mining sentences to jpdb using yomichan via ankiconnect

ankiconnect.patch

save this to the folder for the ankiconnect plugin — you can 
get that by hitting "view files" in the list of addons — and
run

patch __init__.py < ankiconnnect.patch

--- __init__.py 2022-02-11 03:47:57.262335097 +0530
+++ __init__.py 2022-02-10 22:06:13.383091229 +0530
@@ -667,6 +667,14 @@

     @util.api()
     def addNote(self, note):
+        word = note['fields']['front']
+        reading = note['fields']['Reading']
+        sentence = note['fields']['Sentence']
+
+        r = requests.get(f'http://127.0.0.1:5000/add_mined_sentence/{word}/{reading}/{sentence}')
+
+        print(r.status_code)
+
         ankiNote = self.createNote(note)

         self.addMediaFromNote(ankiNote, note)

mining-server.py

# deps: requests furl beautifulsoup4 flask
# run with
#   $ FLASK_APP=mining-server flask run

import requests
from bs4 import BeautifulSoup as Soup
import http
import pathlib
import furl
import urllib
import re
from flask import Flask
app = Flask(__name__)

DOMAIN = "https://jpdb.io"

cookies = pathlib.Path('./jpdb-cookie.txt')
jar = http.cookiejar.MozillaCookieJar(cookies);
jar.load()

sess = requests.Session()
sess.cookies = jar # or maybe s.cookies.update(jar)

# folding the ability to drop a url param into this function
# this allows us to make the urls printed to console look prettier
def make_abs(rel_link, drop=None):
  url = furl.furl(rel_link)
  if drop is not None:
      if drop in url.args:
          del url.args[drop]
  return urllib.parse.urljoin(DOMAIN, url.url)

@app.route("/add_mined_sentence/<vocab>/<reading>/<sentence>")
def add_mined_sentence(vocab, reading, sentence):
    # search for the vocab
    search_page = sess.get(f'https://jpdb.io/search?q={vocab}&lang=english#a')
    search_soup = Soup(search_page.content, "html.parser")

    # find a link to the page with the right reading
    vocab_page_link_rel = search_soup.find("a", href=re.compile(f"{vocab}\/{reading}\?"))['href']
    vocab_page_link_abs = make_abs(vocab_page_link_rel, drop='expand')
    vocab_page = sess.get(vocab_page_link_abs)
    vocab_soup = Soup(vocab_page.content, "html.parser")

    # grab the "Edit sentence" link
    # strip the origin param, since it doesn't really do anything for us
    edit_sentence_link_rel = vocab_soup.find("a", href=re.compile("\/edit-shown-sentence"))['href']
    edit_sentence_link_abs = make_abs(edit_sentence_link_rel, drop='origin')

    # then POST to it with the sentence we want
    payload = { 'sentence': sentence, 'translation': '' }
    resp = sess.post(edit_sentence_link_abs, data=payload)

    # add to a deck (i'm using deck id 20)
    params = urllib.parse.parse_qs(urllib.parse.urlparse(edit_sentence_link_abs).query)
    id_vocab = params['v'][0]
    id_spelling = params['s'][0]
    id_reading = params['r'][0]
    origin = f'/search?q={vocab}&lang=english'
    add_url_abs = make_abs("/deck/20/add")
    payload = { 'v': id_vocab, 's': id_spelling, 'origin': origin }
    resp = sess.post(add_url_abs, data=payload)

    print()
    print(f'{vocab} with reading {reading} added to deck #20')
    print(f'  vocab: {urllib.parse.unquote(vocab_page_link_abs)}')
    print(f'  sentence: {edit_sentence_link_abs}')
    print()

    return "ok"