xpath for img src within element

问题

How would I modify the below code so it picks out the source of any images found within the description element, which contains html? At the moment it just gets the full text from inside the element and I'm not sure how to modify this to get the sources of any img tags.

>>> from lxml import etree
>>> tree = etree.parse('temp.xml')
>>> for guide in tree.xpath('guide'):
...     '---', guide.xpath('id')[0].text
...     for pages in guide.xpath('.//pages'):
...         for page in pages:
...             '------', page.xpath('id')[0].text
...             for description in page.xpath('.//asset/description'):
...                 '---------', description.text

I have also tried this at the end:

print(description.xpath("//img/@src"))

which gives me 'None'

The XML structure is:

<guides>
<guide>
    <id>guide 1</id>
    <group>
    <id></id> 
    <type></type>
    <name></name>
    </group>
    <pages>
        <page>
            <id>page 1</id>
            <name></name>
            <description>&lt;p&gt;Some text. &lt;br /&gt;&lt;img 
            width=&quot;81&quot; 
            src=&quot;http://www.example.com/img.jpg&quot; 
             alt=&quot;wave&quot; height=&quot;63&quot; style=&quot;float: 
              right;&quot; /&gt;&lt;/p&gt;</description>
            <boxes>
                <box>
                    <id></id>
                    <name></name>
                    <type></type>
                    <map_id></map_id>
                    <column></column>
                    <position></position>
                    <hidden></hidden>
                    <created></created>
                    <updated></updated>
                    <assets>
                        <asset>
                            <id></id>
                            <name></name>
                            <type></type>
                       <description>&lt;img src=&quot;https://www.example.com/image.jpg&quot; alt=&quot;image&quot; height=&quot;42&quot; width=&quot;42&quot;&gt;</description>
                            <url/>
                            <owner>
                                <id></id>
                                <email></email>
                                <first_name></first_name>
                                <last_name></last_name>
                            </owner>
                        </asset>
                    </assets>
                </box>
            </boxes>
        </page>
    </pages>
</guide>

回答1:

The content of the description element is HTML. There are various ways of parsing it, one of them being html from lxml.

>>> description.text
'<img src="https://www.example.com/image.jpg" alt="image" height="42" width="42">'
>>> from lxml import html
>>> img = html.fromstring(description.text)
>>> img.attrib['src']
'https://www.example.com/image.jpg'

Edit, in response to comment:

>>> from lxml import etree, html
>>> tree = etree.parse('temp.xml')
>>> for guide in tree.xpath('guide'):
...     '---', guide.xpath('id')[0].text
...     for pages in guide.xpath('.//pages'):
...         for page in pages:
...             '------', page.xpath('id')[0].text
...             for description in page.xpath('.//asset/description'):
...                 '---------', html.fromstring(description.text).attrib['src']
... 
('---', 'guide 1')
('------', 'page 1')
('---------', 'https://www.example.com/image.jpg')

Edit: Handling exception.

Replace

'---------', html.fromstring(description.text).attrib['src']

with

try:
    '---------', html.fromstring(description.text).attrib['src']

except KeyError:
    '--------- No image URL present'

Edit, responding to 9 Nov comment:

from lxml import etree, html
tree = etree.parse('guides.xml')
for guide in tree.xpath('guide'):
    print('---', guide.xpath('id')[0].text)
    for pages in guide.xpath('.//pages'):
        for page in pages:
            print('------', page.xpath('id')[0].text)
            for description in page.xpath('.//asset/description'):
                try:
                    print('---------', html.fromstring(description.text).attrib['src'])
                except TypeError:
                    print('--------- no src identifiable')
                except KeyError:
                    print('--------- no src identifiable')

Output for xml file where 2nd guide element contains no HTML at all, and 3rd contains HTML without a src attribute.

--- guide 1
------ page 1
--------- https://www.example.com/image.jpg
--- guide 2
------ page 1
--------- no src identifiable
--- guide 3
------ page 1
--------- no src identifiable
--- guide 4
------ page 1
--------- https://www.example.com/image.jpg

回答2:

You could try this solution:

description.xpath("//img/@src")

来源：https://stackoverflow.com/questions/47093335/xpath-for-img-src-within-element

标签

python

xpath

lxml