API get only assignments that are not submitted by current student

[BenediktMoore]

i am trying to build a todo list app for personal use. however it returns assignments that i've already submitted and i can't find how i need to filter them

import requests
from datetime import datetime, timedelta
# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("summarization", model="Falconsai/text_summarization")
import re
# as per recommendation from @freylis, compile once only
CLEANR = re.compile('<.*?>|&([a-z0-9]+|#[0-9]{1,6}|#x[0-9a-f]{1,6});')

def cleanhtml(raw_html):
  cleantext = re.sub(CLEANR, '', raw_html)
  return cleantext
def sumtxt(input):
    if (len(input) > 512):
        #print('\n'+input+'\n')
        input = input.strip().replace("\n","")
        if (len(input) > 512):
            input = input[:510]
    return pipe(input,max_length=20,min_length=10,do_sample=False)[0]['summary_text']

def fetch_courses(token):
    url = "https://iusd.instructure.com/api/v1/courses?per_page=100"
    headers = {
        'Authorization': f'Bearer {token}'
    }
    response = requests.get(url, headers=headers)
    response.raise_for_status()  # Raise an error for bad responses
    return response.json()

def fetch_assignments(token, course_id):
    url = f"https://iusd.instructure.com/api/v1/courses/{course_id}/assignments?per_page=100?bucket=upcoming?include[]=submission"
   # url2 = f"https://iusd.instructure.com/api/v1/courses/{course_id}/students/submissions?per_page=100"
    
    headers = {
        'Authorization': f'Bearer {token}'
    }
    response = requests.get(url, headers=headers)
    #response2 = requests.get(url2, headers=headers)
    try:
        response.raise_for_status()  # Raise an error for bad responses
       #print(response.json())
        #print(response2.json())
        print(response.json())
        return response.json()
    except:
        return ""
import requests


def filter_and_sort_assignments(assignments):
    now = datetime.now()
    two_weeks_from_now = now + timedelta(weeks=2)
    two_weeks_from_now.replace(tzinfo=None)

    filtered_assignments = [
        assignment for assignment in assignments
        
        if assignment['due_at'] and datetime.fromisoformat(assignment['due_at'].replace('Z', '+00:00')).replace(tzinfo=None) <= two_weeks_from_now and not ('submission' in assignment )
    ]#or assignment['graded_submissions_exist']

    # Sort by due date
    filtered_assignments.sort(key=lambda x: datetime.fromisoformat(x['due_at'].replace('Z', '+00:00')))
    return filtered_assignments

def main():
    token = "SOMEEXAMPLETOKEN"
    
    try:
        courses = fetch_courses(token)
        all_assignments = []

        for course in courses:
            assignments = fetch_assignments(token, course['id'])
            all_assignments.extend(assignments)

        filtered_assignments = filter_and_sort_assignments(all_assignments)

        print("\nTodo List of Assignments:")
        filtered_assignments=[]
        for assignment in filtered_assignments:
            due_date = datetime.fromisoformat(assignment['due_at'].replace('Z', '+00:00')).strftime('%Y-%m-%d')
            print(assignment)
            print(f"{assignment['name']} - Due: {due_date} - Points: {assignment.get('points', 'N/A')} - Summary: {sumtxt(cleanhtml(assignment.get('description', 'no description available'))) + '...'}")

    except requests.exceptions.RequestException as e:
        print(f"An error occurred: {e}")

if __name__ == "__main__":
    #messages = [
    #{"role": "user", "content": "Who are you?"},
    #]
    #pipe = pipeline("text-generation", model="Qwen/Qwen2.5-1.5B-Instruct")
    #print(pipe(messages))
    main()

[BenediktMoore · Accepted solution]

for anyone else the solution is to get the submissions and then filter assignments with same id as submissions like this

import requests
from datetime import datetime, timedelta
import time
# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("summarization", model="Falconsai/text_summarization")
import re
# as per recommendation from @freylis, compile once only
CLEANR = re.compile('<.*?>|&([a-z0-9]+|#[0-9]{1,6}|#x[0-9a-f]{1,6});')

def cleanhtml(raw_html):
  cleantext = re.sub(CLEANR, '', raw_html)
  return cleantext
def sumtxt(input):
    if (len(input) > 512):
        #print('\n'+input+'\n')
        input = input.strip().replace("\n","")
        if (len(input) > 512):
            input = input[:510]
    return pipe(input,max_length=20,min_length=10,do_sample=False)[0]['summary_text']

def fetch_courses(token):
    url = "https://iusd.instructure.com/api/v1/courses?per_page=100"
    headers = {
        'Authorization': f'Bearer {token}'
    }
    response = requests.get(url, headers=headers)
    response.raise_for_status()  # Raise an error for bad responses
    return response.json()

def fetch_assignments(token, course_id):
    url = f"https://iusd.instructure.com/api/v1/courses/{course_id}/assignments?per_page=100?bucket=upcoming?include[]=submission"
    url2 = f"https://iusd.instructure.com/api/v1/courses/{course_id}/students/submissions?per_page=100"
    
    headers = {
        'Authorization': f'Bearer {token}'
    }
    response = requests.get(url, headers=headers)
    response2 = requests.get(url2, headers=headers)
    try:
        response.raise_for_status()  # Raise an error for bad responses
        print(response.json())
        print(response2.json())

        return [response.json(),response2.json()]
    except:
        return [[],[]]
import requests


def filter_and_sort_assignments(assignments,submissions):
    now = datetime.now()
    two_weeks_from_now = now + timedelta(weeks=2)


    filtered_assignments = []

    for assignment in assignments:
        if assignment['due_at']:
            due_date = datetime.fromisoformat(assignment['due_at'].replace('Z', '+00:00')).replace(tzinfo=None)
            ct=True
            for i in submissions:
                if (i.get('assignment_id') == assignment.get('id')):
                    if (i.get('workflow_state') == 'unsubmitted'):
                        ct=True
                    else:
                        ct=False
                
            if due_date <= two_weeks_from_now and ct:
                filtered_assignments.append(assignment)
    # Sort by due date
    filtered_assignments.sort(key=lambda x: datetime.fromisoformat(x['due_at'].replace('Z', '+00:00')))
    return filtered_assignments

def main():
    token = "SOMEEXAMPLETOKEN"
    
    try:
        courses = fetch_courses(token)
        all_assignments = []
        all_submissions = []

        for course in courses:
            assignments = fetch_assignments(token, course['id'])
            all_assignments.extend(assignments[0])
            all_submissions.extend(assignments[1])

        filtered_assignments = filter_and_sort_assignments(all_assignments,all_submissions)

        print("\nTodo List of Assignments:")
        for assignment in filtered_assignments:
            due_date = datetime.fromisoformat(assignment['due_at'].replace('Z', '+00:00')).strftime('%Y-%m-%d')
            print(assignment)
            print(f"{assignment['name']} - Due: {due_date} - Points: {assignment.get('points', 'N/A')} - Summary: {sumtxt(cleanhtml(assignment.get('description', 'no description available'))) + '...'}")

    except requests.exceptions.RequestException as e:
        print(f"An error occurred: {e}")

if __name__ == "__main__":
    #messages = [
    #{"role": "user", "content": "Who are you?"},
    #]
    #pipe = pipeline("text-generation", model="Qwen/Qwen2.5-1.5B-Instruct")
    #print(pipe(messages))
    main()

[chriscas]

Hi @BenediktMoore,

I know you have found yourself a solution that works, but it may be overcomplicating things.  If you do "https://iusd.instructure.com/api/v1/courses/{course_id}/assignments?per_page=100&bucket=unsubmitted&order_by=due_at" for each course, you could then use list comprehension or other methods on the python side to filter that down to the desired period.  This would have the added benefit of including assignments that may have already been due and you still haven't submitted yet.  I think you may not have been getting the results you expected with your url because you used '?' as a separater for all of your parameters.  You should only use the '?' before your first parameter, then '&' before each additional parameter.

Speaking more generally though, if you plan to use this sort of thing instead of checking in to your courses on the web or mobile apps, you're going to need to do a lot of testing to find all the edge cases and make sure you account for all of them.  Your teachers probably aren't going to accept "the assignment didn't appear on my own custom list" as an excuse for being late or missing work.  Just off the top of my head as a Canvas admin I can think of a couple:

1. Assignments with no due date...  A lot of teachers have assignments like this with no specific due date, but they'd all need to be done by the end of the course.  Maybe you need to substitute the course or term end date as the due date for these type of assignments on the python side so they'd show up.
2. Pagination...  You are currently making one API call with a per_page of 100.  That will probably catch enough assignments from each course, depending how they are sorted, but is courses with a large number of assignments, that will not get a complete list.

Aside from the tech issues mentioned above, I see that you appear to be trying to summarize text with this code.  I don't want to put up too many roadblocks, but you may want to ask your school/institution whether this is okay or not.  There are differing thoughts and policies around the world on the subject of copyright and material use and ownership.  Your school may be okay with you using course content this way, they may leave to to your teacher, or they may have a strict policy forbidding use of course material in this way.  Sometimes it may work to do things and ask for forgiveness later if what you did wasn't in line with school policy, but this is one where I'd personally recommend seeking permission first so you don't face harsh consequences later.

I hope this info helps!

-Chris

[dbrace]

Hi @BenediktMoore,

Something else to consider based on the few edge cases that @chriscas provided. Depending on the institution, course, or teacher, more of your coursework might exist in Canvas but only as a "pointer" or "link" to a third-party resource such as textbook publisher and its content. Dates for those types of items may not exist in Canvas and may only exist on the third-party resource.

-Doug