How to do the following in an async way in Julia:
the JavaScript-ish pseudo-code would be:
for i in 1:100
asyncCall(get(webpageAddr[i]), callback(err, data){
insertToDB(db, coll, data)
})
end
I am interested in how the async part of the code should look like (no so much the web query and db access, I'll take care of those). The 100 queries can run independently (asynchronously) from one another and the rest of the Julia code.
Julia prefers to write code in a synchronous way without losing the power of parallelism, so to speak. You do not have to think in terms of callbacks.
Below, you see @async macro, which runs task (coroutine) on the local process only and @sync macro to wait until all the enclosed tasks are done.
@sync begin
for i=1:10
@async begin
println("I mimic the get for $i"); # -> get(webpageAddr[i])
sleep(rand(1:3)) #waiting for server answer at random number of seconds
println("I mimic the callback for $i") # -> callback(err, data){insertToDB(db, coll, data}
end
end
end
I encouraged you to dig into the details of parallel computing in Julia, because these are enormous.
get with sleep(rand), which makes it crystal clear how async works.