#!/usr/bin/env python3

from bs4 import BeautifulSoup
import os, argparse
from urllib.parse import urlparse, urlsplit
from urllib.request import urlopen
import json

def get_options():
  parser = argparse.ArgumentParser(description='AWS Documentation Downloader')
  parser.add_argument('-d','--documentation', help='Download the Documentation', action='store_true', required=False)
  parser.add_argument('-w','--whitepapers', help='Download White Papers', action='store_true', required=False)
  parser.add_argument('-f','--force', help='Overwrite old files', action='store_true', required=False)
  args = vars(parser.parse_args())
  return (args)

# Build a list of the amazon PDF's
def list_whitepaper_pdfs(start_page):
  html_page = urlopen(start_page)
  # Parse the HTML page
  soup = BeautifulSoup(html_page, 'html.parser')
  pdfs =  set()
  print("Generating PDF list (this may take some time)")
  for link in soup.findAll('a'):
    try:
      uri = link.get('href')
      print('URI: ', uri)
      # Allow whitepapers to be returned
      if "whitepapers" in start_page:
        if uri.endswith("pdf"):
          if "whitepapers" in uri or "enterprise-marketing" in uri:
            pdfs.add(uri)
    except:
     continue
  return pdfs


def find_pdfs_in_html(url):
  html_page_doc = urlopen(url)
  soup_doc = BeautifulSoup(html_page_doc, 'html.parser')
  # Get the A tag from the parsed page
  pdfs = set()
  for link in soup_doc.findAll('a'):
    try:
      sub_url = link.get('href')
      if sub_url.endswith("pdf"):
        pdfs.add(sub_url)
    except:
      continue
  return pdfs


def list_docs_pdfs(start_page):
  locale_path = "en_us/"
  base_url = "http://docs.aws.amazon.com"

  page = urlopen(start_page)
  soup = BeautifulSoup(page, "xml")
  pdfs =  set()
  print("Generating PDF list (this may take some time)")

  for link in soup.findAll('service'):
    try:
      uri = link.get('href')
      print('URI: ', uri)
      # if service uri is .html then parse as HTML
      if '.html' in uri:
        url = base_url + uri
        pdfs = pdfs.union(find_pdfs_in_html(url))
        continue

      # if service uri ends with "/" find and parse xml landing page
      if not uri.startswith('http'):
        url = base_url + uri.split("?")[0] + locale_path + "landing-page.xml"
      
      # Fetch the XML sub page (this is where the links to the pdf's live)
      sub_page_doc = urlopen(url)
      soup_doc = BeautifulSoup(sub_page_doc, 'xml')
      
      # Get the "tile" tag from the parsed page
      for sublink in soup_doc.findAll('tile'):
        try:
          sub_url = sublink.get('href')
          directory = base_url + "/".join(urlsplit(sub_url).path.split('/')[:-1])

          guide_info_url = directory + "/meta-inf/guide-info.json"
          print("Guide info url:", guide_info_url)
          guide_info_doc = urlopen(guide_info_url).read()
          guide_info = json.loads(guide_info_doc)

          if "pdf" in guide_info:
            pdf_url = directory + "/" + guide_info["pdf"]
            pdfs.add(pdf_url)
        except:
          continue
    except:
     continue
  return pdfs


def save_pdf(full_dir,filename,i):
  if not os.path.exists(full_dir):
    os.makedirs(full_dir)
  # Open the URL and retrieve data
  file_loc = full_dir + filename
  if not os.path.exists(file_loc) or force == True:
    if i.startswith("//"):
      i = "http:" + i
    print("Downloading : " + i)
    web = urlopen(i)
    print("Saving to : " + file_loc)
    # Save Data to disk
    output = open(file_loc,'wb')
    output.write(web.read())
    output.close()
  else:
    print("Skipping " + i + " - file exists or is a dated API document, use './getAWSdocs.py --force' to force override")


def get_pdfs(pdf_list, force):
  for i in pdf_list:
    doc = i.split('/')
    doc_location = doc[3]
    filename = urlsplit(i).path.split('/')[-1]
    # Set download dir for whitepapers
    if "whitepapers" in doc_location:
      full_dir = "whitepapers/"
    else:
      # Set download dir and sub directories for documentation
      full_dir = "documentation/"
      directory = urlsplit(i).path.split('/')[:-1]
      for path in directory:
        if path != "":
          full_dir = full_dir + path + "/"
    try:
      save_pdf(full_dir,filename,i)
    except:
      continue

# Main
args = get_options()
# allow user to overwrite files
force = args['force']
if args['documentation']:
  print("Downloading Docs")
  pdf_list = list_docs_pdfs("https://docs.aws.amazon.com/en_us/main-landing-page.xml")
  get_pdfs(pdf_list, force)

if args['whitepapers']:
  print("Downloading Whitepapaers")
  pdf_list = list_whitepaper_pdfs("http://aws.amazon.com/whitepapers/")
  get_pdfs(pdf_list, force)
  print("Downloading SAP Whitepapaers")
  pdf_list = list_whitepaper_pdfs("https://aws.amazon.com/sap/whitepapers/")
  get_pdfs(pdf_list, force)

for p in pdf_list:
  print(p)