I have been following this tutorial: https://kb.objectrocket.com/postgresql/scrape-a-website-to-postgres-with-python-938
My app.py file looks like this (taken from the above tutorial):
from flask import Flask # needed for flask-dependent libraries below
from flask import render_template # to render the error page
from selenium import webdriver # to grab source from URL
from bs4 import BeautifulSoup # for searching through HTML
import psycopg2 # for database access
# set up Postgres database connection and cursor.
t_host = "localhost" # either "localhost", a domain name, or an IP address.
t_port = "5432" # default postgres port
t_dbname = "scrape"
t_user = "postgres"
t_pw = "********"
db_conn = psycopg2.connect(host=t_host, port=t_port, dbname=t_dbname, user=t_user, password=t_pw)
db_cursor = db_conn.cursor()
app = Flask(__name__)
@app.route("/")
@app.route('/import_temp')
def import_temp():
# set up your webdriver to use Chrome web browser
my_web_driver = webdriver.Chrome("/usr/local/bin/chromedriver")
# designate the URL we want to scrape
# NOTE: the long string of characters at the end of this URL below is a clue that
# maybe this page is so dynamic, like maybe refers to a specific web session and/or day/time,
# that we can't necessarily count on it to be the same more than one time.
# Which means... we may want to find another source for our data; one that is more
# dependable. That said, whatever URL you use, the methodology in this lesson stands.
t_url = "https://weather.com/weather/today/l/7ebb344012f0c5ff88820d763da89ed94306a86c770fda50c983bf01a0f55c0d"
# initiate scrape of website page data
my_web_driver.get(t_url)
# return entire page into "t_content"
t_content = my_web_driver.page_source
# use soup to make page content easily searchable
soup_in_bowl = BeautifulSoup(t_content, 'html.parser')
# search for the UNIQUE span and class for the data we are looking for:
o_temp = soup_in_bowl.find('span', attrs={'class': 'deg-feels'})
# from the resulting object, "o_temp", get the text parameter and assign it to "n_temp"
n_temp = o_temp.text
# Build SQL for purpose of:
# saving the temperature data to a new row
s = ""
s = "INSERT INTO tbl_temperatures"
s = "("
s = "n_temp"
s = ") VALUES ("
s = "(%n_temp)"
s = ")"
# Trap errors for opening the file
try:
db_cursor.execute(s, [n_temp, n_temp])
db_conn.commit()
except psycopg2.Error as e:
t_msg = "Database error: " e "/n open() SQL: " s
return render_template("error_page.html", t_msg = t_msg)
# Success!
# Show a message to user.
t_msg = "Successful scrape!"
return render_template("progress.html", t_msg = t_msg)
# Clean up the cursor and connection objects
db_cursor.close()
db_conn.close()
I have successfully imported the required libraries, and created a database to store the scraped data. However, when I run the application I now receive an error:
FLASK_APP = app.py
FLASK_ENV = development
FLASK_DEBUG = 0
In folder /home/lloyd/PycharmProjects/flaskProject
/home/lloyd/PycharmProjects/flaskProject/venv/bin/python -m flask run
* Serving Flask app 'app.py' (lazy loading)
* Environment: development
* Debug mode: off
* Running on http://127.0.0.1:5000/ (Press CTRL C to quit)
[2021-12-29 20:51:32,817] ERROR in app: Exception on / [GET]
Traceback (most recent call last):
File "/home/lloyd/PycharmProjects/flaskProject/venv/lib/python3.6/site-packages/flask/app.py", line 2073, in wsgi_app
response = self.full_dispatch_request()
File "/home/lloyd/PycharmProjects/flaskProject/venv/lib/python3.6/site-packages/flask/app.py", line 1518, in full_dispatch_request
rv = self.handle_user_exception(e)
File "/home/lloyd/PycharmProjects/flaskProject/venv/lib/python3.6/site-packages/flask/app.py", line 1516, in full_dispatch_request
rv = self.dispatch_request()
File "/home/lloyd/PycharmProjects/flaskProject/venv/lib/python3.6/site-packages/flask/app.py", line 1502, in dispatch_request
return self.ensure_sync(self.view_functions[rule.endpoint])(**req.view_args)
File "/home/lloyd/PycharmProjects/flaskProject/app.py", line 41, in import_temp
n_temp = o_temp.text
AttributeError: 'NoneType' object has no attribute 'text'
127.0.0.1 - - [29/Dec/2021 20:51:32] "GET / HTTP/1.1" 500 -
I think the error suggests the data passed into the BeautifulSoup method is incorrect: Line 41, in import_temp n_temp = o_temp.text
If anyone is able to tell me the cause of this exception, and how to resolve it, it would be really appreciated.
Thanks
CodePudding user response:
You need to look at the website structure in the browser dev tools. Here's how it looks on this particular site:
<div
data-testid="FeelsLikeSection"
>
<span
data-testid="TemperatureValue"
>54°</span
><span
data-testid="FeelsLikeLabel"
>Feels Like</span
>
</div>
The <span>
with real feel temperature has a unique class name, but this class name is auto-generated (probably by something like Styled Components) and will likely change each time the site is rebuilt, so you should not rely on that.
The enclosing <div>
on the other hand can be distinguished by the data-testid
attribute, which is put there by the site developers for their automated tests. This is unlikely to change, as it will break their tests.
The <span>
with the temperature also has the data-testid
attribute, but on this site there are many <span>
s with the same attribute value, so a query by {'data-testid': 'TemperatureValue'}
will not give you a unique element.
So you can do a two-level query like this:
o_temp_feel = soup_in_bowl.find('div', attrs={'data-testid': 'FeelsLikeSection'})
o_temp = o_temp_feel.find('span', attrs={'data-testid': 'TemperatureValue'})
# from the resulting object, "o_temp", get the text parameter and assign it to "n_temp"
n_temp = o_temp.text
Note that the n_temp
will be something like "54°", so if you need the numeric value, you need to truncate it before converting to an integer.