Magento产品集合过滤

Question

Why hello there Magento gurus,

I'm importing large product catalog in Magento, so every milisec counts.

Import is kind of done, but I'm now searching for parts that could be improved and found one interesting ( at least for me ) moment -> http://screencast.com/t/Gq8VumQluKr

This is for 5K SKU's and as you can see, there is a big difference.

In second collection I used getData() and then from array fetched entity_id .

But in first collection, I didn't use getData() and fetched ID from model function.

I know that getId() most likely is call to function, but even if I call some getLala() or getData('lala') , using getData() on collection and then looping will be hundreds ( or even thousands ) of times faster.

My question is - why there is such difference in performance and how these methods differ?

Thanks.

Answer 1

I believe your benchmarking code is incorrect. You start benchmarking after getData() function call, but getData() is doing all the heavy lifting. You should start taking time before getData() function call. I will explain it in more detail:

When you use product collection in for loop without getData, Magento internally will use ArrayIterator, to iterate over product list. And to get that list, Magento will call getData function internally (check out load() function in lib/Varien/Data/Collection/DB.php) if it's not already called.

Magento is designed to lazily load the product list only when you request it. If you call getData() function on collection, you request the product list and it gets collected at that time, or if you use for loop, then it will request the product list before the loop.

getId() and $product['entity_id'] do not differ much, internally it is doing almost the same thing. Product is Varien Object that implements ArrayAccess, so you can call $product['id'] as well as getId(), that both call getData('id') function internally. I believe there was a slight difference between 'id' and 'entity_id', but I'm not sure about that.