result.ok())
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Failed to create batch reader: {}", batch_result.status().ToString()); throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Failed to create batch reader: {}", batch_result.status().ToString());
auto batch = batch_result.ValueOrDie(); auto batch = batch_result.ValueOrDie();
if (!batch) if (!batch)
@ -56,7 +56,7 @@ Chunk ORCBlockInputFormat::generate()
auto table_result = arrow::Table::FromRecordBatches({batch}); auto table_result = arrow::Table::FromRecordBatches({batch});
if (!table_result.ok()) if (!table_result.ok())
throw ParsingException( throw Exception(
ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading batch of ORC data: {}", table_result.status().ToString()); ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading batch of ORC data: {}", table_result.status().ToString());
/// We should extract the number of rows directly from the stripe, because in case when /// We should extract the number of rows directly from the stripe, because in case when

2
src/Processors/Formats/Impl/ORCBlockInputFormat.h
View File

@ -32,7 +32,7 @@ public:
size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; } size_t getApproxBytesReadForChunk() const override { return approx_bytes_read_for_chunk; }
protected: protected:
Chunk generate() override; Chunk read() override;
void onCancel() override void onCancel() override
{ {

2
src/Processors/Formats/Impl/OneFormat.cpp
View File

@ -23,7 +23,7 @@ OneInputFormat::OneInputFormat(const Block & header, ReadBuffer & in_) : IInputF
header.getByPosition(0).type->getName()); header.getByPosition(0).type->getName());
} }
Chunk OneInputFormat::generate() Chunk OneInputFormat::read()
{ {
if (done) if (done)
return {}; return {};

2
src/Processors/Formats/Impl/OneFormat.h
View File

@ -14,7 +14,7 @@ public:
String getName() const override { return "One"; } String getName() const override { return "One"; }
protected: protected:
Chunk generate() override; Chunk read() override;
private: private:
bool done = false; bool done = false;

22
src/Processors/Formats/Impl/ParallelParsingInputFormat.cpp
View File

@ -61,7 +61,7 @@ void ParallelParsingInputFormat::segmentatorThreadFunction(ThreadGroupPtr thread
} }
catch (...) catch (...)
{ {
onBackgroundException(successfully_read_rows_count); onBackgroundException();
} }
} }
@ -90,7 +90,7 @@ void ParallelParsingInputFormat::parserThreadFunction(ThreadGroupPtr thread_grou
ReadBuffer read_buffer(unit.segment.data(), unit.segment.size(), 0); ReadBuffer read_buffer(unit.segment.data(), unit.segment.size(), 0);
InputFormatPtr input_format = internal_parser_creator(read_buffer); InputFormatPtr input_format = internal_parser_creator(read_buffer);
input_format->setCurrentUnitNumber(current_ticket_number); input_format->setRowsReadBefore(unit.offset);
input_format->setErrorsLogger(errors_logger); input_format->setErrorsLogger(errors_logger);
InternalParser parser(input_format); InternalParser parser(input_format);
@ -132,28 +132,16 @@ void ParallelParsingInputFormat::parserThreadFunction(ThreadGroupPtr thread_grou
} }
catch (...) catch (...)
{ {
onBackgroundException(unit.offset); onBackgroundException();
} }
} }
void ParallelParsingInputFormat::onBackgroundException(size_t offset) void ParallelParsingInputFormat::onBackgroundException()
{ {
std::lock_guard lock(mutex); std::lock_guard lock(mutex);
if (!background_exception) if (!background_exception)
{
background_exception = std::current_exception(); background_exception = std::current_exception();
if (ParsingException * e = exception_cast<ParsingException *>(background_exception))
{
/// NOTE: it is not that safe to use line number hack here (may exceed INT_MAX)
if (e->getLineNumber() != -1)
e->setLineNumber(static_cast<int>(e->getLineNumber() + offset));
auto file_name = getFileNameFromReadBuffer(getReadBuffer());
if (!file_name.empty())
e->setFileName(file_name);
}
}
if (is_server) if (is_server)
tryLogCurrentException(__PRETTY_FUNCTION__); tryLogCurrentException(__PRETTY_FUNCTION__);
@ -164,7 +152,7 @@ void ParallelParsingInputFormat::onBackgroundException(size_t offset)
segmentator_condvar.notify_all(); segmentator_condvar.notify_all();
} }
Chunk ParallelParsingInputFormat::generate() Chunk ParallelParsingInputFormat::read()
{ {
/// Delayed launching of segmentator thread /// Delayed launching of segmentator thread
if (unlikely(!parsing_started.exchange(true))) if (unlikely(!parsing_started.exchange(true)))

4
src/Processors/Formats/Impl/ParallelParsingInputFormat.h
View File

@ -135,7 +135,7 @@ public:
private: private:
Chunk generate() override final; Chunk read() override final;
void onCancel() override final void onCancel() override final
{ {
@ -333,7 +333,7 @@ private:
/// threads. This function is used by segmentator and parsed threads. /// threads. This function is used by segmentator and parsed threads.
/// readImpl() is called from the main thread, so the exception handling /// readImpl() is called from the main thread, so the exception handling
/// is different. /// is different.
void onBackgroundException(size_t offset); void onBackgroundException();
}; };
} }

6
src/Processors/Formats/Impl/ParquetBlockInputFormat.cpp
View File

@ -570,7 +570,7 @@ void ParquetBlockInputFormat::decodeOneChunk(size_t row_group_batch_idx, std::un
// We may be able to schedule more work now, but can't call scheduleMoreWorkIfNeeded() right // We may be able to schedule more work now, but can't call scheduleMoreWorkIfNeeded() right
// here because we're running on the same thread pool, so it'll deadlock if thread limit is // here because we're running on the same thread pool, so it'll deadlock if thread limit is
// reached. Wake up generate() instead. // reached. Wake up read() instead.
condvar.notify_all(); condvar.notify_all();
}; };
@ -579,7 +579,7 @@ void ParquetBlockInputFormat::decodeOneChunk(size_t row_group_batch_idx, std::un
auto batch = row_group_batch.record_batch_reader->Next(); auto batch = row_group_batch.record_batch_reader->Next();
if (!batch.ok()) if (!batch.ok())
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading Parquet data: {}", batch.status().ToString()); throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading Parquet data: {}", batch.status().ToString());
if (!*batch) if (!*batch)
{ {
@ -637,7 +637,7 @@ void ParquetBlockInputFormat::scheduleMoreWorkIfNeeded(std::optional<size_t> row
} }
} }
Chunk ParquetBlockInputFormat::generate() Chunk ParquetBlockInputFormat::read()
{ {
initializeIfNeeded(); initializeIfNeeded();

12
src/Processors/Formats/Impl/ParquetBlockInputFormat.h
View File

@ -65,7 +65,7 @@ public:
size_t getApproxBytesReadForChunk() const override { return previous_approx_bytes_read_for_chunk; } size_t getApproxBytesReadForChunk() const override { return previous_approx_bytes_read_for_chunk; }
private: private:
Chunk generate() override; Chunk read() override;
void onCancel() override void onCancel() override
{ {
@ -142,7 +142,7 @@ private:
// reading its data (using RAM). Row group becomes inactive when we finish reading and // reading its data (using RAM). Row group becomes inactive when we finish reading and
// delivering all its blocks and free the RAM. Size of the window is max_decoding_threads. // delivering all its blocks and free the RAM. Size of the window is max_decoding_threads.
// //
// Decoded blocks are placed in `pending_chunks` queue, then picked up by generate(). // Decoded blocks are placed in `pending_chunks` queue, then picked up by read().
// If row group decoding runs too far ahead of delivery (by `max_pending_chunks_per_row_group` // If row group decoding runs too far ahead of delivery (by `max_pending_chunks_per_row_group`
// chunks), we pause the stream for the row group, to avoid using too much memory when decoded // chunks), we pause the stream for the row group, to avoid using too much memory when decoded
// chunks are much bigger than the compressed data. // chunks are much bigger than the compressed data.
@ -150,7 +150,7 @@ private:
// Also: // Also:
// * If preserve_order = true, we deliver chunks strictly in order of increasing row group. // * If preserve_order = true, we deliver chunks strictly in order of increasing row group.
// Decoding may still proceed in later row groups. // Decoding may still proceed in later row groups.
// * If max_decoding_threads <= 1, we run all tasks inline in generate(), without thread pool. // * If max_decoding_threads <= 1, we run all tasks inline in read(), without thread pool.
// Potential improvements: // Potential improvements:
// * Plan all read ranges ahead of time, for the whole file, and do prefetching for them // * Plan all read ranges ahead of time, for the whole file, and do prefetching for them
@ -189,7 +189,7 @@ private:
Status status = Status::NotStarted; Status status = Status::NotStarted;
// Window of chunks that were decoded but not returned from generate(): // Window of chunks that were decoded but not returned from read():
// //
// (delivered) next_chunk_idx // (delivered) next_chunk_idx
// v v v // v v v
@ -215,7 +215,7 @@ private:
std::unique_ptr<ArrowColumnToCHColumn> arrow_column_to_ch_column; std::unique_ptr<ArrowColumnToCHColumn> arrow_column_to_ch_column;
}; };
// Chunk ready to be delivered by generate(). // Chunk ready to be delivered by read().
struct PendingChunk struct PendingChunk
{ {
Chunk chunk; Chunk chunk;
@ -265,7 +265,7 @@ private:
// Done NotStarted // Done NotStarted
std::mutex mutex; std::mutex mutex;
// Wakes up the generate() call, if any. // Wakes up the read() call, if any.
std::condition_variable condvar; std::condition_variable condvar;
std::vector<RowGroupBatchState> row_group_batches; std::vector<RowGroupBatchState> row_group_batches;

2
src/Processors/Formats/Impl/ParquetMetadataInputFormat.cpp
View File

@ -140,7 +140,7 @@ ParquetMetadataInputFormat::ParquetMetadataInputFormat(ReadBuffer & in_, Block h
checkHeader(getPort().getHeader()); checkHeader(getPort().getHeader());
} }
Chunk ParquetMetadataInputFormat::generate() Chunk ParquetMetadataInputFormat::read()
{ {
Chunk res; Chunk res;
if (done) if (done)

2
src/Processors/Formats/Impl/ParquetMetadataInputFormat.h
View File

@ -63,7 +63,7 @@ public:
void resetParser() override; void resetParser() override;
private: private:
Chunk generate() override; Chunk read() override;
void onCancel() override void onCancel() override
{ {

2
src/Processors/Formats/Impl/ProtobufListInputFormat.cpp
View File

@ -61,7 +61,7 @@ bool ProtobufListInputFormat::readRow(MutableColumns & columns, RowReadExtension
size_t ProtobufListInputFormat::countRows(size_t max_block_size) size_t ProtobufListInputFormat::countRows(size_t max_block_size)
{ {
if (getTotalRows() == 0) if (getRowNum() == 0)
reader->startMessage(true); reader->startMessage(true);
if (reader->eof()) if (reader->eof())

4
src/Processors/Formats/Impl/TSKVRowInputFormat.cpp
View File

@ -92,7 +92,7 @@ static bool readName(ReadBuffer & buf, StringRef & ref, String & tmp)
} }
} }
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected end of stream while reading key name from TSKV format"); throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected end of stream while reading key name from TSKV format");
} }
@ -161,7 +161,7 @@ bool TSKVRowInputFormat::readRow(MutableColumns & columns, RowReadExtension & ex
if (in->eof()) if (in->eof())
{ {
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected end of stream after field in TSKV format: {}", name_ref.toString()); throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected end of stream after field in TSKV format: {}", name_ref.toString());
} }
else if (*in->position() == '\t') else if (*in->position() == '\t')
{ {

8
src/Processors/Formats/Impl/TemplateRowInputFormat.cpp
View File

@ -21,7 +21,7 @@ namespace ErrorCodes
[[noreturn]] static void throwUnexpectedEof(size_t row_num) [[noreturn]] static void throwUnexpectedEof(size_t row_num)
{ {
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected EOF while parsing row {}. " throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA, "Unexpected EOF while parsing row {}. "
"Maybe last row has wrong format or input doesn't contain specified suffix before EOF.", "Maybe last row has wrong format or input doesn't contain specified suffix before EOF.",
std::to_string(row_num)); std::to_string(row_num));
} }
@ -121,7 +121,7 @@ bool TemplateRowInputFormat::readRow(MutableColumns & columns, RowReadExtension
updateDiagnosticInfo(); updateDiagnosticInfo();
if (likely(row_num != 1)) if (likely(getRowNum() != 0))
format_reader->skipRowBetweenDelimiter(); format_reader->skipRowBetweenDelimiter();
extra.read_columns.assign(columns.size(), false); extra.read_columns.assign(columns.size(), false);
@ -160,7 +160,7 @@ bool TemplateRowInputFormat::deserializeField(const DataTypePtr & type,
catch (Exception & e) catch (Exception & e)
{ {
if (e.code() == ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF) if (e.code() == ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF)
throwUnexpectedEof(row_num); throwUnexpectedEof(getRowNum());
throw; throw;
} }
} }
@ -198,7 +198,7 @@ bool TemplateRowInputFormat::parseRowAndPrintDiagnosticInfo(MutableColumns & col
out << "\nUsing format string (from format_schema_rows): " << row_format.dump() << "\n"; out << "\nUsing format string (from format_schema_rows): " << row_format.dump() << "\n";
out << "\nTrying to parse next row, because suffix does not match:\n"; out << "\nTrying to parse next row, because suffix does not match:\n";
if (likely(row_num != 1) && !parseDelimiterWithDiagnosticInfo(out, *buf, row_between_delimiter, "delimiter between rows", ignore_spaces)) if (likely(getRowNum() != 0) && !parseDelimiterWithDiagnosticInfo(out, *buf, row_between_delimiter, "delimiter between rows", ignore_spaces))
return false; return false;
for (size_t i = 0; i < row_format.columnsCount(); ++i) for (size_t i = 0; i < row_format.columnsCount(); ++i)

2
src/Processors/Formats/Impl/ValuesBlockInputFormat.cpp
View File

@ -98,7 +98,7 @@ bool ValuesBlockInputFormat::skipToNextRow(ReadBuffer * buf, size_t min_chunk_by
return true; return true;
} }
Chunk ValuesBlockInputFormat::generate() Chunk ValuesBlockInputFormat::read()
{ {
if (total_rows == 0) if (total_rows == 0)
readPrefix(); readPrefix();

2
src/Processors/Formats/Impl/ValuesBlockInputFormat.h
View File

@ -58,7 +58,7 @@ private:
using ConstantExpressionTemplates = std::vector<std::optional<ConstantExpressionTemplate>>; using ConstantExpressionTemplates = std::vector<std::optional<ConstantExpressionTemplate>>;
Chunk generate() override; Chunk read() override;
void readRow(MutableColumns & columns, size_t row_num); void readRow(MutableColumns & columns, size_t row_num);
void readUntilTheEndOfRowAndReTokenize(size_t current_column_idx); void readUntilTheEndOfRowAndReTokenize(size_t current_column_idx);

7
src/Processors/Formats/RowInputFormatWithDiagnosticInfo.cpp
View File

@ -26,8 +26,6 @@ RowInputFormatWithDiagnosticInfo::RowInputFormatWithDiagnosticInfo(const Block &
void RowInputFormatWithDiagnosticInfo::updateDiagnosticInfo() void RowInputFormatWithDiagnosticInfo::updateDiagnosticInfo()
{ {
++row_num;
bytes_read_at_start_of_buffer_on_prev_row = bytes_read_at_start_of_buffer_on_current_row; bytes_read_at_start_of_buffer_on_prev_row = bytes_read_at_start_of_buffer_on_current_row;
bytes_read_at_start_of_buffer_on_current_row = in->count() - in->offset(); bytes_read_at_start_of_buffer_on_current_row = in->count() - in->offset();
@ -73,7 +71,7 @@ std::pair<String, String> RowInputFormatWithDiagnosticInfo::getDiagnosticAndRawD
{ {
in->position() = in->buffer().begin() + offset_of_prev_row; in->position() = in->buffer().begin() + offset_of_prev_row;
out_diag << "\nRow " << (row_num - 1) << ":\n"; out_diag << "\nRow " << getRowNum() - 1 << ":\n";
if (!parseRowAndPrintDiagnosticInfo(columns, out_diag)) if (!parseRowAndPrintDiagnosticInfo(columns, out_diag))
return std::make_pair(out_diag.str(), out_data.str()); return std::make_pair(out_diag.str(), out_data.str());
} }
@ -96,7 +94,7 @@ std::pair<String, String> RowInputFormatWithDiagnosticInfo::getDiagnosticAndRawD
++data; ++data;
} }
out_diag << "\nRow " << row_num << ":\n"; out_diag << "\nRow " << getRowNum() << ":\n";
parseRowAndPrintDiagnosticInfo(columns, out_diag); parseRowAndPrintDiagnosticInfo(columns, out_diag);
out_diag << "\n"; out_diag << "\n";
@ -193,7 +191,6 @@ bool RowInputFormatWithDiagnosticInfo::deserializeFieldAndPrintDiagnosticInfo(co
void RowInputFormatWithDiagnosticInfo::resetParser() void RowInputFormatWithDiagnosticInfo::resetParser()
{ {
IRowInputFormat::resetParser(); IRowInputFormat::resetParser();
row_num = 0;
bytes_read_at_start_of_buffer_on_current_row = 0; bytes_read_at_start_of_buffer_on_current_row = 0;
bytes_read_at_start_of_buffer_on_prev_row = 0; bytes_read_at_start_of_buffer_on_prev_row = 0;
offset_of_current_row = std::numeric_limits<size_t>::max(); offset_of_current_row = std::numeric_limits<size_t>::max();

3
src/Processors/Formats/RowInputFormatWithDiagnosticInfo.h
View File

@ -29,9 +29,6 @@ protected:
virtual void tryDeserializeField(const DataTypePtr & type, IColumn & column, size_t file_column) = 0; virtual void tryDeserializeField(const DataTypePtr & type, IColumn & column, size_t file_column) = 0;
virtual bool isGarbageAfterField(size_t after_input_pos_idx, ReadBuffer::Position pos) = 0; virtual bool isGarbageAfterField(size_t after_input_pos_idx, ReadBuffer::Position pos) = 0;
/// For convenient diagnostics in case of an error.
size_t row_num = 0;
private: private:
/// How many bytes were read, not counting those still in the buffer. /// How many bytes were read, not counting those still in the buffer.
size_t bytes_read_at_start_of_buffer_on_current_row = 0; size_t bytes_read_at_start_of_buffer_on_current_row = 0;

11
src/Processors/Formats/RowInputFormatWithNamesAndTypes.cpp
View File

@ -66,11 +66,6 @@ RowInputFormatWithNamesAndTypes::RowInputFormatWithNamesAndTypes(
void RowInputFormatWithNamesAndTypes::readPrefix() void RowInputFormatWithNamesAndTypes::readPrefix()
{ {
/// This is a bit of abstraction leakage, but we need it in parallel parsing:
/// we check if this InputFormat is working with the "real" beginning of the data.
if (getCurrentUnitNumber() != 0)
return;
/// Search and remove BOM only in textual formats (CSV, TSV etc), not in binary ones (RowBinary*). /// Search and remove BOM only in textual formats (CSV, TSV etc), not in binary ones (RowBinary*).
/// Also, we assume that column name or type cannot contain BOM, so, if format has header, /// Also, we assume that column name or type cannot contain BOM, so, if format has header,
/// then BOM at beginning of stream cannot be confused with name or type of field, and it is safe to skip it. /// then BOM at beginning of stream cannot be confused with name or type of field, and it is safe to skip it.
@ -206,7 +201,7 @@ bool RowInputFormatWithNamesAndTypes::readRow(MutableColumns & columns, RowReadE
updateDiagnosticInfo(); updateDiagnosticInfo();
if (likely(row_num != 1 || getCurrentUnitNumber() != 0 || (getCurrentUnitNumber() == 0 && (with_names || with_types || is_header_detected)))) if (likely(getRowNum() != 0 || with_names || with_types || is_header_detected))
format_reader->skipRowBetweenDelimiter(); format_reader->skipRowBetweenDelimiter();
format_reader->skipRowStartDelimiter(); format_reader->skipRowStartDelimiter();
@ -270,7 +265,7 @@ size_t RowInputFormatWithNamesAndTypes::countRows(size_t max_block_size)
return 0; return 0;
size_t num_rows = 0; size_t num_rows = 0;
bool is_first_row = getTotalRows() == 0 && !with_names && !with_types && !is_header_detected; bool is_first_row = getRowNum() == 0 && !with_names && !with_types && !is_header_detected;
while (!format_reader->checkForSuffix() && num_rows < max_block_size) while (!format_reader->checkForSuffix() && num_rows < max_block_size)
{ {
if (likely(!is_first_row)) if (likely(!is_first_row))
@ -323,7 +318,7 @@ bool RowInputFormatWithNamesAndTypes::parseRowAndPrintDiagnosticInfo(MutableColu
if (!format_reader->tryParseSuffixWithDiagnosticInfo(out)) if (!format_reader->tryParseSuffixWithDiagnosticInfo(out))
return false; return false;
if (likely(row_num != 1) && !format_reader->parseRowBetweenDelimiterWithDiagnosticInfo(out)) if (likely(getRowNum() != 0) && !format_reader->parseRowBetweenDelimiterWithDiagnosticInfo(out))
return false; return false;
if (!format_reader->parseRowStartWithDiagnosticInfo(out)) if (!format_reader->parseRowStartWithDiagnosticInfo(out))

16
src/Server/ProxyV1Handler.cpp
View File

@ -29,38 +29,38 @@ void ProxyV1Handler::run()
// read "PROXY" // read "PROXY"
if (!readWord(5, word, eol) || word != "PROXY" || eol) if (!readWord(5, word, eol) || word != "PROXY" || eol)
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
// read "TCP4" or "TCP6" or "UNKNOWN" // read "TCP4" or "TCP6" or "UNKNOWN"
if (!readWord(7, word, eol)) if (!readWord(7, word, eol))
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
if (word != "TCP4" && word != "TCP6" && word != "UNKNOWN") if (word != "TCP4" && word != "TCP6" && word != "UNKNOWN")
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
if (word == "UNKNOWN" && eol) if (word == "UNKNOWN" && eol)
return; return;
if (eol) if (eol)
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
// read address // read address
if (!readWord(39, word, eol) || eol) if (!readWord(39, word, eol) || eol)
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
stack_data.forwarded_for = std::move(word); stack_data.forwarded_for = std::move(word);
// read address // read address
if (!readWord(39, word, eol) || eol) if (!readWord(39, word, eol) || eol)
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
// read port // read port
if (!readWord(5, word, eol) || eol) if (!readWord(5, word, eol) || eol)
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
// read port and "\r\n" // read port and "\r\n"
if (!readWord(5, word, eol) || !eol) if (!readWord(5, word, eol) || !eol)
throw ParsingException(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation"); throw Exception(ErrorCodes::CANNOT_PARSE_INPUT_ASSERTION_FAILED, "PROXY protocol violation");
if (!stack_data.forwarded_for.empty()) if (!stack_data.forwarded_for.empty())
LOG_TRACE(log, "Forwarded client address from PROXY header: {}", stack_data.forwarded_for); LOG_TRACE(log, "Forwarded client address from PROXY header: {}", stack_data.forwarded_for);

16
tests/integration/test_kafka_bad_messages/test.py
View File

@ -294,7 +294,7 @@ def test_bad_messages_parsing_exception(kafka_cluster, max_retries=20):
]: ]:
print(format_name) print(format_name)
kafka_create_topic(admin_client, f"{format_name}_err") kafka_create_topic(admin_client, f"{format_name}_parsing_err")
instance.query( instance.query(
f""" f"""
@ -305,7 +305,7 @@ def test_bad_messages_parsing_exception(kafka_cluster, max_retries=20):
CREATE TABLE kafka_{format_name} (key UInt64, value UInt64) CREATE TABLE kafka_{format_name} (key UInt64, value UInt64)
ENGINE = Kafka ENGINE = Kafka
SETTINGS kafka_broker_list = 'kafka1:19092', SETTINGS kafka_broker_list = 'kafka1:19092',
kafka_topic_list = '{format_name}_err', kafka_topic_list = '{format_name}_parsing_err',
kafka_group_name = '{format_name}', kafka_group_name = '{format_name}',
kafka_format = '{format_name}', kafka_format = '{format_name}',
kafka_num_consumers = 1; kafka_num_consumers = 1;
@ -316,16 +316,18 @@ def test_bad_messages_parsing_exception(kafka_cluster, max_retries=20):
) )
kafka_produce( kafka_produce(
kafka_cluster, f"{format_name}_err", ["qwertyuiop", "asdfghjkl", "zxcvbnm"] kafka_cluster,
f"{format_name}_parsing_err",
["qwertyuiop", "asdfghjkl", "zxcvbnm"],
) )
expected_result = """avro::Exception: Invalid data file. Magic does not match: : while parsing Kafka message (topic: Avro_err, partition: 0, offset: 0)\\'|1|1|1|default|kafka_Avro expected_result = """avro::Exception: Invalid data file. Magic does not match: : while parsing Kafka message (topic: Avro_parsing_err, partition: 0, offset: 0)\\'|1|1|1|default|kafka_Avro
Cannot parse input: expected \\'{\\' before: \\'qwertyuiop\\': while parsing Kafka message (topic: JSONEachRow_err, partition: 0, offset: 0|1|1|1|default|kafka_JSONEachRow Cannot parse input: expected \\'{\\' before: \\'qwertyuiop\\': (at row 1)\\n: while parsing Kafka message (topic: JSONEachRow_parsing_err, partition:|1|1|1|default|kafka_JSONEachRow
""" """
# filter out stacktrace in exceptions.text[1] because it is hardly stable enough # filter out stacktrace in exceptions.text[1] because it is hardly stable enough
result_system_kafka_consumers = instance.query_with_retry( result_system_kafka_consumers = instance.query_with_retry(
""" """
SELECT substr(exceptions.text[1], 1, 131), length(exceptions.text) > 1 AND length(exceptions.text) < 15, length(exceptions.time) > 1 AND length(exceptions.time) < 15, abs(dateDiff('second', exceptions.time[1], now())) < 40, database, table FROM system.kafka_consumers WHERE table in('kafka_Avro', 'kafka_JSONEachRow') ORDER BY table, assignments.partition_id[1] SELECT substr(exceptions.text[1], 1, 139), length(exceptions.text) > 1 AND length(exceptions.text) < 15, length(exceptions.time) > 1 AND length(exceptions.time) < 15, abs(dateDiff('second', exceptions.time[1], now())) < 40, database, table FROM system.kafka_consumers WHERE table in('kafka_Avro', 'kafka_JSONEachRow') ORDER BY table, assignments.partition_id[1]
""", """,
retry_count=max_retries, retry_count=max_retries,
sleep_time=1, sleep_time=1,
@ -338,7 +340,7 @@ Cannot parse input: expected \\'{\\' before: \\'qwertyuiop\\': while parsing Kaf
"Avro", "Avro",
"JSONEachRow", "JSONEachRow",
]: ]:
kafka_delete_topic(admin_client, f"{format_name}_err") kafka_delete_topic(admin_client, f"{format_name}_parsing_err")
def test_bad_messages_to_mv(kafka_cluster, max_retries=20): def test_bad_messages_to_mv(kafka_cluster, max_retries=20):

40
tests/queries/0_stateless/02889_file_log_save_errors.reference
View File

@ -1,20 +1,20 @@
Cannot parse input: expected \'{\' before: \'Error 0\' Error 0 a.jsonl Cannot parse input: expected \'{\' before: \'Error 0\': (at row 1)\n Error 0 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 1\' Error 1 a.jsonl Cannot parse input: expected \'{\' before: \'Error 1\': (at row 1)\n Error 1 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 2\' Error 2 a.jsonl Cannot parse input: expected \'{\' before: \'Error 2\': (at row 1)\n Error 2 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 3\' Error 3 a.jsonl Cannot parse input: expected \'{\' before: \'Error 3\': (at row 1)\n Error 3 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 4\' Error 4 a.jsonl Cannot parse input: expected \'{\' before: \'Error 4\': (at row 1)\n Error 4 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 5\' Error 5 a.jsonl Cannot parse input: expected \'{\' before: \'Error 5\': (at row 1)\n Error 5 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 6\' Error 6 a.jsonl Cannot parse input: expected \'{\' before: \'Error 6\': (at row 1)\n Error 6 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 7\' Error 7 a.jsonl Cannot parse input: expected \'{\' before: \'Error 7\': (at row 1)\n Error 7 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 8\' Error 8 a.jsonl Cannot parse input: expected \'{\' before: \'Error 8\': (at row 1)\n Error 8 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 9\' Error 9 a.jsonl Cannot parse input: expected \'{\' before: \'Error 9\': (at row 1)\n Error 9 a.jsonl
Cannot parse input: expected \'{\' before: \'Error 10\' Error 10 b.jsonl Cannot parse input: expected \'{\' before: \'Error 10\': (at row 1)\n Error 10 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 11\' Error 11 b.jsonl Cannot parse input: expected \'{\' before: \'Error 11\': (at row 1)\n Error 11 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 12\' Error 12 b.jsonl Cannot parse input: expected \'{\' before: \'Error 12\': (at row 1)\n Error 12 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 13\' Error 13 b.jsonl Cannot parse input: expected \'{\' before: \'Error 13\': (at row 1)\n Error 13 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 14\' Error 14 b.jsonl Cannot parse input: expected \'{\' before: \'Error 14\': (at row 1)\n Error 14 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 15\' Error 15 b.jsonl Cannot parse input: expected \'{\' before: \'Error 15\': (at row 1)\n Error 15 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 16\' Error 16 b.jsonl Cannot parse input: expected \'{\' before: \'Error 16\': (at row 1)\n Error 16 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 17\' Error 17 b.jsonl Cannot parse input: expected \'{\' before: \'Error 17\': (at row 1)\n Error 17 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 18\' Error 18 b.jsonl Cannot parse input: expected \'{\' before: \'Error 18\': (at row 1)\n Error 18 b.jsonl
Cannot parse input: expected \'{\' before: \'Error 19\' Error 19 b.jsonl Cannot parse input: expected \'{\' before: \'Error 19\': (at row 1)\n Error 19 b.jsonl