Hello!
I think this is probably a simple answer, but I've been banging on it for too
long now. I want to process the lines of a file in parallel. The R code I am
trying to emulate is:
library(parallel)
library(data.table)
file <- commandArgs(trailingOnly=T)
df <- fread('cat /dev/stdin', header=F, sep="\t")
totals <- mclapply(df, function(x) {
sum(grepl("bc", x, ignore.case=T))
}, mc.cores=ncol(df))
print(sum(unlist(totals))
Run
Here is my Nim code:
import strutils, sequtils, threadpool
{.experimental: "parallel".}
func countBC(line:string): int =
for val in line.split('\t'):
if "bc" in toLower(val[1..<4]):
inc(result)
proc main() =
var count = 0
var lines = stdin.readAll.splitLines[0 .. ^2]
var counts = newSeqOfCap[int](lines.len)
parallel:
for i in 0 .. lines.high:
counts[i] = spawn countBC(lines[i])
echo foldl(counts, a + b, 0)
main()
Run
The error I get is count_lines_parallel_nim.nim(18, 20) Error: cannot prove: i
<= len(counts) + -1 (bounds check)
What do I need to do to let the compiler know that counts has the same length
as lines?
