Cleanup CSV WriterBuilder, Default to AutoSI Second Precision (#4735)#4909
Cleanup CSV WriterBuilder, Default to AutoSI Second Precision (#4735)#4909tustvold merged 4 commits intoapache:masterfrom
Conversation
| } | ||
|
|
||
| /// Set whether to write headers | ||
| #[deprecated(note = "Use Self::with_header")] |
There was a problem hiding this comment.
| const DEFAULT_DATE_FORMAT: &str = "%F"; | ||
| const DEFAULT_TIME_FORMAT: &str = "%T"; | ||
| const DEFAULT_TIMESTAMP_FORMAT: &str = "%FT%H:%M:%S.%9f"; | ||
| const DEFAULT_TIMESTAMP_TZ_FORMAT: &str = "%FT%H:%M:%S.%9f%:z"; |
There was a problem hiding this comment.
These are just the definitions for an RFC3339 representation
arrow-csv/src/writer.rs
Outdated
| @@ -97,26 +92,14 @@ pub struct Writer<W: Write> { | |||
| /// Is the beginning-of-writer | |||
| beginning: bool, | |||
| /// The value to represent null entries | |||
There was a problem hiding this comment.
| /// The value to represent null entries | |
| /// The value to represent null entries. If not specified, [`DEFAULT_NULL_VALUE`] is used |
There was a problem hiding this comment.
It might help to add information about the default values for the formats above as well
| pub struct WriterBuilder { | ||
| /// Optional column delimiter. Defaults to `b','` | ||
| delimiter: Option<u8>, | ||
| delimiter: u8, |
There was a problem hiding this comment.
is the idea that a u8 is more space efficient than Option<u8>?
There was a problem hiding this comment.
It's more just unnecessary 😅
| } | ||
| drop(writer); | ||
|
|
||
| let left = "c1,c2 |
There was a problem hiding this comment.
Why was this test removed?
I see the output is identical in the two cases -- is there still coverage showing that `
arrow_csv::WriterBuilder::new()
.with_rfc3339()
Produces the default values?
Perhaps given .with_rfc3339() is the default, maybe it should should be deprecated too 🤔 ?
2 participants
Which issue does this PR close?
Closes #4735
Closes #4906
Rationale for this change
Cleans up
WriterBuilderto not default the format strings, this not only saves allocations, but also allows using the fast formatting path for RFC3339 formatting.What changes are included in this PR?
Are there any user-facing changes?
This changes timestamps to default to AutoSI. I'm not sure this really counts as a breaking change, any well-behaved parser should handle this