#!/usr/bin/env python3

"""

Simple requests-based (non-cached) interface

"""

import requests

from ansicolor import black

import json

from collections import defaultdict

from multiprocessing import Pool

def getExercises():

return requests.get("https://www.khanacademy.org/api/v1/exercises").json()

def getExerciseVideos(name):

return requests.get("https://www.khanacademy.org/api/v1/exercises/{0}/videos".format(name)).json()

def hasYTTimedText(v, lang="de"):

url = "https://www.youtube.com/api/timedtext?caps=asr&key=yttt1&v=%s&lang=%s&fmt=srv2" % (v, lang)

response = requests.get(url)

return bool(response.text) # Empty if not available

def fetchVideos(lang):

return json["nodes"]["videos"]

def findDubbedVideos(lang):

"Find only videos which are dubbed"

videos = fetchVideos(lang)

return [k for k, v in videos.items() if v["dubbed"]]

def fetchOriginalVideoURL(vid):

return response["url"].partition("&")[0].replace("http:", "https:")

timedtextRE = re.compile(r"^https?://www\.youtube\.com/timedtext_video\?v=(.*)$")

def getTranslatedVideoId(vid, lang):

else: return None

def fetchVideoMap(pool, lang):

if url is not None}

if __name__ == "__main__":

#Download exercise list

print(black("Downloading master exercise list...", bold=True))

exercises = getExercises()

#Download videos

print(black("Downloading exercise video list for {0} exercises..."

.format(len(exercises)), bold=True))

pool = Pool(32)

exVideos = pool.map(getExerciseVideos, [e["name"] for e in exercises])

with open("fo.json","w") as outf:json.dump(exVideos, outf)

# Perform mapping

print(black("Mapping videos...", bold=True))

result = []

allVideoIDs = set() # Need that for subtitle mapping

for exercise, videos in zip(exercises, exVideos):

current_videos = []

for video in videos:

current_videos.append({

"title": video["title"],

"youtube_id": video["youtube_id"],

"duration": video["duration"],

})

allVideoIDs.add(video["youtube_id"])

exercise = {

"videos": current_videos,

"id": exercise["name"],

"title": exercise["title"],

"url": exercise["ka_url"],

}

result.append(exercise)

allVideoIDs = list(allVideoIDs) # Need guaranteed order

# Fetch subtitles for videos (reuse pool)

print(black("Checking subtitles for {0} videos..."

.format(len(allVideoIDs)), bold=True))

hasSubtitles = pool.map(hasYTTimedText, allVideoIDs)

hasSubtitleMap = {

v: hasSubs for v, hasSubs in zip(allVideoIDs, hasSubtitles)}

# Injects

print(black("Remapping subtitle information to video...", bold=True))

for exercise in result:

for video in exercise["videos"]:

video["has_subtitles"] = hasSubtitleMap[video["youtube_id"]]

# Write to output file

print(black("Writing to videos.json...", bold=True))

with open("videos.json", "w") as outfile:

json.dump(result, outfile)