Get tumblr posts by tag

output_data.csv

tumblscrape.py

import pytumblr
import os
import time
from datetime import datetime
import csv


client = pytumblr.TumblrRestClient(
  '..................................................',
  '..................................................',
  '..................................................',
  '..................................................'
)

posts = []
max_calls = 200

last_timestamp = int(time.time())
for _ in range(max_calls):
    posts.extend(client.tagged('goals', limit=20, filter='text', before=last_timestamp))
    last_timestamp = posts[-1]['timestamp']


# massage the data into a consistent dictionary
content_key_map = {'text': 'body', 'photo': 'caption', 'quote': 'text', 'link': 'description', 'chat': 'body',
           'audio': 'caption', 'video': 'caption', 'answer': 'answer'}
output_list = []
add_this_item = False
or_tags = ['food', 'fit', 'weight', 'health', 'life']

for p in posts:
    for t in or_tags:
        if t in p['tags']:
            output_list.append(
                {
                    'content': p[content_key_map[p['type']]],
                    'date': datetime.utcfromtimestamp(int(p['timestamp'])).strftime('%Y-%m-%d'),
                    'time': datetime.utcfromtimestamp(int(p['timestamp'])).strftime('%H:%M:%S'),
                    'tags': p['tags'],
                    'id': p['id'],
                    'blog_name': p['blog_name'],
                    'post_url': p['post_url'],
                    'type': p['type']
                }
            )
            break


# write the output to csv

with open('output_data.csv', 'w', newline='') as csvfile:
    fieldnames = ['content', 'date', 'time', 'tags', 'id', 'blog_name', 'post_url', 'type']
    writer = csv.DictWriter(csvfile, fieldnames=fieldnames)

    writer.writeheader()
    for p in output_list:
        writer.writerow(p)