Can langextract be used to extract desired text from html?

[2020-abhilash]

I am trying to use langextract to extract useful information that I needed from a web page in wikipedia where the content itself was given to it as a URL which I am assuming would be html. So the question is that if langextract is capable of extracting useful text information from html content?

I get the error like below when I provide a wikipedia URL to extract place name entity from it.

raise ResolverParsingError("Content must contain an 'extractions' key.")
langextract.resolver.ResolverParsingError: Content must contain an 'extractions' key.

[KikeVen]

If you're working with HTML files, you can use a library like BeautifulSoup to extract the text content, then pass it to LangExtract. Here's an example:

from bs4 import BeautifulSoup
import requests
import langextract as lx

# Extract text from HTML
url = "https://example.com/some-html-file"
response = requests.get(url)
soup = BeautifulSoup(response.content, 'html.parser')
text_content = soup.get_text()

# Feed the extracted text to LangExtract
result = lx.extract(
    text_or_documents=text_content,
    prompt_description="Extract relevant details",
    examples=[],
    model_id="gemini-2.5-flash"
)

If you expect to work with different types of document formats, you can use a library like MarkItDown

from markitdown import MarkItDown
import re


def read_document(file_path):
    md = MarkItDown()
    result = md.convert(file_path)
    return result.text_content


def simple_md_to_text(md_content):
    # Remove headers
    text = re.sub(r'^#+\s+', '', md_content, flags=re.MULTILINE)
    # Remove bold/italic
    text = re.sub(r'\*\*(.+?)\*\*', r'\1', text)
    text = re.sub(r'\*(.+?)\*', r'\1', text)
    # Remove links
    text = re.sub(r'\[(.+?)\]\(.+?\)', r'\1', text)
    return text

my_file = "/path/to/file.html"
doc_loader = read_document(my_file)
plain_text = simple_md_to_text(doc_loader)

# Feed the extracted text to LangExtract
result = lx.extract(
    text_or_documents=plain_text,
    prompt_description="Extract relevant details",
    examples=[],
    model_id="gemini-2.5-flash"
)

In short:

• Yes: url_text = "https://www.gutenberg.org/files/1513/1513-0.txt"
• No: url_text = "https://www.gutenberg.org/files/1513/1513-0.html"

[2020-abhilash]

Thank you. I will try this out to see if it works.

[aksg87]

Hello, I think this was due to a subtle bug that was just fixed. @2020-abhilash please let me know if the merge fix resolves your issue.