|
| 1 | +// Licensed to the Apache Software Foundation (ASF) under one |
| 2 | +// or more contributor license agreements. See the NOTICE file |
| 3 | +// distributed with this work for additional information |
| 4 | +// regarding copyright ownership. The ASF licenses this file |
| 5 | +// to you under the Apache License, Version 2.0 (the |
| 6 | +// "License"); you may not use this file except in compliance |
| 7 | +// with the License. You may obtain a copy of the License at |
| 8 | +// |
| 9 | +// http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | +// |
| 11 | +// Unless required by applicable law or agreed to in writing, |
| 12 | +// software distributed under the License is distributed on an |
| 13 | +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| 14 | +// KIND, either express or implied. See the License for the |
| 15 | +// specific language governing permissions and limitations |
| 16 | +// under the License. |
| 17 | + |
| 18 | +//! DataFusion error types |
| 19 | +
|
| 20 | +use std::error; |
| 21 | +use std::fmt::{Display, Formatter}; |
| 22 | +use std::io; |
| 23 | +use std::result; |
| 24 | + |
| 25 | +use arrow::error::ArrowError; |
| 26 | +#[cfg(feature = "avro")] |
| 27 | +use avro_rs::Error as AvroError; |
| 28 | +use parquet::errors::ParquetError; |
| 29 | +#[cfg(feature = "pyarrow")] |
| 30 | +use pyo3::exceptions::PyException; |
| 31 | +#[cfg(feature = "pyarrow")] |
| 32 | +use pyo3::prelude::PyErr; |
| 33 | +use sqlparser::parser::ParserError; |
| 34 | + |
| 35 | +/// Result type for operations that could result in an [DataFusionError] |
| 36 | +pub type Result<T> = result::Result<T, DataFusionError>; |
| 37 | + |
| 38 | +/// Error type for generic operations that could result in DataFusionError::External |
| 39 | +pub type GenericError = Box<dyn error::Error + Send + Sync>; |
| 40 | + |
| 41 | +/// DataFusion error |
| 42 | +#[derive(Debug)] |
| 43 | +pub enum DataFusionError { |
| 44 | + /// Error returned by arrow. |
| 45 | + ArrowError(ArrowError), |
| 46 | + /// Wraps an error from the Parquet crate |
| 47 | + ParquetError(ParquetError), |
| 48 | + /// Wraps an error from the Avro crate |
| 49 | + #[cfg(feature = "avro")] |
| 50 | + AvroError(AvroError), |
| 51 | + /// Error associated to I/O operations and associated traits. |
| 52 | + IoError(io::Error), |
| 53 | + /// Error returned when SQL is syntactically incorrect. |
| 54 | + SQL(ParserError), |
| 55 | + /// Error returned on a branch that we know it is possible |
| 56 | + /// but to which we still have no implementation for. |
| 57 | + /// Often, these errors are tracked in our issue tracker. |
| 58 | + NotImplemented(String), |
| 59 | + /// Error returned as a consequence of an error in DataFusion. |
| 60 | + /// This error should not happen in normal usage of DataFusion. |
| 61 | + // DataFusions has internal invariants that we are unable to ask the compiler to check for us. |
| 62 | + // This error is raised when one of those invariants is not verified during execution. |
| 63 | + Internal(String), |
| 64 | + /// This error happens whenever a plan is not valid. Examples include |
| 65 | + /// impossible casts, schema inference not possible and non-unique column names. |
| 66 | + Plan(String), |
| 67 | + /// Error returned during execution of the query. |
| 68 | + /// Examples include files not found, errors in parsing certain types. |
| 69 | + Execution(String), |
| 70 | + /// This error is thrown when a consumer cannot acquire memory from the Memory Manager |
| 71 | + /// we can just cancel the execution of the partition. |
| 72 | + ResourcesExhausted(String), |
| 73 | + /// Errors originating from outside DataFusion's core codebase. |
| 74 | + /// For example, a custom S3Error from the crate datafusion-objectstore-s3 |
| 75 | + External(GenericError), |
| 76 | +} |
| 77 | + |
| 78 | +impl From<io::Error> for DataFusionError { |
| 79 | + fn from(e: io::Error) -> Self { |
| 80 | + DataFusionError::IoError(e) |
| 81 | + } |
| 82 | +} |
| 83 | + |
| 84 | +impl From<ArrowError> for DataFusionError { |
| 85 | + fn from(e: ArrowError) -> Self { |
| 86 | + DataFusionError::ArrowError(e) |
| 87 | + } |
| 88 | +} |
| 89 | + |
| 90 | +#[cfg(feature = "pyarrow")] |
| 91 | +impl From<DataFusionError> for PyErr { |
| 92 | + fn from(err: DataFusionError) -> PyErr { |
| 93 | + PyException::new_err(err.to_string()) |
| 94 | + } |
| 95 | +} |
| 96 | + |
| 97 | +impl From<DataFusionError> for ArrowError { |
| 98 | + fn from(e: DataFusionError) -> Self { |
| 99 | + match e { |
| 100 | + DataFusionError::ArrowError(e) => e, |
| 101 | + DataFusionError::External(e) => ArrowError::ExternalError(e), |
| 102 | + other => ArrowError::ExternalError(Box::new(other)), |
| 103 | + } |
| 104 | + } |
| 105 | +} |
| 106 | + |
| 107 | +impl From<ParquetError> for DataFusionError { |
| 108 | + fn from(e: ParquetError) -> Self { |
| 109 | + DataFusionError::ParquetError(e) |
| 110 | + } |
| 111 | +} |
| 112 | + |
| 113 | +#[cfg(feature = "avro")] |
| 114 | +impl From<AvroError> for DataFusionError { |
| 115 | + fn from(e: AvroError) -> Self { |
| 116 | + DataFusionError::AvroError(e) |
| 117 | + } |
| 118 | +} |
| 119 | + |
| 120 | +impl From<ParserError> for DataFusionError { |
| 121 | + fn from(e: ParserError) -> Self { |
| 122 | + DataFusionError::SQL(e) |
| 123 | + } |
| 124 | +} |
| 125 | + |
| 126 | +impl From<GenericError> for DataFusionError { |
| 127 | + fn from(err: GenericError) -> Self { |
| 128 | + DataFusionError::External(err) |
| 129 | + } |
| 130 | +} |
| 131 | + |
| 132 | +impl Display for DataFusionError { |
| 133 | + fn fmt(&self, f: &mut Formatter) -> std::fmt::Result { |
| 134 | + match *self { |
| 135 | + DataFusionError::ArrowError(ref desc) => write!(f, "Arrow error: {}", desc), |
| 136 | + DataFusionError::ParquetError(ref desc) => { |
| 137 | + write!(f, "Parquet error: {}", desc) |
| 138 | + } |
| 139 | + #[cfg(feature = "avro")] |
| 140 | + DataFusionError::AvroError(ref desc) => { |
| 141 | + write!(f, "Avro error: {}", desc) |
| 142 | + } |
| 143 | + DataFusionError::IoError(ref desc) => write!(f, "IO error: {}", desc), |
| 144 | + DataFusionError::SQL(ref desc) => { |
| 145 | + write!(f, "SQL error: {:?}", desc) |
| 146 | + } |
| 147 | + DataFusionError::NotImplemented(ref desc) => { |
| 148 | + write!(f, "This feature is not implemented: {}", desc) |
| 149 | + } |
| 150 | + DataFusionError::Internal(ref desc) => { |
| 151 | + write!(f, "Internal error: {}. This was likely caused by a bug in DataFusion's \ |
| 152 | + code and we would welcome that you file an bug report in our issue tracker", desc) |
| 153 | + } |
| 154 | + DataFusionError::Plan(ref desc) => { |
| 155 | + write!(f, "Error during planning: {}", desc) |
| 156 | + } |
| 157 | + DataFusionError::Execution(ref desc) => { |
| 158 | + write!(f, "Execution error: {}", desc) |
| 159 | + } |
| 160 | + DataFusionError::ResourcesExhausted(ref desc) => { |
| 161 | + write!(f, "Resources exhausted: {}", desc) |
| 162 | + } |
| 163 | + DataFusionError::External(ref desc) => { |
| 164 | + write!(f, "External error: {}", desc) |
| 165 | + } |
| 166 | + } |
| 167 | + } |
| 168 | +} |
| 169 | + |
| 170 | +impl error::Error for DataFusionError {} |
| 171 | + |
| 172 | +#[cfg(test)] |
| 173 | +mod test { |
| 174 | + use crate::error::DataFusionError; |
| 175 | + use arrow::error::ArrowError; |
| 176 | + |
| 177 | + #[test] |
| 178 | + fn arrow_error_to_datafusion() { |
| 179 | + let res = return_arrow_error().unwrap_err(); |
| 180 | + assert_eq!( |
| 181 | + res.to_string(), |
| 182 | + "External error: Error during planning: foo" |
| 183 | + ); |
| 184 | + } |
| 185 | + |
| 186 | + #[test] |
| 187 | + fn datafusion_error_to_arrow() { |
| 188 | + let res = return_datafusion_error().unwrap_err(); |
| 189 | + assert_eq!(res.to_string(), "Arrow error: Schema error: bar"); |
| 190 | + } |
| 191 | + |
| 192 | + /// Model what happens when implementing SendableRecrordBatchStream: |
| 193 | + /// DataFusion code needs to return an ArrowError |
| 194 | + #[allow(clippy::try_err)] |
| 195 | + fn return_arrow_error() -> arrow::error::Result<()> { |
| 196 | + // Expect the '?' to work |
| 197 | + let _foo = Err(DataFusionError::Plan("foo".to_string()))?; |
| 198 | + Ok(()) |
| 199 | + } |
| 200 | + |
| 201 | + /// Model what happens when using arrow kernels in DataFusion |
| 202 | + /// code: need to turn an ArrowError into a DataFusionError |
| 203 | + #[allow(clippy::try_err)] |
| 204 | + fn return_datafusion_error() -> crate::error::Result<()> { |
| 205 | + // Expect the '?' to work |
| 206 | + let _bar = Err(ArrowError::SchemaError("bar".to_string()))?; |
| 207 | + Ok(()) |
| 208 | + } |
| 209 | +} |
0 commit comments