[
https://issues.apache.org/jira/browse/ARROW-3547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16654082#comment-16654082
]
Wes McKinney commented on ARROW-3547:
-------------------------------------
I don't think it should be possible to obtain an R6 object instance with a
nullptr inside. Can you return NULL or NA instead?
> [R] Protect against Null crash when reading from RecordBatch
> ------------------------------------------------------------
>
> Key: ARROW-3547
> URL: https://issues.apache.org/jira/browse/ARROW-3547
> Project: Apache Arrow
> Issue Type: Improvement
> Components: R
> Reporter: Javier Luraschi
> Priority: Minor
>
> Reprex:
>
> {code:java}
> tbl <- tibble::tibble(
> int = 1:10, dbl = as.numeric(1:10),
> lgl = sample(c(TRUE, FALSE, NA), 10, replace = TRUE),
> chr = letters[1:10]
> )
> batch <- record_batch(tbl)
> bytes <- write_record_batch(batch, raw())
> stream_reader <- record_batch_stream_reader(bytes)
> batch1 <- read_record_batch(stream_reader)
> batch2 <- read_record_batch(stream_reader)
>
> # Crash
> as_tibble(batch2){code}
>
> While users should check for Null entries by running:
>
> {code:java}
> if(!batch2$is_null()) as_tibble(batch2)
> {code}
> It's harsh to trigger a crash, we should consider protecting all functions
> that use RecordBatch pointers to return NULL instead, for instance:
>
> {code:java}
> List RecordBatch__to_dataframe(const std::shared_ptr<arrow::RecordBatch>&
> batch) {
> if (batch->get() == nullptr) Rcpp::stop("Can't read from NULL record
> batch.")
> }{code}
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
