raindrop-spider

A simple distribute spider based on scrapy framework.

• Reference
• Install
• Fast start
• Development

Reference

scrapy
scrapyd
scrapyd-client
mongodb

Install

$ apt-get install curl libffi-dev libxml2-dev libxslt1-dev -y
$ apt-get install mongodb -y

$ cd raindrop-spider
$ virtualenv env
$ source env/local/bin/activate
$ pip install -r requirement.txt

Fast start

config scrapyd and then start a scrapyd server

$ cd raindrop-spider
$ mkdir /etc/scrapyd
$ cp etc/scrapyd.conf /etc/scrapyd
$ source env/local/bin/activate
$ scrapyd

deploy scrapy spider

open a new ssh tab, excute commands to deploy scrapy spider on scrapyd server

$ cd raindrop-spider
$ source env/local/bin/activate
$ cd spider
$ scrapyd-deploy

test spider

params: uid is the sns account id that you wanna spider

$ curl http://localhost:6800/schedule.json -d project=spider -d spider=gh_user -d uid=account-id 

check the result

1.check log
you would see the logs below if the spider task succeed

2015-10-22 15:44:44+0000 [-] Process started:  project='spider' spider='gh_user' 
job='cf1e9a2678d311e5bdd5080027880ca6' pid=4199 
log='logs/spider/gh_user/cf1e9a2678d311e5bdd5080027880ca6.log'

2015-10-22 15:44:47+0000 [-] Process finished:  project='spider' spider='gh_user' 
job='cf1e9a2678d311e5bdd5080027880ca6' pid=4199 
log='logs/spider/gh_user/cf1e9a2678d311e5bdd5080027880ca6.log' 

2.check mongodb
you could check mongodb after seconds

$ mongo
> use raindrop

you can check the collections by command like this

> show collections

you can check data in collections by commad like this

> db.gh_user.find({})

Development

develop your own spider