Motivation
Obsidian is a great productivity tool that I use to keep my thoughts and ideas in place. One area I want to log and review are movies. When I sat down and listed all my favourite movies, it came down to a list of more than 100 movies! Trying to search through the cast and directors for all these movies is time consuming and I just want to get a quick summary of the movie before trying to write my own thoughts on them. So I decided to use Python to automate this for me.
How this works
- Input all the movies to search into the csv
movies_to_search.csv. - The script makes use of the
imdbpackage to download the metadata from IMDB. This does not require you to have an IMDB api account. You can check out their documentations on what data is available to adjust as required. I created a function that allows me to down the information regarding:- title
- plot
- genre
- country
- cover url
- director
- kind
- rating
- synopsis
- year
from imdb import IMDb
def movie_search(my_title):
"""
Given a string (movie title), find the metadata for that movie (or the closest
one matching that name).
Parameters
----------
my_title : string
returns a dictionary of metadata
"""
# create a dictionary to store data
information = dict()
# create an instance of the IMDb class
ia = IMDb()
# search for a movie
lookups = ia.search_movie(my_title)
# use the first search result to return the id
try:
# Test if any movies are found
movie_id = lookups[0].movieID
# get a movie
movie = ia.get_movie(movie_id)
# RETRIEVE DATA
# Get the movie title searched for
t = movie['title']
information['title'] = t
print(f'Retrieving data for {t}')
# Get the plot outline
try:
information['plot'] = movie['plot outline']
except:
print('plot outline not available')
try:
information['plot'] = movie['plot'][0]
except:
print('plot also not available')
information['plot'] = ''
# Get the genre
try:
information['genre'] = ','.join(movie['genre'])
except:
print('no information on genre')
information['genre'] = ''
# Get the country
try:
information['country'] = ','.join(movie['countries'])
except:
print('no information on country')
information['country'] = ''
# Get the director
# Extract the names in each Person object
try:
director_names = [d['name'] for d in movie['directors']]
information['director'] = ','.join(director_names)
except:
print('director information not available')
information['director'] = ''
# Get the cast
# Extract the top 5 names in cast (if it exists)
try:
cast_names = [d['name'] for d in movie['cast']]
if len(cast_names) >5:
cast_names = cast_names[0:5]
information['cast'] = ','.join(cast_names)
except:
print('No information on cast')
information['cast'] = ''
# Get the rating
try:
information['rating'] = movie['rating']
except:
print('rating info is not available')
information['rating'] = 'N/A'
# Get the year
try:
information['year'] = movie['year']
except:
print('no information on year')
information['year'] = ''
# Get the type
try:
information['kind'] = movie['kind']
except:
print('type not available')
information['kind'] = ''
# Get the cover url
try:
information['cover url'] = movie['cover url']
except:
print('cover url not available')
information['cover url'] = ''
# Get sypnosis
try:
information['synopsis'] = movie['synopsis'][0]
except:
print('no synopsis')
information['synopsis'] = ''
except:
print('NO RESULTS FOUND FOR {my_title}')
information['title'] = my_title
information['plot'] = ''
information['genre'] = ''
information['country'] = ''
information['cover url'] = ''
information['director'] = ''
information['kind'] = ''
information['rating'] = ''
information['synopsis'] = ''
information['year'] = ''
return(information)
Obsidian reads markdown pages so I automate the creation of markdown pages using python using mdutils. I create a function to automate the creation of a standard template filled in with the new metadata:
def create_markdown_page(myresults):
"""
Given a dataframe containing all the information about movies,
we can create individual pages
Parameters
----------
myresults : DataFrame
Contains essential headers: title, year, rating, genre, country, director
cast, synopsis and plot
"""
# Create iterations
for index, row in myresults.iterrows():
# Create a file
mdFile = MdUtils(file_name=row['title'])
# Create a metadata section
mdFile.new_line(text='---\n')
facts = ['year', 'rating','genre', 'country', 'director', 'cast']
for f in facts:
mdFile.new_line(text=f'{f.title()}: {row[f]}')
mdFile.new_line(text='Type: Review')
mdFile.new_line(text='')
mdFile.new_line(text='---\n')
# Create a title
mdFile.new_header(level=1, title=row['title'])
# Create a cover
# i could use new_inline_image but the text wraps at 20 characters
# creating a break. Hence, I code my own image line and unwrap it
cover_link = '!['+ row['title'] + '](' + row['cover url'] + ')'
mdFile.write(text = cover_link, wrap_width=0)
# Create a section for the plot
mdFile.new_header(level=1, title="Plot")
# Clean up plot text
plot_text = row['plot']
plot_text = ' '.join(plot_text.split())
plot_text = plot_text.replace('\ ','')
mdFile.new_line(text=plot_text, wrap_width=0)
# Create my own review section
mdFile.new_header(level=1, title="My own thoughts")
# Create markdown
mdFile.create_md_file()
You can drag and drop these newly created markdown sheets into Obsidian. For the full code - see my github