How to access to meta values?

[llermaly]

Hi,

I'm doing this request:
curl -XPOST -d '{ "spider_name":"quotes", "start_requests":true, "request":{ "meta": {
"test": "1", } } }' "http://138.219.228.215:9080/crawl.json"

Then I try to access from my spider by print(response.meta) and this is what it shows:

{'depth': 0, 'download_latency': 0.03323054313659668, 'download_slot': 'URL', 'download_timeout': 180.0}

of course response.meta["test"] throws error.

I need to use this "test" parameter to fill the form request

EDIT: spider : https://pastebin.com/EFz818qL

thanks!

[changtingQ]

Do you have been slove the issues? can you tell me the method

[pawelmhm]

indeed that doesn't work, I'm going to fix it.

edit: it only happens when you set start_requests: True and provide request object with meta. In this case meta is not passed to spider.

[rythm-of-the-red-man]

I guess this workaround might help you for now.
You can patch scrapyrt to take parameters from json directly. I mean, not under meta key, but as key. To achieve this, modification of resources.py is required. In tl;dr version just paste

        crawler_params = api_params.copy()
        for api_param in ['max_requests', 'start_requests', 'spider_name', 'url']:
            crawler_params.pop(api_param, None)
        kwargs.update(crawler_params)

below first try:except block in prepare_crawl method, so it looks as on attached screenshot

.
After that when you set attributes in spider constructor like this:

    def __init__(self, name=name, **kwargs):
        super().__init__(name=name, **kwargs)
        ### Getting get/post args ###
        for k, v in kwargs.items():
            setattr(self, k, v)

test parameter should be avaiable as self.test

I'm not sure, but I think I found this solution in open PR, @pawelmhm maybe it is good idea to merge it? I'm using this patch around half year and it seems to be fine.