inscriptis

2.7.1 · active · verified Sat Apr 11

Inscriptis is a Python-based HTML to text conversion library, command line client, and Web service (v2.7.1). It specializes in providing high-quality, layout-aware text representations of HTML content, including support for nested tables and a subset of CSS, and offers optional annotated output. The library is actively maintained with regular releases addressing new Python versions and feature enhancements.

Warnings

breaking The `XmlAnnotationProcessor` (introduced in 2.6.0) now requires a mandatory root element. The generated XML will contain a `<content>` root element by default. If you were using this processor directly, your XML output structure will change.
Fix: If using `XmlAnnotationProcessor`, be aware of the new `<content>` root element. The name can be overwritten by providing the `root_element` parameter to the processor call.
deprecated Support for Python 3.9 has been removed as of version 2.7.0. Python 3.8 support was deprecated in 2.5.1 and subsequently removed.
Fix: Upgrade your Python environment to version 3.10 or newer (up to <3.15) to maintain compatibility with `inscriptis`.
gotcha When processing very complex HTML pages, `inscriptis` (which uses `lxml` internally) may exhibit increased memory consumption due to `lxml`'s tendency to reuse memory rather than releasing it back to the operating system.
Fix: For long-running services processing many complex HTML documents, monitor memory usage and consider restarting processes periodically or optimizing the HTML input where possible. This is a characteristic of `lxml` rather than a direct `inscriptis` bug.

Install

pip install inscriptis Core library
pip install inscriptis[web-service] With Web service (FastAPI/Uvicorn)

Imports

get_text
```
from inscriptis import get_text
```

Quickstart

Convert HTML from a URL to plain text, preserving layout and structure. The example fetches content from 'https://www.informationscience.ch' and prints its text representation.

import urllib.request
from inscriptis import get_text

url = "https://www.informationscience.ch"
try:
    with urllib.request.urlopen(url) as response:
        html_content = response.read().decode('utf-8')
except Exception as e:
    html_content = f"<html><body><p>Error fetching URL: {e}</p></body></html>"

text = get_text(html_content)
print(text)

view raw JSON →