import json, re
import urllib.parse
from urllib.request import urlopen, Request
from selenium import webdriver
from selenium.webdriver.common.keys import Keys
from bs4 import BeautifulSoup
from selenium.webdriver.common.action_chains import ActionChains
from time import sleep

LIKES_THRESHOLD = 400
INSTAGRAM_EMAIL = "chase2332@gmail.com"
INSTAGRAM_PASSWORD = "MIqUfII7ADzF84Eb"
HASHTAG = "puravidabracelets" # Don't include the actual '#'.

def write(words):
    handle = open("test.html", "w")
    handle.write(words)
    handle.close()
    print("written")

def fetch(url):
    try:
        q = Request(url)
        q.add_header("User-Agent", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/55.0.2883.95 Safari/537.36")
        q.add_header("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8")
        response = urlopen(q, timeout=20)
        html = response.read()
        try:
            html = html.decode("utf-8")
        except:
            html = html.decode("latin1")
        return html
    except:
        sleep(.5)
    try:
        q = Request(url)
        q.add_header("User-Agent", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/55.0.2883.95 Safari/537.36")
        q.add_header("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8")
        response = urlopen(q, timeout=20)
        html = response.read()
        try:
            html = html.decode("utf-8")
        except:
            html = html.decode("latin1")
        return html
    except:
        print("Couldn't Load Page")

def strip_html(data):
    p = re.compile(r'<.*?>')
    return p.sub('', data)

def scroll_page(driver, delay):
    driver.execute_script("window.scrollTo(0, 0);")
    sleep(delay/2)
    driver.execute_script("window.scrollTo(0, document.body.scrollHeight);")
    sleep(delay)

def count_likes(driver, photo):
    hover = ActionChains(driver).move_to_element(photo)
    hover.perform()
    likes = photo.get_attribute('innerHTML')
    link = photo.get_attribute("href")
    likes = strip_html(likes)
    if "video" in likes.lower():
        return None
    likes = likes.split(">")[-1].split(";")[-1].split("like")[0].replace(",", "")
    try:
        likes_f = float(likes.replace("k", "").replace("m", ""))
    except:
        print(likes)
        return None
    if "k" in likes:
        likes = int(likes_f*1000)
    elif "m" in likes:
        likes = int(likes_f*1000000)
    else:
        likes = int(likes_f)
    if likes > LIKES_THRESHOLD:
        print("Likes Num", likes)
        return link
    return None

def remove_images(driver, photos):
    for photo in photos:
        link = photo.get_attribute("href")
        driver.execute_script("document.querySelectorAll(\"a[href='"+link.split(".com")[1]+"']\")[0].remove();")
    driver.execute_script('''var a = document.getElementsByClassName('_myci9');
        while(a.length > 20){
            a[0].remove();
        }''')

chromeOptions = webdriver.ChromeOptions()
prefs = {"profile.managed_default_content_settings.images":2}
chromeOptions.add_experimental_option("prefs",prefs)
driver = webdriver.Chrome(chrome_options=chromeOptions)

# driver = webdriver.Chrome()
driver.get("https://instagram.com/")
elem = driver.find_element_by_xpath("//a[text()='Log in']")
elem.click()

elem = driver.find_element_by_name("username")
elem.clear()
elem.send_keys(INSTAGRAM_EMAIL)
elem = driver.find_element_by_name("password")
elem.clear()
elem.send_keys(INSTAGRAM_PASSWORD)
elem.send_keys(Keys.RETURN)
sleep(6)
driver.get("https://www.instagram.com/explore/tags/"+HASHTAG+"/")
sleep(1)
elem = driver.find_element_by_xpath("//a[text()='Load more']")
elem.click()

photos = []

timeouts = 0
links = []
count = 0
reset = True
while timeouts < 20:
    driver.execute_script("window.scrollTo(0, document.body.scrollHeight);")
    photos = driver.find_elements_by_xpath("//a[contains(@href, '?tagged=')]")
    for photo in photos:
        timeouts = 0
        link = count_likes(driver, photo)
        if link:
            links.append(link)
    count += len(photos)
    remove_images(driver, photos)#remove thumbnails so things don't get too packed in the web page.
    print(str(count)+" total posts.  "+str(len(links))+" with more than "+str(LIKES_THRESHOLD)+" likes.")
    scroll_page(driver, .6)
    if timeouts == 18:
        driver.execute_script("window.scrollTo(0, document.body.scrollHeight/2);")
        scroll_page(driver, 8)
    timeouts += 1


write(json.dumps(links))
results = []
count = 0
for link in links:
    print(count, len(links))
    count += 1
    page = fetch(link)
    if not page:
        continue
    json_string = page.split("window._sharedData = ")[1].split(";</script>")[0]
    data = json.loads(json_string)
    username = data["entry_data"]["PostPage"][0]["media"]["owner"]["username"]
    link = "https://www.instagram.com/"+username+"/"
    if link in [x["link"] for x in results]:
        continue
    profile_pic_url = data["entry_data"]["PostPage"][0]["media"]["owner"]["profile_pic_url"]
    image_src = data["entry_data"]["PostPage"][0]["media"]["display_src"]
    page = fetch(link)
    if not page:
        results.append({
            "link":link,
            "profile_pic_url":"",
            "image_src":"",
            "ave_comments":"",
            "ave_likes":""
        })
        continue
    json_string = page.split("window._sharedData = ")[1].split(";</script>")[0]
    data = json.loads(json_string)
    posts = data["entry_data"]["ProfilePage"][0]["user"]["media"]["nodes"]
    posts = posts[:10]
    comments = 0
    likes = 0
    for post in posts:
        comments += post["comments"]["count"]
        likes += post["likes"]["count"]
    ave_comments = comments/len(posts)
    ave_likes = likes/len(posts)
    results.append({
        "link":link,
        "profile_pic_url":profile_pic_url,
        "image_src":image_src,
        "ave_comments":ave_comments,
        "ave_likes":ave_likes
    })

results = sorted(results, key=lambda k: k['ave_likes'], reverse=True)
result_string = "Average Likes (last 10 posts), Average # Comments (last 10 posts), Profile Link, Image Link, Profile Pic\n"
for row in results:
    result_string += str(row["ave_likes"])+","+str(row["ave_comments"])+","+row["link"]+","+row["image_src"]+","+row["profile_pic_url"]+"\n"

handle = open(HASHTAG+".csv", "w")
handle.write(result_string)
handle.close()


# elem.clear()
# elem.send_keys("madsongha+1@gmail.com")
# elem = driver.find_element_by_name("password")
# elem.clear()
# elem.send_keys("fullcontact")
# elem.send_keys(Keys.RETURN)