Version 2.0: chunk_size parameter broken in read_delim_chunked function

[timothy-barry]

Hi,

I just updated to version 2.0. The chunk_size parameter of the read_delim_chunked function does not seem to work. Regardless of the argument that I pass, the chunk size remains fixed at 10000.

Thanks.

EDIT: Here is a reproducible example. Chunk size argument does not work when reading mtcars.csv:

library(readr)
f <- function(x, pos) print(nrow(x))
read_delim_chunked(file = readr_example("mtcars.csv"),
                   delim = ",",
                   callback = SideEffectChunkCallback$new(f),
                   chunk_size = 5)

The output is "32," indicating that all 32 rows were read at once.

[svraka]

Ran into the same problem, it looks like chunk_size is not passed to the read_delimited_chunked() call in read_delim_chunked().

readr/R/read_delim_chunked.R

Lines 78 to 82 in 3c0e4d8

	read_delimited_chunked(file,
	callback = callback, tokenizer = tokenizer, col_names = col_names, col_types = col_types,
	locale = locale, skip = skip, skip_empty_rows = skip_empty_rows,
	comment = comment, guess_max = guess_max, progress = progress
	)

[jimhester]

Thanks for opening the issue and providing a reproducible example.
This was a regression in readr 2.0.0, it should be fixed now in the current master branch.