Júlio César Batista

Interesting. Can you try? ``` class CustomSpidermon(Spidermon): def spider_closed(self, spider): super().spider_closed(spider) spider.logger.info(f"JSON Stats {json.dumps(self.crawler.stats)}") # or any other logger EXTENSIONS = { "CustomSpidermon": 0 # instead of spidermon } ```...

Awesome! I'd just push the script you probably used to create it so people won't need to update it manually.

My two cents here: 1. The `signal` idea is fine, though I think it can be quite hard to track 2. We could have an accessible instance of `FeedExporter` in...

Hi @MikeoPerfect, thanks for your interest in scrapy for GSoC 2019. I don't know if including this queue in scrapy is in the project roadmap. @dangra and @kmike are better...

This issue may provide some other ideas as well https://github.com/scrapy/itemadapter/issues/30

Refer to https://github.com/scrapy/itemloaders/pull/13#discussion_r444905860 for a few ideas

Quoting @kmike here (from https://github.com/scrapy/itemloaders/pull/13#discussion_r445184176) > > About the change to use dict internally, we expose it to the outside as a property item in ItemLoader. So, need to be...

Just adding a new bit of info. An item is created in `__init__` and returned in `load_item`, but the user can access it through `item` property in `ItemLoader`. Also, the...

I guess it goes down to the change. If we decide to change the API to accommodate this limitation, then a feature request. If we consider the API is fine...

I was thinking a bit about it and it seems to be a good solution, even though it might not solve all the cases. It'll allow users to not set...