Scraping new ESPN site using xpath [Python]

Question

I am trying to scrape the new ESPN NBA scoreboard. Here is a simple script which should return the start times for all games on 4/5/15:

import requests
import lxml.html
from lxml.cssselect import CSSSelector

doc =  lxml.html.fromstring(requests.get('http://scores.espn.go.com/nba/scoreboard?date=20150405').text)

#xpath
print doc.xpath("//title/text()") #print page title
print doc.xpath("//span/@time")
print doc.xpath("//span[@class='time']")
print doc.xpath("//span[@class='time']/text()")

#CCS Selector
sel = CSSSelector('span.time')
for i in sel(doc):
    print i.text

It doesn't return anything, but the page title:

['NBA Basketball Scores - NBA Scoreboard - ESPN']
[]
[]
[]

Can anyone help? Thanks

Answer 1

The nature of the page is quite dynamic - there are asynchronous XHR requests, javascript logic involved. requests is not a browser and downloads only the initial HTML page and there are no span elements with class="time" in the HTML that requests gets.

One of the options to approach the problem would be to involve a real browser using selenium. Here is an example using PhantomJS headless browser:

>>> from selenium import webdriver
>>> 
>>> url = "http://scores.espn.go.com/nba/scoreboard?date=20150405"
>>> 
>>> driver = webdriver.PhantomJS()
>>> driver.get(url)
>>> 
>>> elements = driver.find_elements_by_css_selector("span.time")
>>> for element in elements:
...     print element.text
... 

1:00 PM ET
3:30 PM ET
6:00 PM ET
7:00 PM ET
7:30 PM ET
9:00 PM ET
9:30 PM ET 

---

Alternatively, you can look for the desired data in the data-data attribute of the div with id="scoreboard-page":

import json
from pprint import pprint

import lxml.html
import requests

response = requests.get('http://scores.espn.go.com/nba/scoreboard?date=20150405')
doc = lxml.html.fromstring(response.content)

data = doc.xpath("//div[@id='scoreboard-page']/@data-data")[0]
data = json.loads(data)

pprint(data)