CODEWITHSUNDEEP
pythonregexhtml-parsingnested-lists
XdansLaFoule
XdansLaFoule

Reputation: 87

Automatically generate nested table of contents based on heading tags using python

I'm trying to create a nested table of content based on heading tags of HTML.

My HTML file:

<html>
<head>
  <META http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
  <h1>
            My report Name
  </h1>
  <h1 id="2">First Chapter                          </h1>
  <h2 id="3"> First Sub-chapter of the first chapter</h2>
  <ul>
    <h1 id="text1">Useless h1</h1>
    <p>
      some text
    </p>
  </ul>
  <h2 id="4">Second Sub-chapter of the first chapter </h2>
  <ul>
    <h1 id="text2">Useless h1</h1>
    <p>
      some text
    </p>
  </ul>
  <h1 id="5">Second Chapter                          </h1>
  <h2 id="6">First Sub-chapter of the Second chapter </h2>
  <ul>
    <h1 id="text6">Useless h1</h1>
    <p>
      some text
    </p>
  </ul>
  <h2 id="7">Second Sub-chapter of the Second chapter </h2>
  <ul>
    <h1 id="text6">Useless h1</h1>
    <p>
      some text
    </p>
  </ul>
</body>
</html>

My python code:

import from lxml import html
from bs4 import BeautifulSoup as soup
import re
import codecs
#Access to the local URL(Html file)
f = codecs.open("C:\\x\\test.html", 'r')
page = f.read()
f.close()
#html parsing
page_soup = soup(page,"html.parser")
tree = html.fromstring(page)#extract report name
ref = page_soup.find("h1",{"id": False}).text.strip()
print("the name of the report is : " + ref + " \n")

chapters = page_soup.findAll('h1', attrs={'id': re.compile("^[0-9]*$")})
print("We have " + str(len(chapters)) + " chapter(s)")
for index, chapter in enumerate(chapters):
    print(str(index+1) +"-" + str(chapter.text.strip()) + "\n")

sub_chapters = page_soup.findAll('h2', attrs={'id': re.compile("^[0-9]*$")})
print("We have " + str(len(sub_chapters)) + " sub_chapter(s)")
for index, sub_chapter in enumerate(sub_chapters):
    print(str(index+1) +"-" +str(sub_chapter.text.strip()) + "\n")

With this code, I am able to get all the chapters and all the sub-chapters but it is not my goal.

My goal is to get the below as my table of contents:

1-First Chapter
    1-First sub-chapter of the first chapter
    2-Second sub-chapter of the first chapter
2-Second Chapter    
    1-First sub-chapter of the Second chapter
    2-Second sub-chapter of the Second chapter

Any recommendation or ideas on how to achieve my desired table of contents format?

Upvotes: 4

Views: 592

Answers (2)

bob0the0mighty
bob0the0mighty

Reputation: 782

If your willing to changer your HTML layout to something similar to below:

<html>

<head>
  <META http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>

<body>
  <article>
    <h1>
      My report Name
    </h1>
    <section>
      <h2 id="chapter-one">First Chapter</h2>
      <section>
        <h3 id="one-one"> First Sub-chapter of the first chapter</h3>
        <ul>
          <h4 id="text1">Useless h4</h4>
          <p>
            some text
          </p>
        </ul>
      </section>
      <section>
        <h3 id="one-two">Second Sub-chapter of the first chapter</h3>
        <ul>
          <h4 id="text2">Useless h4</h4>
          <p>
            some text
          </p>
        </ul>
      </section>
    </section>
    <section>
      <h2 id="chapter-two">Second Chapter </h2>
      <section>
        <h3 id="two-one">First Sub-chapter of the Second chapter</h3>
        <ul>
          <h4 id="text6">Useless h4</h4>
          <p>
            some text
          </p>
        </ul>
      </section>
      <section>
        <h3 id="two-two">Second Sub-chapter of the Second chapter</h3>
        <ul>
          <h4 id="text6">Useless h4</h4>
          <p>
            some text
          </p>
        </ul>
      </section>
    </section>
  </article>
</body>

</html>

Then your Python code becomes a bit simpler:

from lxml import html
from bs4 import BeautifulSoup as soup
import re
import codecs

#Access to the local URL(Html file)
with codecs.open("index.html", 'r') as f:
  page = f.read()

#html parsing
page_soup = soup(page,"html.parser")
tree = html.fromstring(page)#extract report name
ref = page_soup.find("h1").text.strip()
print("the name of the report is : " + ref + " \n")

chapters = page_soup.findAll('h2')
for index, chapter in enumerate(chapters):
    print(str(index+1) +"-" + str(chapter.text.strip()))
    sub_chapters = chapter.find_parent().find_all("h3")
    for index2, sub_chapter in enumerate(sub_chapters):
       print("\t" + str(index2+1) +"-" +str(sub_chapter.text.strip()))

I updated the page reading code a little and tried to use more idiomatic python in the updated script.

Also, note that:

sub_chapters = chapter.find_parent().find_all("h3")

find_all is relative to the parent of the chapter and not the entire document

Upvotes: 1

Ajax1234
Ajax1234

Reputation: 71471

You can use itertools.groupby after finding all the data associated with each chapter:

from itertools import groupby, count
import re
from bs4 import BeautifulSoup as soup
data = [[i.name, re.sub('\s+$', '', i.text)] for i in soup(content, 'html.parser').find_all(re.compile('h1|h2'), {'id':re.compile('^\d+$')})]
grouped, _count = [[a, list(b)] for a, b in groupby(data, key=lambda x:x[0] == 'h1')], count(1)
new_grouped = [[grouped[i][-1][0][-1], [c for _, c in grouped[i+1][-1]]] for i in range(0, len(grouped), 2)]
final_string = '\n'.join(f'{next(_count)}-{a}\n'+'\n'.join(f'\t{i}-{c}' for i, c in enumerate(b, 1)) for a, b in new_grouped)
print(final_string)

Output:

1-First Chapter
    1- First Sub-chapter of the first chapter
    2-Second Sub-chapter of the first chapter
2-Second Chapter
    1-First Sub-chapter of the Second chapter
    2-Second Sub-chapter of the Second chapter

Upvotes: 1

Related Questions