热门文章

最新文章

python 提取html文本的方法

发布时间:2021-06-22 15:26:45

在解决自然语言处理问题时,有时你需要获得大量的文本集。互联网是文本的最大来源,但是从任意HTML页面提取文本是一项艰巨而痛苦的任务。本文将讲述python高效提取html文本的方法

假设我们需要从各种网页中提取全文,并且要剥离所有HTML标记。通常,默认解决方案是使用BeautifulSoup软件包中的get_text方法,该方法内部使用lxml。这是一个经过充分测试的解决方案,但是在处理成千上万个HTML文档时可能会非常慢。
通过用selectolax替换BeautifulSoup,您几乎可以免费获得5-30倍的加速!
这是一个简单的基准测试,可分析commoncrawl(`处理NLP问题时,有时您需要获得大量的文本集。互联网是文本的最大来源,但是不幸的是,从任意HTML页面提取文本是一项艰巨而痛苦的任务。
假设我们需要从各种网页中提取全文,并且要剥离所有HTML标记。通常,默认解决方案是使用BeautifulSoup软件包中的get_text方法,该方法内部使用lxml。这是一个经过充分测试的解决方案,但是在处理成千上万个HTML文档时可能会非常慢。
通过用selectolax替换BeautifulSoup,您几乎可以免费获得5-30倍的加速!这是一个简单的基准测试,可分析commoncrawl(https://commoncrawl.org/)的10,000个HTML页面:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
# coding: utf-8
 
from time import time
 
import warc
from bs4 import BeautifulSoup
from selectolax.parser import HTMLParser
  
def get_text_bs(html):
    tree = BeautifulSoup(html, 'lxml')
 
    body = tree.body
    if body is None:
        return None
 
    for tag in body.select('script'):
        tag.decompose()
    for tag in body.select('style'):
        tag.decompose()
 
    text = body.get_text(separator='\n')
    return text
  
def get_text_selectolax(html):
    tree = HTMLParser(html)
 
    if tree.body is None:
        return None
 
    for tag in tree.css('script'):
        tag.decompose()
    for tag in tree.css('style'):
        tag.decompose()
 
    text = tree.body.text(separator='\n')
    return text
  
def read_doc(record, parser=get_text_selectolax):
    url = record.url
    text = None
 
    if url:
        payload = record.payload.read()
        header, html = payload.split(b'\r\n\r\n', maxsplit=1)
        html = html.strip()
 
        if len(html) > 0:
            text = parser(html)
 
    return url, text
  
def process_warc(file_name, parser, limit=10000):
    warc_file = warc.open(file_name, 'rb')
    t0 = time()
    n_documents = 0
    for i, record in enumerate(warc_file):
        url, doc = read_doc(record, parser)
 
        if not doc or not url:
            continue
 
        n_documents += 1
 
        if i > limit:
            break
 
    warc_file.close()
    print('Parser: %s' % parser.__name__)
    print('Parsing took %s seconds and produced %s documents\n' % (time() - t0, n_documents))
1
2
3
4
5
6
7
8
>>> ! wget https://commoncrawl.s3.amazonaws.com/crawl-data/CC-MAIN-2018-05/segments/1516084886237.6/warc/CC-MAIN-20180116070444-20180116090444-00000.warc.gz
>>> file_name = "CC-MAIN-20180116070444-20180116090444-00000.warc.gz"
>>> process_warc(file_name, get_text_selectolax, 10000)
Parser: get_text_selectolax
Parsing took 16.170367002487183 seconds and produced 3317 documents
>>> process_warc(file_name, get_text_bs, 10000)
Parser: get_text_bs
Parsing took 432.6902508735657 seconds and produced 3283 documents

显然,这并不是对某些事物进行基准测试的最佳方法,但是它提供了一个想法,即selectolax有时比lxml快30倍。
selectolax最适合将HTML剥离为纯文本。如果我有10,000多个HTML片段,需要将它们作为纯文本索引到Elasticsearch中。(Elasticsearch有一个html_strip文本过滤器,但这不是我想要/不需要在此上下文中使用的过滤器)。事实证明,以这种规模将HTML剥离为纯文本实际上是非常低效的。那么,最有效的方法是什么?

  • PyQuery

1
2
3
from pyquery import PyQuery as pq
 
text = pq(html).text()
  • selectolax

1
2
3
from selectolax.parser import HTMLParser
 
text = HTMLParser(html).text()
  • 正则表达式

1
2
3
4
import re
 
regex = re.compile(r'<.*?>')
text = clean_regex.sub('', html)

结果

我编写了一个脚本来计算时间,该脚本遍历包含HTML片段的10,000个文件。注意!这些片段不是完整的<html>文档(带有<head>和<body>等),只是HTML的一小部分。平均大小为10,314字节(中位数为5138字节)。结果如下:

1
2
3
4
5
6
7
8
9
10
11
12
pyquery
  SUM:    18.61 seconds
  MEAN:   1.8633 ms
  MEDIAN: 1.0554 ms
selectolax
  SUM:    3.08 seconds
  MEAN:   0.3149 ms
  MEDIAN: 0.1621 ms
regex
  SUM:    1.64 seconds
  MEAN:   0.1613 ms
  MEDIAN: 0.0881 ms

我已经运行了很多次,结果非常稳定。重点是:selectolax比PyQuery快7倍。

正则表达式好用?真的吗?

对于最基本的HTML Blob,它可能工作得很好。实际上,如果HTML是<p> Foo&amp; Bar </ p>,我希望纯文本转换应该是Foo&Bar,而不是Foo&amp; bar。
更重要的一点是,PyQuery和selectolax支持非常特定但对我的用例很重要的内容。在继续之前,我需要删除某些标签(及其内容)。例如:

1
2
3
<h4 class="warning">This should get stripped.</h4>
<p>Please keep.</p>
<div style="display: none">This should also get stripped.</div>

正则表达式永远无法做到这一点。

2.0 版本

因此,我的要求可能会发生变化,但基本上,我想删除某些标签。例如:<div class =“ warning”>  、 <div class =“ hidden”> 和 <div style =“ display:none”>。因此,让我们实现一下:

  • PyQuery

1
2
3
4
5
6
7
8
9
10
11
from pyquery import PyQuery as pq
 
_display_none_regex = re.compile(r'display:\s*none')
 
doc = pq(html)
doc.remove('div.warning, div.hidden')
for div in doc('div[style]').items():
    style_value = div.attr('style')
    if _display_none_regex.search(style_value):
        div.remove()
text = doc.text()
  • selectolax

1
2
3
4
5
6
7
8
9
10
11
12
from selectolax.parser import HTMLParser
 
_display_none_regex = re.compile(r'display:\s*none')
 
tree = HTMLParser(html)
for tag in tree.css('div.warning, div.hidden'):
    tag.decompose()
for tag in tree.css('div[style]'):
    style_value = tag.attributes['style']
    if style_value and _display_none_regex.search(style_value):
        tag.decompose()
text = tree.body.text()

这实际上有效。当我现在为10,000个片段运行相同的基准时,新结果如下:

1
2
3
4
5
6
7
8
9
10
pyquery
  SUM:    21.70 seconds
  MEAN:   2.1701 ms
  MEDIAN: 1.3989 ms
selectolax
  SUM:    3.59 seconds
  MEAN:   0.3589 ms
  MEDIAN: 0.2184 ms
regex
  Skip

同样,selectolax击败PyQuery约6倍。

结论

正则表达式速度快,但功能弱。selectolax的效率令人印象深刻。

以上就是python 提取html文本的方法的详细内容


返回顶部