arxiv_rss_email_alert.py

#!/usr/bin/env python3

"""
The arXiv RSS (https://arxiv.org/help/rss) is separate from
the arXiv API (https://arxiv.org/help/api/index)

The API supports /query
The RSS provides XML for a given topic

https://arxiv.org/help/robots
"""


# python3 rss_filter_email.py

import xmltodict
import datetime
import json
import requests
import os
#import smtplib

with open('.env','r') as fil:
    api_key = fil.read().strip()

#print(api_key)

r = requests.get('https://export.arxiv.org/rss/quant-ph')
rss_as_rdf = r.text
rss_as_dict = xmltodict.parse(rss_as_rdf)

str_to_send = "search of arxiv RSS feed: \n\n"

if 'rdf:RDF' in rss_as_dict.keys():
    if 'item' in rss_as_dict['rdf:RDF'].keys():
        for entry_dict in rss_as_dict['rdf:RDF']['item']:
        #    for keyword_tuple in [('keyword1','keyword2')]:

            if 'title' in entry_dict.keys():
                if ((('keyword1'  in entry_dict['title'].lower()) and ('keyword2' in entry_dict['title'].lower())) or 
                     ('keyword3'    in entry_dict['title'].lower())):
                    str_to_send += ('found match for '+'keyword1 keyword2'+' in title for ' + entry_dict['link'] + "\n")
                    str_to_send += ("TITLE: " + entry_dict['title'] + '\n')
                    if 'description' in entry_dict.keys() and '#text' in entry_dict['description']:
                        str_to_send += ("ABSTRACT: " + entry_dict['description']['#text'] + '\n\n')
                    break
            else: 
                str_to_send += "'title' is not in RSS item"

            if 'description' in entry_dict.keys() and '#text' in entry_dict['description']:
                if ((('keyword1' in entry_dict['description']['#text'].lower()) and
                    ('keyword2' in entry_dict['description']['#text'].lower())) or 
                    ('keyword3' in entry_dict['description']['#text'].lower())):
                    str_to_send += ('found match for '+'keyword1 keyword2'+' in abstract for ' + entry_dict['link'] + "\n")
                    str_to_send += ("TITLE: " + entry_dict['title'] + '\n')
                    str_to_send += ("ABSTRACT: " + entry_dict['description']['#text'] + '\n\n')
                break
            else:
                str_to_send += "'description' or '#text' not found in RSS item"

            if 'dc:creator' in entry_dict.keys():
                list_of_authors = entry_dict['dc:creator'].replace('a>, <a','a>\n<a').split('\n')
                for author_url in list_of_authors:
                    for name_tuple in [('John','Doe'),
                                       ('John','Smith')]:
                        if (name_tuple[0] in author_url) and (name_tuple[1] in author_url):
                            str_to_send += ('found '+name_tuple[0]+' '+name_tuple[1]+' as author for ' + entry_dict['link'] + "\n")
                            str_to_send += ("TITLE: " + entry_dict['title'] + '\n')
                            str_to_send += ("ABSTRACT: " + entry_dict['description']['#text'] + '\n\n')
                            break
            else:
                str_to_send += "'dc:creator' not found in RSS item"
        else:
            str_to_send += "'item' not in RSS"
    else:
        str_to_send += "'rdf:RDF' not in RSS"

# DigitalOcean blocks port 25, so this won't work:
# # http://effbot.org/pyfaq/how-do-i-send-mail-from-a-python-script.htm
#server = smtplib.SMTP("localhost")
#server.sendmail("ben", ["ben@gmail.com"], "hello")

# I set up a SendMail account, so this works:
# curl --request POST   --url https://api.sendgrid.com/v3/mail/send   --header "Authorization: Bearer $SENDGRID_API_KEY"   --header 'Content-Type: application/json'   --data '{"personalizations": [{"to": [{"email": "ben@gmail.com"}]}],"from": {"email": "ben@gmail.com"},"subject": "Sending with SendGrid is Fun","content": [{"type": "text/plain", "value": "and easy to do anywhere, even with cURL"}]}'

# using the site https://curl.trillworks.com/#python
# I translated the above command into

headers = {
#    'Authorization': 'Bearer ' + os.environ.get('SENDGRID_API_KEY'),
  'Authorization': 'Bearer ' + api_key,
  'Content-Type': 'application/json',
}
#print(headers)

if len(str_to_send)<30:
#    str_to_send = "no matches to existing filters for today"
    subject = "no matches to report on "
else:
    subject = "keyword match for arxiv quant-ph on "

data_dict = {"personalizations":
               [{"to": [{"email": "ben@gmail.com"}, #}]}], 
                        {"email": "another@uni.edu"}]}],
               "from": {"email": "ben@gmail.com"},
               "subject": subject + str(datetime.date.today()),
               "content": [{"type": "text/plain", "value": str_to_send}]}
# send the Email as a POST

data_dict_no_email = {"personalizations":
               [{"to": [{"email": "ben@gmail.com"}]}], 
               "from": {"email": "ben@gmail.com"},
               "subject": subject + str(datetime.date.today()),
               "content": [{"type": "text/plain", "value": str_to_send}]}

#if not str_to_send=="no matches to existing filters for today":
#    response = requests.post('https://api.sendgrid.com/v3/mail/send',
#                             headers=headers,
#                             data=json.dumps(data_dict))
    # if the POST to Sendmail is successful, nothing is produced
#    if len(response.text)>0:
#        print(response.text)
#else:
response = requests.post('https://api.sendgrid.com/v3/mail/send',
                             headers=headers,
                             data=json.dumps(data_dict_no_email))