update repo to work with latest checkvisaslot screenshots

.gitignore

@@ -142,3 +142,5 @@ dmypy.json
 # Cython debug symbols
 cython_debug/
 
+.history
+.space

Pipfile

@@ -0,0 +1,21 @@
+[[source]]
+url = "https://pypi.org/simple"
+verify_ssl = true
+name = "pypi"
+
+[packages]
+opencv-python-headless = "~=4.5.5.64"
+numpy = "~=1.21.6"
+pytesseract = "~=0.3.9"
+requests = "~=2.27.1"
+python-dotenv = "~=0.20.0"
+python-telegram-bot = "~=13.11"
+python-dateutil = "*"
+datefinder = "*"
+dateparser = "*"
+thefuzz = {extras = ["speedup"], version = "*"}
+
+[dev-packages]
+
+[requires]
+python_version = "3.8"

README.md

@@ -30,15 +30,18 @@ DevTools Networks tab + Python + Tesseract + 🧠 = 💣
 
 ## Usage 👨‍💻
 
-- Clone and create heroku app linked to this.
-- Add [apt buildpack](https://elements.heroku.com/buildpacks/heroku/heroku-buildpack-apt) (we need to install `libgl1`
-  to make cv2 work).
-- `heroku config:set TESSDATA_PREFIX=/app/.apt/usr/share/tesseract-ocr/tessdata`
+- Clone repo
+- install pipenv
+- install tesseract ocr for your machine
 - Setup config vars of
     - `API_KEY`: From your https://checkvisaslots.com.
     - `BOT_TOKEN`: From BotFather in Telegram.
     - `CHAT_ID`: ID of group you want to run it in.
+    - `YEAR`: which year you're looking for slots in
+    - `MONTHS`: which months to notify you for
 - Change the location (default Chennai) and checking interval (default 10 mins).
+- run `pipenv install`
+- run `pipenv run python bot.py`
 
 Some info on using the bot. 
 - Monitoring start immediately - runs once every 10 mins. 

bot.py

@@ -4,6 +4,8 @@
 import traceback
 from datetime import datetime, timedelta
 from urllib.request import urlopen
+from urllib.parse import quote
+import re
 
 import cv2
 import numpy as np
@@ -12,6 +14,7 @@
 from dotenv import dotenv_values
 from telegram import ParseMode, Update
 from telegram.ext import CallbackContext, CommandHandler, Updater
+from thefuzz import fuzz, process
 
 SLOTS_URL = "https://app.checkvisaslots.com/slots/v1"
 SCREENSHOTS_URL = "https://app.checkvisaslots.com/retrieve/v1"
@@ -21,6 +24,8 @@
 CHAT_ID = dotenv_values("config.env").get("CHAT_ID")
 BOT_TOKEN = dotenv_values("config.env").get("BOT_TOKEN")
 SEND_DUPES = eval(dotenv_values("config.env").get("SEND_DUPES"))
+YEAR = dotenv_values("config.env").get("YEAR")
+MONTHS = dotenv_values("config.env").get("MONTHS").split(',')
 
 logging.basicConfig(format="%(asctime)s - %(name)s - %(levelname)s - %(message)s", level=logging.INFO)
 logger = logging.getLogger(__name__)
@@ -55,12 +60,12 @@ def image_to_bytes(img):
     return bts
 
 
-def check_chennai_consulate(results):
-    return [[each["slots"], each["createdon"]] for each in results if each["visa_location"] == "CHENNAI"][0]
+def check_NEWDELHI_consulate(results):
+    return [[each["slots"], each["createdon"]] for each in results['slotDetails'] if each["visa_location"] == "NEW DELHI VAC"][0]
 
 
-def get_chennai_screenshots(results):
-    screenshot_urls = [(each["img_url"], each["createdon"]) for each in results if "CHENNAI" in each["img_url"]]
+def get_NEWDELHI_screenshots(results):
+    screenshot_urls = [(each["img_url"], each["createdon"]) for each in results if quote("NEW DELHI VAC") in each["img_url"]]
     screenshots = []
     for url, timestamp in screenshot_urls:
         screenshots.append([url_to_image(SCREENSHOTS_FILE_URL + url), timestamp])
@@ -76,36 +81,49 @@ def run_once(bot, chat_id, old_log_text=None, old_log_pics=None, api_key=None, s
         old_log_pics = []
     if api_key is None:
         api_key = random.choice(API_KEYS)
-    response = requests.get(SLOTS_URL, headers=get_slot_header(api_key))
-    results = eval(response.text)
-    slots = check_chennai_consulate(results)
-
-    gmt_time = datetime.strptime(slots[1], "%a, %d %b %Y %H:%M:%S %Z")
-    ist_time = gmt_to_ist(gmt_time)
-    log_text = f'{slots[0]} slots available at {datetime.strftime(ist_time, "%I:%M:%S %p")}'
+    # commenting this request because we're making double calls to api and exhausting api calls faster(1st is this call, 2nd being screenshot call)
+    # response = requests.get(SLOTS_URL, headers=get_slot_header(api_key)) 
+    # mockRespText = '{"slotDetails":[{"img_id":7602746,"visa_location":"CHENNAI","slots":0,"createdon":"Sat, 29 Apr 2023 17:10:46 GMT"},{"img_id":7602836,"visa_location":"CHENNAI VAC","slots":1,"createdon":"Sat, 29 Apr 2023 17:21:03 GMT"},{"img_id":7602748,"visa_location":"HYDERABAD","slots":0,"createdon":"Sat, 29 Apr 2023 17:10:49 GMT"},{"img_id":7602839,"visa_location":"HYDERABAD VAC","slots":6,"createdon":"Sat, 29 Apr 2023 17:21:11 GMT"},{"img_id":7602640,"visa_location":"KOLKATA","slots":0,"createdon":"Sat, 29 Apr 2023 16:58:24 GMT"},{"img_id":7602785,"visa_location":"KOLKATA VAC","slots":4,"createdon":"Sat, 29 Apr 2023 17:17:35 GMT"},{"img_id":7602765,"visa_location":"MUMBAI","slots":0,"createdon":"Sat, 29 Apr 2023 17:14:19 GMT"},{"img_id":7602830,"visa_location":"MUMBAI VAC","slots":30,"createdon":"Sat, 29 Apr 2023 17:20:48 GMT"},{"img_id":7602766,"visa_location":"NEW DELHI","slots":0,"createdon":"Sat, 29 Apr 2023 17:14:22 GMT"},{"img_id":7602831,"visa_location":"NEW DELHI VAC","slots":6,"createdon":"Sat, 29 Apr 2023 17:20:49 GMT"}],"userDetails":{"visa_type":"B1/B2","appointment_type":"Regular","subscription":"FREE"}}\n'
+    # results = eval(mockRespText)
+    # results = eval(response.text)
+    # slots = check_NEWDELHI_consulate(results)
+
+    # gmt_time = datetime.strptime(slots[1], "%a, %d %b %Y %H:%M:%S %Z")
+    # ist_time = gmt_to_ist(gmt_time)
+    # log_text = f'{slots[0]} slots available at {datetime.strftime(ist_time, "%I:%M:%S %p")}'
+    log_text = "running now"
     logger.info(log_text)
-    if send_dupes or log_text != old_log_text:
-        if slots[0] > 0:
-            bot.send_message(chat_id=chat_id, text=f"🤯🤯🤯 {log_text} ({api_key}) @Syzygianinfern0 @M_N_Sathish")
-        else:
-            bot.send_message(chat_id=chat_id, text=f"{log_text} ({api_key})")
+    # if send_dupes or log_text != old_log_text:
+    #     if slots[0] > 0:
+    #         bot.send_message(chat_id=chat_id, text=f"🤯🤯🤯 {log_text} ({api_key}) ")
+    #     else:
+    #         bot.send_message(chat_id=chat_id, text=f"{log_text} ({api_key})")
     log_pics = []
-    if slots[0] > 0:
-        response = requests.get(SCREENSHOTS_URL, headers=get_screenshots_header(api_key))
-        results = eval(response.text)
-        screenshots = get_chennai_screenshots(results)
-        for screenshot, timestamp in screenshots:
-            text = pytesseract.image_to_string(screenshot)
-            if "consular" in text.lower():
-                gmt_time = datetime.strptime(timestamp, "%a, %d %b %Y %H:%M:%S %Z")
-                ist_time = gmt_to_ist(gmt_time)
-                if send_dupes or not any(np.array_equal(screenshot, each) for each in old_log_pics):
-                    bot.send_photo(
-                        chat_id=chat_id,
-                        photo=image_to_bytes(screenshot),
-                        caption=datetime.strftime(ist_time, "%I:%M:%S %p") + " @Syzygianinfern0 @M_N_Sathish",
-                    )
-                log_pics.append(screenshot)
+    # if slots[0] > 0:
+    response = requests.get(SCREENSHOTS_URL, headers=get_screenshots_header(api_key))
+    results = eval(response.text)
+    screenshots = get_NEWDELHI_screenshots(results)
+    for screenshot, timestamp in screenshots:
+        text = pytesseract.image_to_string(screenshot)
+        # if "consular" in text.lower():
+
+        # remove line with 'current appointment' in it
+        fuzzyMatch = process.extractOne('current appointment', text.splitlines(), scorer=fuzz.partial_ratio)
+        matchThreshold = 80
+        if fuzzyMatch is not None and fuzzyMatch[1] > matchThreshold:
+            text = ''.join(text.split(fuzzyMatch[0]))
+        if YEAR in text.lower() and len([month for month in MONTHS if month.lower() in text.lower()]) > 0:
+            text_with_dates = [re.sub(r"[^a-zA-Z0-9 ]", "", line) for line in text.splitlines() if YEAR in line]
+            gmt_time = datetime.strptime(timestamp, "%a, %d %b %Y %H:%M:%S %Z")
+            ist_time = gmt_to_ist(gmt_time)
+            if send_dupes or not any(np.array_equal(screenshot, each) for each in old_log_pics):
+                bot.send_message(chat_id=CHAT_ID, text="<b>"+'\n\n'.join(text_with_dates)+"</b>", parse_mode=ParseMode.HTML)
+                bot.send_photo(
+                    chat_id=chat_id,
+                    photo=image_to_bytes(screenshot),
+                    caption=datetime.strftime(ist_time, "%I:%M:%S %p") + "",
+                )
+            log_pics.append(screenshot)
 
     return log_text, log_pics
 
@@ -121,7 +139,7 @@ def monitor(bot, chat_id, old_log_text=None, old_log_pics=None):
         minutes = 5
     except Exception as e:
         logger.error(e)
-        bot.send_message(chat_id=CHAT_ID, text=f"Bot Crashed @Syzygianinfern0: {e}\n{traceback.format_exc()}")
+        bot.send_message(chat_id=CHAT_ID, text=f"Bot Crashed : {e}\n{traceback.format_exc()}")
         log_text, log_pics = old_log_text, old_log_pics
         minutes = 1
 
@@ -151,7 +169,7 @@ def run_once_handler(update: Update, context: CallbackContext):
         except Exception as e:
             logger.error(e)
             context.bot.send_message(
-                chat_id=update.effective_chat.id, text=f"Bot Crashed @Syzygianinfern0: {e}\n{traceback.format_exc()}"
+                chat_id=update.effective_chat.id, text=f"Bot Crashed : {e}\n{traceback.format_exc()}"
             )
     else:
         context.bot.send_message(chat_id=update.effective_chat.id, text="Unauthorized!")
@@ -165,7 +183,7 @@ def main():
 
     # Welcome message
     bot.send_message(
-        chat_id=CHAT_ID, text="<b>Bot Started!</b> @Syzygianinfern0 @M_N_Sathish", parse_mode=ParseMode.HTML
+        chat_id=CHAT_ID, text="<b>Bot Started!</b>", parse_mode=ParseMode.HTML
     )
     logger.info("Bot started!")
 

config.env

@@ -1,3 +1,6 @@
+API_KEY=<FILL ME>
 CHAT_ID=<FILL ME>
 BOT_TOKEN=<FILL ME>
 SEND_DUPES=True
+YEAR=2023
+MONTHS=may,june,july,august