Github user HyukjinKwon commented on the issue:
https://github.com/apache/spark/pull/17136
Oh, @maropu, I have been looking into R's `read.csv` before adding some
comments in the JIRAs.
with the data below:
```
a,b,c
a,b,c,d,e,d,d
```
```
> read.csv("test.csv")
Error in read.table(file = file, header = header, sep = sep, quote = quote,
:
```
with the data below:
```csv
a,b,c,d,e,d,d
a,b,c
```
```r
> read.csv("test.csv")
a b c d e d.1 d.2
1 a b c NA NA NA NA
```
So, IMHO, we might better follow R's `read.csv` for now but of course I
guess we should take a look for other libraries.
I am actually a bit worried of behaviour change because `PERMISSIVE` has
been the default mode and since it was as the thirdparty library (Spark 1.3+).
Another concern is, it seems we should produce `columnNameOfCorruptRecord`
as we are doing in JSON datasource if we will treat those tokens as malformed
ones in `PERMISSIVE` mode.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
