I'm having some trouble in understanding how jsonify works even though I went through the documentation. As you can see below, I'm calling the lookup() function which returns a dictionary, then I'm trying to jsonify it.
#app.route("/articles")
def articles():
a = lookup(33496)
return jsonify([link=a["link"], title = a["title"]]) #invalid syntax error
my helpers.py:
import feedparser
import urllib.parse
def lookup(geo):
"""Looks up articles for geo.""" #this function already parses the 'link' and 'title' form rss feed
# check cache for geo
if geo in lookup.cache:
return lookup.cache[geo]
# get feed from Google
feed = feedparser.parse("http://news.google.com/news?geo={}&output=rss".format(urllib.parse.quote(geo, safe="")))
# if no items in feed, get feed from Onion
if not feed["items"]:
feed = feedparser.parse("http://www.theonion.com/feeds/rss")
# cache results
lookup.cache[geo] = [{"link": item["link"], "title": item["title"]} for item in feed["items"]]
# return results
return lookup.cache[geo]
# initialize cache
lookup.cache = {}
The error that I'm getting is of invalid syntax. Any idea into what I'm doing wrong? Thanks
I think your dict syntax is wrong. You can read about more in official documentation.
The code that I think you are trying for is as follows:
#app.route("/articles")
def articles():
a = lookup(33496)
return jsonify({"link" : a["link"], "title" : a["title"]})
Specifically you should use curly braces instead of brackets ({}) and colon (:) instead of equals sign.
Another option is to let jsonify() to do the conversion (as pointed out in the other answer):
#app.route("/articles")
def articles():
a = lookup(33496)
return jsonify(link = a["link"], title = a["title"])
Nevertheless, I think you would be well advised to use create a dict. It becomes more flexible when you need to create larger JSON objects.
Hope this helps.
You dont need the square brackets, get rid of them.
return jsonify(link=a["link"], title=a["title"])
# ^At this point ^ and this one.
Read about keyword arguments in python.
Related
I am building grafana links in python with urllib like the following:
from urllib.parse import urlencode, urlunsplit
parameters = {
"parameter1":"value1",
"parameter2":"value2"
}
query = urlencode(
query = parameters,
doseq = True
)
link = urlunsplit((
"https",
"my_grafana.com",
"/graph",
query,
""
))
link will be in this case 'https://my_grafana.com/graph?parameter1=value1¶meter2=value2'. I now want to add parameters with no keyword for example "kiosk". The link should look like 'https://my_grafana.com/graph?parameter1=value1¶meter2=value2&kiosk&other_parameter'
As urlencode returns a string with the parameters I could manipulate the string like in the following example before I give it to urlunsplit:
no_keyword_parameters = ["kiosk","other_parameter"]
query = "&".join([query, *no_keyword_parameters])
I wonder if you can put parameters with and without keyword directly with urlencode together. I tried giving "kiosk" as a dictionary entry with None as content ({"kiosk": None}) but it includes the None in the url. Approaches, where I give a list of tuples instead of a dictionary for the parameters, were also unsuccessful.
Thank you for any help.
As mentioned by Ondrej, urlencode builds the query using k + '=' + v.
You could add non value parameters manually:
from urllib.parse import urlencode, urlunsplit, quote_plus
parameters = {"parameter1": "value1", "parameter2": "value2"}
no_value_parameters = ["kiosk", "other_parameter"]
no_value_parameters_quoted = [quote_plus(p) for p in no_value_parameters]
query = urlencode(query=parameters, doseq=True)
link = urlunsplit(("https", "my_grafana.com", "/graph", query, ""))
link = f"{link}&{'&'.join(no_value_parameters_quoted)}"
print(link)
Out:
https://my_grafana.com/graph?parameter1=value1¶meter2=value2&kiosk&other_parameter
What you've done seems sound and you could either do it like that or formalize it a bit more in your own encoding function, but urllib.parse.urlencode does not seem to understand the notion of parameters without value. If you look at the implementation (with doseq you get a variation of the same for the part relevant to your question):
for k, v in query:
...
l.append(k + '=' + v)
I.e. you have to have a key, value pair (to unpack two values) and whatever they are quoted to (that happens in the ellipses) will be a str joined over =. So even using custom qoute_via you cannot really change its function.
That linked implementation is the one provided with CPython, but also the documentation expects: key/value pairs, so that behavior really is as specified / documented:
The resulting string is a series of key=value pairs separated by '&' characters...
Studying Python, I am following an excellent Corey Schafer tutorial on Flask, he does this (I have extracted and summarized it for obvious reasons):
from folder_app import app # I did it to follow the structure and that the code is equal to the original
s = Serializer(app.config['SECRET_KEY'], 1800) # key, seconds
token = s.dumps({'user_id': 1}).decode('utf-8')
s = Serializer(app.config['SECRET_KEY'])
user_id = s.loads(token)['user_id'] # This is where I have the doubt
print(user_id)
print(type(s.loads(token)))
The code works, the problem I have is that although as you can see (s.loads (token)) is a dict, I expected to see something like this s.loads ({token ['user_id']}), or s.loads (token ['user_id']) or something like that. That is, it is a dict but it does not seem so. And my doubt goes in the sense if this comes from a greater concept of those they call "pythonic" (which I have not seen so far), or is something that only happens particularly as in this case. Incidentally, https://itsdangerous.palletsprojects.com/en/1.1.x/jws/ this appears: loads (self, s, salt = None, return_header = False) the arguments are in parentheses. I hope it is clear what my doubt is :)
I know this is not answer per say but just to add to my comment. This is an example of how the loads function works on dictionaries with the json module. https://docs.python.org/3/library/json.html#json.loads. What it does is take a json string and return the dictionary type object in Python. Your Serializer is doing something similar. It takes the token string and represents it as an object like dict
The s.dumps I am assuming is similar to json.dumps which gives you the json string representation of python dictionary.
import json
my_dict = json.loads('{"user_id": "Mane", "name": "Joe"}')
my_dict['user_id']
So you could just do json.loads('{"user_id": "Mane", "name": "Joe"}')['user_id'] which is just chaining the operations.
I am trying to create API for my Flask project. I have data stored in mongoDB and for building API I am using flask_restful. The problem is that in JSON are added escaped quotes and I cannot figure why and I rather have my JSON without them.
This is how my get function looks like:
from flask_restful import Resource
import json
from bson import json_util
class Harvests(Resource):
def get(self):
json_docs = []
for doc in db.collection.find():
json_doc = json.dumps(doc, default=json_util.default)
json_docs.append(json_doc)
return json_docs
In app.py it is just like that
api = Api(app)
api.add_resource(Harvests, '/api/harvests')
And I get JSON with escaped quotes (in browser or with curl)
[
"{\"_id\": {\"$oid\": \"5c05429cc4247917d66163a7\"},...
]
If I try this outside Flask (print JSON from mongo) and it works just fine. I tried use .replace(), but I think is not most elegant solution, but it did not work anyway. Any idea how I should get rid off these backslashes?
What you see is absolutely what you should expect to see according to your code, so I think there is a misunderstanding at some point. Let me explain what you are doing.
You convert each doc (a data structure) into a jsonified version (a string) of this data. Then you gather these strings in a list. Later you see this list, and of course you see a list of strings. Each of these strings contains a jsonified version of a data structure (a dictionary with opening braces, keys and values inside, and each key is a string itself with quotes, so these quotes are escaped within the jsonified string).
I recommend to collect your documents into a list and then convert that list to json instead:
def get(self):
docs = []
for doc in db.collection.find():
docs.append(doc)
return json.dumps(docs, default=json_util.default)
This way you get one json string representing the list of docs.
Maybe your framework is already applying a jsonifying automatically, in this case just don't do this step yourself:
return docs
Just use this instead.
So I'm trying to learn Python here, and would appreciate any help you guys could give me. I've written a bit of code that asks one of my favorite websites for some information, and the api call returns an answer in a dictionary. In this dictionary is a list. In that list is a dictionary. This seems crazy to me, but hell, I'm a newbie.
I'm trying to assign the answers to variables, but always get various error messages depending on how I write my {},[], or (). Regardless, I can't get it to work. How do I read this return? Thanks in advance.
{
"answer":
[{"widgets":16,
"widgets_available":16,
"widgets_missing":7,
"widget_flatprice":"156",
"widget_averages":15,
"widget_cost":125,
"widget_profit":"31",
"widget":"90.59"}],
"result":true
}
Edited because I put in the wrong sample code.
You need to show your code, but the de-facto way of doing this is by using the requests module, like this:
import requests
url = 'http://www.example.com/api/v1/something'
r = requests.get(url)
data = r.json() # converts the returned json into a Python dictionary
for item in data['answer']:
print(item['widgets'])
Assuming that you are not using the requests library (see Burhan's answer), you would use the json module like so:
data = '{"answer":
[{"widgets":16,
"widgets_available":16,
"widgets_missing":7,
"widget_flatprice":"156",
"widget_averages":15,
"widget_cost":125,
"widget_profit":"31",
"widget":"90.59"}],
"result":true}'
import json
data = json.loads(data)
# Now you can use it as you wish
data['answer'] # and so on...
First I will mention that to access a dictionary value you need to use ["key"] and not {}. see here an Python dictionary syntax.
Here is a step by step walkthrough on how to build and access a similar data structure:
First create the main dictionary:
t1 = {"a":0, "b":1}
you can access each element by:
t1["a"] # it'll return a 0
Now lets add the internal list:
t1["a"] = ["x",7,3.14]
and access it using:
t1["a"][2] # it'll return 3.14
Now creating the internal dictionary:
t1["a"][2] = {'w1':7,'w2':8,'w3':9}
And access:
t1["a"][2]['w3'] # it'll return 9
Hope it helped you.
I'm doing a program using Twitter API and MongoDB in 2.7 Python language.
I get a timeline and put it in a dictionary, which I want to store in a MongoDB database. To do this I have next code:
def saveOnBD(self, dic):
client = MongoClient("xxxx", "port")
db = client.DB_Tweets_User_Date
collection = db.tweets
collection.insert_many(dic)
I'm debbuging and dic it's not empty but I get next error:
TypeError: documents must be a non-empty list
How can I fix it?
I trying many options, but i solved that question changing the post method.
Instead of:
collection.insert_many(dic)
I used this:
collection.insert_one(dic)
I supose that, as I try to post only a variable(dic) and "insert_many()" is for many variables that retun me the error. That change solved me the question
you can either put in an entry before running the bulk entry function or use insert()
A list of documents must be passed to insert_many method
E.g.:
collection.insert_many([dic])