#!/usr/bin/env python
# coding: utf-8

# # Beyond the copyright cliff of death
# 
# Most of the newspaper articles on Trove were published before 1955, but there are some from the later period. Let's find out how many, and which newspapers they were published in.

# In[1]:


import requests
import pandas as pd
from IPython.display import display, FileLink


# In[2]:


trove_api_key = 'YOUR API KEY'


# ## Search for articles published after 1955
# 
# First we're going to run a date query to find all the articles published after 1954. But instead of looking at the articles themselves, we're going to get the `title` facet – this will tell us the number of articles for each newspaper.

# In[3]:


params = {
    'q': 'date:[1955 TO *]', # date range query
    'zone': 'newspaper',
    'facet': 'title', # get the newspaper facets
    'encoding': 'json',
    'n': 0, # no articles thanks
    'key': trove_api_key
}


# In[4]:


# Make our API request
response = requests.get('https://api.trove.nla.gov.au/v2/result', params=params)
data = response.json()


# In[5]:


# Get the facet data
facets = data['response']['zone'][0]['facets']['facet']['term']


# In[6]:


# Convert to a dataframe
df_articles = pd.DataFrame(facets)
# Get rid of some columns
df_articles = df_articles[['count', 'display']]
# Rename columns
df_articles.columns = ['number_of_articles', 'id']
# Change id to string, so we can merge on it later
df_articles['id'] = df_articles['id'].astype('str')
# Preview results
df_articles.head()


# ## Match the facets with newspapers
# 
# As you can see from the data above, the `title` facet only gives us the identifier for a newspaper, not its title or date range. To get more information about each newspaper, we're going to get a list of newspapers from the Trove API and then merge the two datasets.

# In[7]:


# Get ALL the newspapers
response = requests.get('https://api.trove.nla.gov.au/v2/newspaper/titles', params={'encoding': 'json', 'key': trove_api_key})
newspapers_data = response.json()


# In[8]:


newspapers = newspapers_data['response']['records']['newspaper']
# Convert to a dataframe
df_newspapers = pd.DataFrame(newspapers)


# In[9]:


# Merge the two dataframes by doing a left join on the 'id' column
df_newspapers_post54 = pd.merge(df_articles, df_newspapers, how='left', on='id')
df_newspapers_post54.head()


# ## Results

# In[10]:


# How many newspapers?
df_newspapers_post54.shape[0]


# In[11]:


# Reorder columns and save as CSV
df_newspapers_post54[['title', 'state', 'id', 'startDate', 'endDate', 'issn', 'number_of_articles', 'troveUrl']].to_csv('newspapers_post_54.csv', index=False)
# Display a link for easy download
display(FileLink('newspapers_post_54.csv'))


# ----
# 
# Created by [Tim Sherratt](https://timsherratt.org/) for the [GLAM Workbench](https://glam-workbench.github.io/).