How to efficiently match offerlines based on a rolling window in R - r

Question

I currently have 600 000 + offerlines, where I want to efficiently match them based on the product bought & the timeframe.
With timeframe I mean that from the base line, I look at all offerlines that are maximally either 10 days before the base line or 10 days after. Everything in between with the same product should be matched.
However, it is very time expensive & after running it for a complete night, I only got to line 45000.
I know parallelism is one option, but I want to know if there are better ways (packages, functions, logic).
Input data
Offerline n°,Customer n°,Offerdate,Product
(we clean to 1 offerline n° per day per custno, for a certain product)
Logic => match lines with same product, different Customer n°
Desired output
Base customer, Related Customer, Offerline n°, Matched Offerline n°, Product, Offerdate base, Offerdate matched line.
Current code
for(i in 1:nrow(X)){
sku <- X[i,]$product
date <- X[i,]$order.offer_date
cust <- X[i,]$customer_code
oon <- X[i,]$order.offer_number
F <- data.frame()
F <- X %>%
filter(product == (X[i,]$product) & (order.offer_date <= date + 10 & order.offer_date >= date - 10)& customer_code != cust)
if(nrow(F)== 0){next}
else{
for(j in 1:nrow(F)){
skuc <- F[j,]$product
datec <- F[j,]$order.offer_date
custc <- F[j,]$customer_code
oonc <- F[j,]$order.offer_number
if(custc == cust | oon == oonc){next}
else if(skuc != sku){next}
else if(skuc == sku){
if(datec <= date + 10 & datec >= date - 10){
z <- z + 1
Y[z,]$count <- j
Y[z,]$base <- oon
Y[z,]$related <- oonc
Y[z,]$baseSku <- sku
Y[z,]$relSku <- skuc
Y[z,]$basedate <- as.Date(date)
Y[z,]$reldate <- as.Date(datec)
Y[z,]$basecust <- cust
Y[z,]$relcust <- custc
}
else{next}
}
}
next
}
}