Python script to generate blog post filenames from titles

Posted by Saugata Chatterjee on February 05, 2021 · 8 mins read

Jekyll makes it easy to publish posts. The url for the post is automatically generated from the file name. The date and time is extracted too. We, however, need to provide a filename for the markdown file which becomes its URL. The following set of python commands will generate the filename from the post title and save us the trouble of typing the long filename. We are running the commands in an interpreter but this can be automated into a script.

Python 3.8.5 (default, Jul 28 2020, 12:59:40) 
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import re
>>> import os
>>> os.listdir()
['index.html', '.jekyll-cache', 'jekyll-theme-clean-blog.gemspec', '.idea', 'README.md', 'screenshot.png', 'gulpfile.js', 'about.md', '_sass', 'assets', '_config.yml', '_includes', '_layouts', '.gitignore', 'Gemfile.lock', 'posts', '_site', 'LICENSE', '.git', '_posts', 'bin', 'contact.html', 'img', 'Gemfile', 'package-lock.json', '.travis.yml', 'package.json']
>>> os.chdir('_posts')
>>> os.listdir()
['2020-01-26-dinosaurs.html', '2021-02-05-post1.md', '.ipynb_checkpoints', '2020-02-04-getting-attendance-from-ringcentral-chat-messages.md', 'renameposts.ipynb']
>>> os.listdir()[1]
'2021-02-05-post1.md'
>>> f = open(os.listdir()[1])
>>> while True:
...     line = f.readline()
...     title = re.findall('title: \"(.*?)\"', line)
...     if len(title) > 0:
...             print(re.sub(' ', '-', title[0].lower()))
...             break
... 
python-script-to-read-generate-blog-post-filenames-from-titles
>>> f.close()

Explanation

import os
os.chdir('_posts')

changes to the directory containing the posts.

os.listdir()

lists the file name so that we can select the correct markdown file containing the post.

f = open(os.listdir()[1])

opens the file. In this case we chose the file 2021-02-05-post1.md which doesn’t have proper name yet.

while True:

will keep looping till we find the lines containing the data for title and date and then it will break.

line = f.readline()

reads the next line into a variable line.

title = re.findall('title: \"(.*?)\"', line)

Here, we use the regex pattern match using the re library. re.findall() searches for the regular expression (regex) pattern 'title: \"(.*?)\"' and every time it finds title: "<your title>" it extracts <your title>. If re() doesn’t find anything, it returns an empty list.

if len(title) > 0:

if the list returned by re.findall() is empty then the line does not contain the title and we move on.

print(re.sub(' ', '-', title[0].lower()))

if the list returned by re.findall() is not empty, then we have the title in the first element of the title variable, title[0]. We first convert it to lowercase title[0].lower() and then substitute the spaces in the title with hyphens using re.sub(). re.sub() performs substitution of regex pattern using the format re.sub(pattern to replace, replace pattern, string). Once we found the title and extracted it, we break out of the loop.

The full script

The following script gets the full blog title from both the date and the title.

#!/bin/env python
import re
import sys

fname = sys.argv[1]
f = open(fname)

while True:
    line = f.readline()
    title = re.findall('title: \"(.*?)\"', line)
    if len(title) > 0:
        title = re.sub(' ', '-', title[0].lower())
        break

while True:
    line = f.readline()
    postdate = re.findall('date: (\d\d\d\d-\d\d-\d\d)', line)
    if len(postdate) > 0:
        postdate = re.sub(' ', '-', postdate[0].lower())
        break
    
fullfilename = postdate+'-'+title+'.md'
print(fullfilename)
f.close()

Running the script on our sample markdown file with the following content

---
layout: post
title: "Python script to read generate blog post filenames from titles"
date: 2020-02-05
background: '/img/posts/01.jpg'
---

will generate the following filename.

2020-02-05-python-script-to-read-generate-blog-post-filenames-from-titles.md

You can modify this script to accept the current filename as an argument (using argparse for example). We leave that as an exercise for the reader.

Next Post →