Support newlines_in_values CSV option (apache#11533)

* feat!: support `newlines_in_values` CSV option

This significantly simplifies the UX when dealing with large CSV files
that must support newlines in (quoted) values. By default, large CSV
files will be repartitioned into multiple parallel range scans. This is
great for performance in the common case but when large CSVs contain
newlines in values the parallel scan will fail due to splitting on
newlines within quotes rather than actual line terminators.

With the current implementation, this behaviour can be controlled by the
session-level `datafusion.optimizer.repartition_file_scans` and
`datafusion.optimizer.repartition_file_min_size` settings.

This commit introduces a `newlines_in_values` option to `CsvOptions` and
plumbs it through to `CsvExec`, which includes it in the test for whether
parallel execution is supported. This provides a convenient and
searchable way to disable file scan repartitioning on a per-CSV basis.

BREAKING CHANGE: This adds new public fields to types with all public
fields, which is a breaking change.

* docs: normalise `newlines_in_values` documentation

* test: add/fix sqllogictests for `newlines_in_values`

* docs: document `datafusion.catalog.newlines_in_values`

* fix: typo in config.md

* chore: suppress lint on too many arguments for `CsvExec::new`

* fix: always checkout `*.slt` with LF line endings

This is a bit of a stab in the dark, but it might fix multiline tests on
Windows.

* fix: always checkout `newlines_in_values.csv` with `LF` line endings

The default git behaviour of converting line endings for checked out files causes the `csv_files.slt` test to fail when testing `newlines_in_values`. This appears to be due to the quoted newlines being converted to CRLF, which are not then normalised when the CSV is read. Assuming that the sqllogictests do normalise line endings in the expected output, this could then lead to a "spurious" diff from the actual output.

---------

Co-authored-by: Andrew Lamb <[email protected]>

Author: 2 people

.gitattributes

@@ -1,3 +1,4 @@
 .github/ export-ignore
+datafusion/core/tests/data/newlines_in_values.csv text eol=lf
 datafusion/proto/src/generated/prost.rs linguist-generated
 datafusion/proto/src/generated/pbjson.rs linguist-generated

datafusion/common/src/config.rs

@@ -184,6 +184,16 @@ config_namespace! {
         /// Default value for `format.has_header` for `CREATE EXTERNAL TABLE`
         /// if not specified explicitly in the statement.
         pub has_header: bool, default = false
+
+        /// Specifies whether newlines in (quoted) CSV values are supported.
+        ///
+        /// This is the default value for `format.newlines_in_values` for `CREATE EXTERNAL TABLE`
+        /// if not specified explicitly in the statement.
+        ///
+        /// Parsing newlines in quoted values may be affected by execution behaviour such as
+        /// parallel file scanning. Setting this to `true` ensures that newlines in values are
+        /// parsed successfully, which may reduce performance.
+        pub newlines_in_values: bool, default = false
     }
 }
 
@@ -1593,6 +1603,14 @@ config_namespace! {
         pub quote: u8, default = b'"'
         pub escape: Option<u8>, default = None
         pub double_quote: Option<bool>, default = None
+        /// Specifies whether newlines in (quoted) values are supported.
+        ///
+        /// Parsing newlines in quoted values may be affected by execution behaviour such as
+        /// parallel file scanning. Setting this to `true` ensures that newlines in values are
+        /// parsed successfully, which may reduce performance.
+        ///
+        /// The default behaviour depends on the `datafusion.catalog.newlines_in_values` setting.
+        pub newlines_in_values: Option<bool>, default = None
         pub compression: CompressionTypeVariant, default = CompressionTypeVariant::UNCOMPRESSED
         pub schema_infer_max_rec: usize, default = 100
         pub date_format: Option<String>, default = None
@@ -1665,6 +1683,18 @@ impl CsvOptions {
         self
     }
 
+    /// Specifies whether newlines in (quoted) values are supported.
+    ///
+    /// Parsing newlines in quoted values may be affected by execution behaviour such as
+    /// parallel file scanning. Setting this to `true` ensures that newlines in values are
+    /// parsed successfully, which may reduce performance.
+    ///
+    /// The default behaviour depends on the `datafusion.catalog.newlines_in_values` setting.
+    pub fn with_newlines_in_values(mut self, newlines_in_values: bool) -> Self {
+        self.newlines_in_values = Some(newlines_in_values);
+        self
+    }
+
     /// Set a `CompressionTypeVariant` of CSV
     /// - defaults to `CompressionTypeVariant::UNCOMPRESSED`
     pub fn with_file_compression_type(

datafusion/core/src/datasource/file_format/csv.rs

@@ -233,6 +233,18 @@ impl CsvFormat {
         self
     }
 
+    /// Specifies whether newlines in (quoted) values are supported.
+    ///
+    /// Parsing newlines in quoted values may be affected by execution behaviour such as
+    /// parallel file scanning. Setting this to `true` ensures that newlines in values are
+    /// parsed successfully, which may reduce performance.
+    ///
+    /// The default behaviour depends on the `datafusion.catalog.newlines_in_values` setting.
+    pub fn with_newlines_in_values(mut self, newlines_in_values: bool) -> Self {
+        self.options.newlines_in_values = Some(newlines_in_values);
+        self
+    }
+
     /// Set a `FileCompressionType` of CSV
     /// - defaults to `FileCompressionType::UNCOMPRESSED`
     pub fn with_file_compression_type(
@@ -330,6 +342,9 @@ impl FileFormat for CsvFormat {
             self.options.quote,
             self.options.escape,
             self.options.comment,
+            self.options
+                .newlines_in_values
+                .unwrap_or(state.config_options().catalog.newlines_in_values),
             self.options.compression.into(),
         );
         Ok(Arc::new(exec))
@@ -1052,6 +1067,41 @@ mod tests {
         Ok(())
     }
 
+    #[rstest(n_partitions, case(1), case(2), case(3), case(4))]
+    #[tokio::test]
+    async fn test_csv_parallel_newlines_in_values(n_partitions: usize) -> Result<()> {
+        let config = SessionConfig::new()
+            .with_repartition_file_scans(true)
+            .with_repartition_file_min_size(0)
+            .with_target_partitions(n_partitions);
+        let csv_options = CsvReadOptions::default()
+            .has_header(true)
+            .newlines_in_values(true);
+        let ctx = SessionContext::new_with_config(config);
+        let testdata = arrow_test_data();
+        ctx.register_csv(
+            "aggr",
+            &format!("{testdata}/csv/aggregate_test_100.csv"),
+            csv_options,
+        )
+        .await?;
+
+        let query = "select sum(c3) from aggr;";
+        let query_result = ctx.sql(query).await?.collect().await?;
+        let actual_partitions = count_query_csv_partitions(&ctx, query).await?;
+
+        #[rustfmt::skip]
+        let expected = ["+--------------+",
+            "| sum(aggr.c3) |",
+            "+--------------+",
+            "| 781          |",
+            "+--------------+"];
+        assert_batches_eq!(expected, &query_result);
+        assert_eq!(1, actual_partitions); // csv won't be scanned in parallel when newlines_in_values is set
+
+        Ok(())
+    }
+
     /// Read a single empty csv file in parallel
     ///
     /// empty_0_byte.csv:

datafusion/core/src/datasource/file_format/options.rs

@@ -63,6 +63,14 @@ pub struct CsvReadOptions<'a> {
     pub escape: Option<u8>,
     /// If enabled, lines beginning with this byte are ignored.
     pub comment: Option<u8>,
+    /// Specifies whether newlines in (quoted) values are supported.
+    ///
+    /// Parsing newlines in quoted values may be affected by execution behaviour such as
+    /// parallel file scanning. Setting this to `true` ensures that newlines in values are
+    /// parsed successfully, which may reduce performance.
+    ///
+    /// The default behaviour depends on the `datafusion.catalog.newlines_in_values` setting.
+    pub newlines_in_values: bool,
     /// An optional schema representing the CSV files. If None, CSV reader will try to infer it
     /// based on data in file.
     pub schema: Option<&'a Schema>,
@@ -95,6 +103,7 @@ impl<'a> CsvReadOptions<'a> {
             delimiter: b',',
             quote: b'"',
             escape: None,
+            newlines_in_values: false,
             file_extension: DEFAULT_CSV_EXTENSION,
             table_partition_cols: vec![],
             file_compression_type: FileCompressionType::UNCOMPRESSED,
@@ -133,6 +142,18 @@ impl<'a> CsvReadOptions<'a> {
         self
     }
 
+    /// Specifies whether newlines in (quoted) values are supported.
+    ///
+    /// Parsing newlines in quoted values may be affected by execution behaviour such as
+    /// parallel file scanning. Setting this to `true` ensures that newlines in values are
+    /// parsed successfully, which may reduce performance.
+    ///
+    /// The default behaviour depends on the `datafusion.catalog.newlines_in_values` setting.
+    pub fn newlines_in_values(mut self, newlines_in_values: bool) -> Self {
+        self.newlines_in_values = newlines_in_values;
+        self
+    }
+
     /// Specify the file extension for CSV file selection
     pub fn file_extension(mut self, file_extension: &'a str) -> Self {
         self.file_extension = file_extension;
@@ -490,6 +511,7 @@ impl ReadOptions<'_> for CsvReadOptions<'_> {
             .with_delimiter(self.delimiter)
             .with_quote(self.quote)
             .with_escape(self.escape)
+            .with_newlines_in_values(self.newlines_in_values)
             .with_schema_infer_max_rec(self.schema_infer_max_records)
             .with_file_compression_type(self.file_compression_type.to_owned());
 

datafusion/core/src/datasource/physical_plan/csv.rs

@@ -59,6 +59,7 @@ pub struct CsvExec {
     quote: u8,
     escape: Option<u8>,
     comment: Option<u8>,
+    newlines_in_values: bool,
     /// Execution metrics
     metrics: ExecutionPlanMetricsSet,
     /// Compression type of the file associated with CsvExec
@@ -68,13 +69,15 @@ pub struct CsvExec {
 
 impl CsvExec {
     /// Create a new CSV reader execution plan provided base and specific configurations
+    #[allow(clippy::too_many_arguments)]
     pub fn new(
         base_config: FileScanConfig,
         has_header: bool,
         delimiter: u8,
         quote: u8,
         escape: Option<u8>,
         comment: Option<u8>,
+        newlines_in_values: bool,
         file_compression_type: FileCompressionType,
     ) -> Self {
         let (projected_schema, projected_statistics, projected_output_ordering) =
@@ -91,6 +94,7 @@ impl CsvExec {
             delimiter,
             quote,
             escape,
+            newlines_in_values,
             metrics: ExecutionPlanMetricsSet::new(),
             file_compression_type,
             cache,
@@ -126,6 +130,17 @@ impl CsvExec {
         self.escape
     }
 
+    /// Specifies whether newlines in (quoted) values are supported.
+    ///
+    /// Parsing newlines in quoted values may be affected by execution behaviour such as
+    /// parallel file scanning. Setting this to `true` ensures that newlines in values are
+    /// parsed successfully, which may reduce performance.
+    ///
+    /// The default behaviour depends on the `datafusion.catalog.newlines_in_values` setting.
+    pub fn newlines_in_values(&self) -> bool {
+        self.newlines_in_values
+    }
+
     fn output_partitioning_helper(file_scan_config: &FileScanConfig) -> Partitioning {
         Partitioning::UnknownPartitioning(file_scan_config.file_groups.len())
     }
@@ -196,15 +211,15 @@ impl ExecutionPlan for CsvExec {
     /// Redistribute files across partitions according to their size
     /// See comments on [`FileGroupPartitioner`] for more detail.
     ///
-    /// Return `None` if can't get repartitioned(empty/compressed file).
+    /// Return `None` if can't get repartitioned (empty, compressed file, or `newlines_in_values` set).
     fn repartitioned(
         &self,
         target_partitions: usize,
         config: &ConfigOptions,
     ) -> Result<Option<Arc<dyn ExecutionPlan>>> {
         let repartition_file_min_size = config.optimizer.repartition_file_min_size;
-        // Parallel execution on compressed CSV file is not supported yet.
-        if self.file_compression_type.is_compressed() {
+        // Parallel execution on compressed CSV files or files that must support newlines in values is not supported yet.
+        if self.file_compression_type.is_compressed() || self.newlines_in_values {
             return Ok(None);
         }
 
@@ -589,6 +604,7 @@ mod tests {
             b'"',
             None,
             None,
+            false,
             file_compression_type.to_owned(),
         );
         assert_eq!(13, csv.base_config.file_schema.fields().len());
@@ -658,6 +674,7 @@ mod tests {
             b'"',
             None,
             None,
+            false,
             file_compression_type.to_owned(),
         );
         assert_eq!(13, csv.base_config.file_schema.fields().len());
@@ -727,6 +744,7 @@ mod tests {
             b'"',
             None,
             None,
+            false,
             file_compression_type.to_owned(),
         );
         assert_eq!(13, csv.base_config.file_schema.fields().len());
@@ -793,6 +811,7 @@ mod tests {
             b'"',
             None,
             None,
+            false,
             file_compression_type.to_owned(),
         );
         assert_eq!(14, csv.base_config.file_schema.fields().len());
@@ -858,6 +877,7 @@ mod tests {
             b'"',
             None,
             None,
+            false,
             file_compression_type.to_owned(),
         );
         assert_eq!(13, csv.base_config.file_schema.fields().len());
@@ -953,6 +973,7 @@ mod tests {
             b'"',
             None,
             None,
+            false,
             file_compression_type.to_owned(),
         );
 

datafusion/core/src/physical_optimizer/enforce_distribution.rs

@@ -1472,6 +1472,7 @@ pub(crate) mod tests {
             b'"',
             None,
             None,
+            false,
             FileCompressionType::UNCOMPRESSED,
         ))
     }
@@ -1496,6 +1497,7 @@ pub(crate) mod tests {
             b'"',
             None,
             None,
+            false,
             FileCompressionType::UNCOMPRESSED,
         ))
     }
@@ -3770,6 +3772,7 @@ pub(crate) mod tests {
                     b'"',
                     None,
                     None,
+                    false,
                     compression_type,
                 )),
                 vec![("a".to_string(), "a".to_string())],

datafusion/core/src/physical_optimizer/projection_pushdown.rs

@@ -186,6 +186,7 @@ fn try_swapping_with_csv(
             csv.quote(),
             csv.escape(),
             csv.comment(),
+            csv.newlines_in_values(),
             csv.file_compression_type,
         )) as _
     })
@@ -1700,6 +1701,7 @@ mod tests {
             0,
             None,
             None,
+            false,
             FileCompressionType::UNCOMPRESSED,
         ))
     }
@@ -1723,6 +1725,7 @@ mod tests {
             0,
             None,
             None,
+            false,
             FileCompressionType::UNCOMPRESSED,
         ))
     }

datafusion/core/src/physical_optimizer/replace_with_order_preserving_variants.rs

@@ -1503,6 +1503,7 @@ mod tests {
             b'"',
             None,
             None,
+            false,
             FileCompressionType::UNCOMPRESSED,
         ))
     }

datafusion/core/src/test/mod.rs

@@ -99,6 +99,7 @@ pub fn scan_partitioned_csv(partitions: usize, work_dir: &Path) -> Result<Arc<Cs
         b'"',
         None,
         None,
+        false,
         FileCompressionType::UNCOMPRESSED,
     )))
 }
@@ -283,6 +284,7 @@ pub fn csv_exec_sorted(
         0,
         None,
         None,
+        false,
         FileCompressionType::UNCOMPRESSED,
     ))
 }
@@ -339,6 +341,7 @@ pub fn csv_exec_ordered(
         b'"',
         None,
         None,
+        false,
         FileCompressionType::UNCOMPRESSED,
     ))
 }

datafusion/core/tests/data/newlines_in_values.csv

@@ -0,0 +1,13 @@
+id,message
+1,"hello
+world"
+2,"something
+else"
+3,"
+many
+lines
+make
+good test
+"
+4,unquoted
+value,end