XPath

Krzysztof "noisy" Szumny

source: http://www.wykop.pl/wpis/13367209/, #programowanie

https://chrome.google.com/webstore/detail/xpath-helper/hgimnogjllphhhkhlmebbmlgjoejdpjl

Quick demo

https://github.com/noisy?tab=repositories

# Title of repositories
//*[@id="site-container"]/div/div/div[2]/div[2]/div/ul/li/h3/a


# Information whether repository was forked
//*[@id="site-container"]/div/div/div[2]/div[2]/div/ul/li/p[@class='repo-list-info']


# filtering by forked repositories
//*[@id="site-container"]/div/div/div[2]/div[2]/div/ul/li[p[@class='repo-list-info']]/h3/a


# filtering by not forked repositories
//*[@id="site-container"]/div/div/div[2]/div[2]/div/ul/li[not(p[@class='repo-list-info'])]/h3/a

HTML, Xpath and Python

import urllib2
from lxml import etree
from lxml.etree import tostring

url =  "https://github.com/noisy?tab=repositories"

response = urllib2.urlopen(url)
tree = etree.parse(response, etree.HTMLParser())

xpath_selector1 = '//div/ul/li/h3/a'
xpath_selector2 = '//div/ul/li/p[@class="repo-list-info"]'
xpath_selector3 = '//div/ul/li[p[@class="repo-list-info"]]/h3/a'
xpath_selector4 = '//div/ul/li[not(p[@class="repo-list-info"])]/h3/a'


elements = tree.xpath(xpath_selector4)

for elem in elements:
    # print tostring(elem)
    print elem.text

Case Study - parsing XMLs

<offers xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" version="1">
  <group name="books">
    <o id="28" price="34.58" url="https://bookstore2.pl/44334002/33">
      <name>Czeski underground></name>
      <desc></desc>
      <cat>Księgarnia/Książki/Pozostałe książki></cat>
      <imgs>
        <main url="http://bookstore2.pl/gfx2/logo.jpg"/>
      </imgs>
      <attrs>
        <a name="Autor">Martin Machovec></a>
        <a name="ISBN">9788374323581</a>
        <a name="Ilosc_stron">408</a>
        <a name="Wydawnictwo">Oficyna Wydawnicza Atut</a>
        <a name="Rok_wydania">2008</a>
      </attrs>
    </o>
  </group>
</offers>

http://SpisTresci.pl/

from spistresci.connectors.common import Ceneo

class Bookstore2(Ceneo):

    xml_tag_dict = {
        'external_id': "@id",
        'price': "@price",
        'url': "@url",
        'title': "./name",
        'description': "./desc",
        'category': "./cat",
        'cover': "./imgs/main/@url",
        'authors': "./attrs/a[@name='Autor']",
        'isbns': "./attrs/a[@name='ISBN']",
        'page_count': "./attrs/a[@name='Ilosc_stron']",
        'publisher': "./attrs/a[@name='Wydawnictwo']",
        'date': "./attrs/a[@name='Rok_wydania']",
    }

Case Study - parsing XMLs

<offers>
  <offer>
    <id>a-tale-of-two-cities</id>
    <name>A Tale Of Two Cities></name>
    <price>12,00</price>
    <url>http://some-bookstore.pl/37f416,,1a881,,,</url>
    <categoryId>Księgarnia,Powiesc></categoryId>
    <description/>
    <image>http://some-bookstore.pl/37f416.jpg</image>
    <attributes>
      <attribute>
        <name>Wydawca</name>
        <value>Macmillan></value>
      </attribute>
      <attribute>
        <name>Autor</name>
        <value>Charles Dickens></value>
      </attribute>
      <attribute>
        <name>Dugość</name>
        <value>1 godz. 9 min.></value>
      </attribute>
    </attributes>
  </offer>
  <offer>...</offer>
  <offer>...</offer>
  <offer>...</offer>
</offers>

http://SpisTresci.pl/

from connectors.common import Afiliant
from connectors.generic import GenericBook

class SomeBookstore(Afiliant):

    xml_tag_dict = dict(Afiliant.xml_tag_dict.items() + [
        ('cover', u"./image"),
        ('publisher', u"./attributes/attribute[name='Wydawca']/value"),
        ('authors', u"./attributes/attribute[name='Autor']/value"),
        ('audio_time', u"./attributes/attribute[name='Dugość']/value"),
    ])

Axes

child::
/child::foo/child::bar/child::baz - /foo/bar/baz
attribute::
/child::foo/attribute::bar - /foo/@bar
descendant::
parent::
ancestor::
following-sibling::
preceding-sibling::
following::
preceding::
namespace::
self::
descendand-or-self::
ancestor-or-self::

XPath Tutorials

Useful Cheat Sheets

http://bit.ly/xpath-cheetsheets

link do slajdów: http://slides.com/noisy/xpath/