Skip to content

Instantly share code, notes, and snippets.

View originalankur's full-sized avatar
🖥️
Coding underway.

Ankur Gupta originalankur

🖥️
Coding underway.
  • Bengaluru, India
View GitHub Profile
@originalankur
originalankur / spell_and_write.py
Created February 26, 2023 07:34
tokenise the text and sort it. Using it for ensuring that son can spell and write all the words.
import sys
import contextualSpellCheck
import spacy
def filter_words(text):
text = text.strip()
if text in nlp.Defaults.stop_words:
print("in stop word", text)
return False
<script src="https://cdn.safetycop.io/bundle.embed.1982642.min.js" crossorigin="anonymous"></script>
@originalankur
originalankur / feedfinder.py
Created September 3, 2016 21:40
feedfinder with newsblur bug fixes
"""feedfinder: Find the Web feed for a Web page
http://www.aaronsw.com/2002/feedfinder/
Usage:
feed(uri) - returns feed found for a URI
feeds(uri) - returns all feeds found for a URI
>>> import feedfinder
>>> feedfinder.feed('scripting.com')
'http://scripting.com/rss.xml'
#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
pip install networkx distance pattern
In Flipboard's article[1], they kindly divulge their interpretation
of the summarization technique called LexRank[2].
def normalize_line(line):
return [piece.strip() for piece in line.split("|")[1:-1]]
def is_valid_line(line):
return "|" in line
def load(text):
lines = map(normalize_line,
filter(is_valid_line,
text.strip().splitlines()))