Update README.md

.dockerignore

@@ -1 +1,2 @@
 ./run
+./.venv

Dockerfile

@@ -17,8 +17,9 @@ RUN pip3 install --break-system-packages -r requirements.txt
 # Copy source code
 COPY . .
 
-# Mount working directory
+# Mount persistent working directory
 VOLUME ./run
 
 # Run the bot
-CMD ["python3", "src/main.py"]
+CMD ["python3", "-u", "src/main.py"]
+#CMD ["python3", "-u", "src/main.py", "--straight-to-queue"]

README.md

@@ -5,6 +5,8 @@ Twitter bot that tracks cross-company interactions between the non-JP branches o
 
 **This project was created to run [this account](https://twitter.com/NijiHolo_EN_ID).**
 
+![Screenshot_20230912_235245](https://github.com/muskit/muskit/assets/15199219/0359fb26-8a48-4698-9b78-66d7d852099e)
+
 ## Running
 With the way packages are setup, **you must have Docker installed and running!!**
 
@@ -12,11 +14,17 @@ Setup the `.env` in the project root. Refer to the [`.env`](#env) section for va
 
 Build and run the Docker container:
 ```bash
-# to run attached (can CTRL+P,CTRL+Q to detach)
-sh run.sh
+# to delete container and built image
+sh scripts/delete.sh
 
-# ... or to run headless
-sh run_detached.sh
+# to build image
+sh scripts/build.sh
+
+# to create container and run attached (can CTRL+P,CTRL+Q to detach)
+sh scripts/run.sh
+
+# ... or to run headless/detached
+sh scripts/run_detached.sh
 ```
 
 If attached to a container prepared by Dockerfile, you can run the program from project root (not in `src`). Refer to the following section for options.
@@ -36,17 +44,17 @@ These need to be defined in a `.env` file in the `run` ephemeral directory.
 ### Scraper Credentials
 To get around rate limitations imposed on users, we scrape with multiple accounts. Each account is defined in the file using the following format:
 ```
-scraper_usernameX=twitter_username
-scraper_passwordX=twitter_password
+scraperX_username=twitter_username
+scraperX_password=twitter_auth_token
 ```
 where `X` is a number starting from 0, increasing by 1 for each account added. For instance:
 ```
-scraper_username0=
-scraper_password0=
-scraper_username1=
-scraper_password1=
+scraper0_username=
+scraper0_password=
+scraper1_username=
+scraper1_password=
 ```
-The first account (`scraper_username0` and `scraper_password0`) **MUST be defined (`scraper_username` and `scraper_password` without number will not work!)**  and will be used to attempt scraping private accounts. Make sure this account follows any private accounts that you want to scrape!
+The first account (`scraper0_username` and `scraper0_password`) **MUST be defined (`scraper_username` and `scraper_password` without number will not work!)**  and will be used to attempt scraping private accounts. Make sure this account follows any private accounts that you want to scrape!
 ### Twitter API Stuff
 The following keys/tokens are used for the official API via `tweepy`. We mainly use these to just post tweets.
 ```
@@ -56,20 +64,16 @@ user_token=
 user_secret=
 ```
 ### Screenshot Cookie *(optional)*
-This is the authentication token obtained from a browser when signed in on the Twitter website. It's only needed if you want to screenshot tweets from privated accounts. Make sure the token belongs to an account that follows desired private accounts! Maybe have it belong to `scraper_username0`?
+This is the authentication token obtained from a browser when signed in on the Twitter website. It's only needed if you want to screenshot tweets from privated accounts. Make sure the token belongs to an account that follows desired private accounts! Maybe have it belong to `scraper0`?
 ```
 web_auth_token=
 ```
 ### Example `.env` without values
 ```
-scraper_username0=
-scraper_password0=
-scraper_username1=
-scraper_password1=
-scraper_username2=
-scraper_password2=
-scraper_username3=
-scraper_password3=
+scraper0_username=
+scraper0_password=
+scraper1_username=
+scraper1_password=
 web_auth_token=
 app_key=
 app_secret=

build.sh

@@ -1,3 +0,0 @@
-#!/bin/sh
-
-docker build -t nijiholo_bot .

requirements.txt

@@ -1,6 +1,6 @@
 python-dotenv
 nest-asyncio
 pytz
-git+https://github.com/mahrtayyab/tweety.git@e3d330280cb3b2e8f9d2bf2f20425c476f7671a5
+git+https://github.com/mahrtayyab/tweety.git
 tweepy
 tweet-capture

requirements_dev.txt

@@ -1,7 +1,7 @@
 python-dotenv
 nest-asyncio
 pytz
-git+https://github.com/mahrtayyab/tweety.git@e3d330280cb3b2e8f9d2bf2f20425c476f7671a5
+git+https://github.com/mahrtayyab/tweety.git
 tweepy
 tweet-capture
 opencv-python-headless

run.sh

@@ -1,4 +0,0 @@
-#!/bin/sh
-
-mkdir -p run
-docker run -v ./run:/app/run --name bot -it nijiholo_bot

run_detached.sh

@@ -1,4 +0,0 @@
-#!/bin/sh
-
-mkdir -p run
-docker run -v ./run:/app/run --name bot -d nijiholo_bot

scripts/build.sh

@@ -0,0 +1,6 @@
+#!/bin/sh
+
+CURPATH="$(dirname `realpath "$0"`)/.."
+#sudo docker build -t nijiholo_bot --no-cache "$CURPATH"
+sudo docker build -t nijiholo_bot "$CURPATH"
+sudo docker container create -v "$CURPATH/run:/app/run" --name bot nijiholo_bot

scripts/delete.sh

@@ -0,0 +1,4 @@
+#!/bin/sh
+
+sudo docker container rm bot
+sudo docker image rm nijiholo_bot

scripts/run.sh

@@ -0,0 +1,6 @@
+#!/bin/sh
+CURPATH="$(dirname `realpath "$0"`)/.."
+cd "$CURPATH"
+mkdir -p run
+#sudo docker run -v "$CURPATH/run:/app/run" --name bot -it nijiholo_bot
+sudo docker container start -a -i bot

scripts/run_detached.sh

@@ -0,0 +1,6 @@
+#!/bin/sh
+CURPATH="$(dirname `realpath "$0"`)/.."
+cd "$CURPATH"
+mkdir -p run
+#sudo docker run -v "$CURPATH/run:/app/run" --name bot -d nijiholo_bot
+sudo docker container start bot

scripts/shell.sh

@@ -0,0 +1,2 @@
+#!/bin/sh
+sudo docker exec -it bot sh

src/account_pool.py

@@ -11,13 +11,14 @@ class AccountPool:
         creds = dotenv_values(working_path(file=".env"))
         i = 0
         while True:
-            if f"scraper_username{i}" in creds and f"scraper_password{i}" in creds:
+            if f"scraper{i}_username" in creds and f"scraper{i}_password" in creds:
                 self.__accounts.append(
-                    (creds[f"scraper_username{i}"], creds[f"scraper_password{i}"])
+                    (creds[f"scraper{i}_username"], creds[f"scraper{i}_password"])
                 )
                 i += 1
             else:
                 break
+        print(f"{len(self.__accounts)} scraper credentials found!")
 
     def use_index(self, idx):
         self.__idx = idx

src/catchup.py

@@ -16,8 +16,9 @@ import ttweetqueue as ttq
 
 PROGRAM_ARGS = None
 
+preempt_done = False
 safe_to_post_tweets = True
-scraper: Scraper
+scraper = Scraper()
 
 
 # Updates TTweetQueue
@@ -87,7 +88,7 @@ async def process_queue() -> bool:
 
     queued_ttweets_count = queue.get_count()
 
-    WAIT_TIME = 60 * 15
+    WAIT_TIME = 60 * 30 # 30 minutes
     ttweets_posted = 0
 
     if queued_ttweets_count == 0:
@@ -110,7 +111,7 @@ async def process_queue() -> bool:
                 ttweets_posted += 1
                 print(f"({ttweets_posted}/{queued_ttweets_count}) done")
                 if not queue.is_empty():
-                    print(f"resting for {WAIT_TIME}s...")
+                    print(f"resting for {WAIT_TIME/60} minutes...")
                     await asyncio.sleep(WAIT_TIME - 5)
                     print("5 second warning!")
                     await asyncio.sleep(5)
@@ -127,13 +128,13 @@ async def process_queue() -> bool:
 # return False = issue occurred where we couldn't post all past tweets properly
 async def run(PROGRAM_ARGS):
     global safe_to_post_tweets
+    global preempt_done
     global scraper
     global queue
 
-    scraper = Scraper()
     queue = ttq.TalentTweetQueue.instance
 
-    # post tweets given in command line first
+    # OPTION: post tweets given in command line first
     if PROGRAM_ARGS.post_id is not None and len(PROGRAM_ARGS.post_id) > 0:
         PROGRAM_ARGS.post_id.sort()
         print("Posting specified tweets first.")
@@ -150,11 +151,38 @@ async def run(PROGRAM_ARGS):
                 print("Successfully posted tweet. Sleeping for 5 minutes")
                 await asyncio.sleep(60 * 5)
             else:
-                print("Did not post tweet")
+                print("Did not post tweet\n")
         print("Done processing specified tweets")
         PROGRAM_ARGS.post_id = None
 
-    # refresh stored queue first
+    # PREEMPT: post tweet IDs in preempt.txt if exists and not empty
+    if not preempt_done:
+        try:
+            with open(working_path(file="preempt.txt"), "r") as preempt_file:
+                print("Found preempt.txt! Posting stored IDs unconditionally...")
+
+                for l in preempt_file:
+                    if len(l) == 0: continue
+                    try:
+                        id = int(l.strip().split()[0])
+                    except:
+                        print(f"Error occurred processing {l}, skipping...")
+                        continue
+
+                    posted = await TwAPI.instance.post_ttweet_by_id(id, PROGRAM_ARGS.dry_run)
+                    if posted:
+                        queue.add_finished_tweet(id)
+                        print("Successfully posted tweet. Sleeping for 5 minutes")
+                        await asyncio.sleep(60 * 5)
+                    else:
+                        print("Could not post tweet\n")
+
+                print("Finished processing preempt.txt")
+                preempt_done = True
+        except FileNotFoundError:
+            print("preempt.txt wasn't found")
+
+    # OPTION: refresh stored queue first
     if PROGRAM_ARGS.refresh_queue:
         PROGRAM_ARGS.refresh_queue = False
         print("Refreshing queue tweets...")

src/scraper.py

@@ -1,6 +1,7 @@
 from os.path import exists
 from time import sleep
 from datetime import datetime, timedelta
+import traceback
 
 import pytz
 
@@ -14,8 +15,10 @@ from tweety_utils import *
 from talenttweet import *
 import talent_lists
 
-
+# TODO: on RateLimit encounter, determine when it will probably
+# unlock and wait just until then
 class Scraper:
+    COOLDOWN = 16 # minutes
     def __init__(self):
         Scraper.instance = self
         self.__account = AccountPool()
@@ -47,16 +50,16 @@ class Scraper:
     def login_wait(self, private=False):
         if private:
             print(
-                f"keeping pvt-accessible account ({self.__account.use_index(0)[0]}). sleeping for 4 minutes..."
+                f"keeping pvt-accessible account ({self.__account.use_index(0)[0]}). sleeping for {Scraper.COOLDOWN} minutes..."
             )
-            sleep(240)
+            sleep(60*Scraper.COOLDOWN)
             print()
             l = self.try_login(0)
         else:
             l = self.try_login()
         if not l:
-            print("sleeping for 4 minutes...")
-            sleep(240)
+            print(f"sleeping for {Scraper.COOLDOWN} minutes...")
+            sleep(60*Scraper.COOLDOWN)
             print()
             self.try_login()
 
@@ -86,7 +89,7 @@ class Scraper:
         return tweet
 
     def get_tweet(self, id: int, private_user=False):
-        # print(f'{id}{" on private" if private_user else ""}')
+        # print(f'getting {id}{" on private" if private_user else ""}')
         if private_user:
             self.try_login(0)
         while True:
@@ -96,19 +99,29 @@ class Scraper:
             except RateLimitReached:
                 print("RateLimitReached occurred")
                 self.login_wait(private_user)
-            except UnknownError:
-                print("UnknownError occurred, probably rate-limited")
+            except UnknownError as e:
+                print(f"UnknownError occurred: {e.message.rstrip()}")
+                print(f"skipping attempt to get tweet {id}...")
+                return None
+                # if any(x in e.message.lower() for x in ["missing", "post is unavailable", "delete"]) :  # tweet is probably unavailable
+                #     print(f"tweet {id} seems unavailable; skipping...")
+                #     return None
+                # if "account owner limits" in e.message.lower(): # private tweet
+                #     print("trying again as pvt-accessible...\n")
+                #     return self.get_tweet(id, True)
+                # print("treating like RateLimitReached and using the next scraper...")
                 # traceback.print_exc()
-                self.login_wait(private_user)
+                # self.login_wait(private_user)
             except Exception as e:
-                if not private_user:
-                    print("Unhandled exception occurred, trying again as private...")
-                    return self.get_tweet(id, True)
-                else:
-                    print(
-                        f"Unhandled exception occurred, tweet {id} is probably unavailable"
-                    )
-                    print(e)
+                # if not private_user:
+                #     print("Unhandled exception occurred getting tweet!")
+                #     traceback.print_exc()
+                #     print("trying again as pvt-accessible...\n")
+                #     return self.get_tweet(id, True)
+                # else:
+                    print("Unhandled exception occurred")
+                    traceback.print_exc()
+                    print(f"skipping tweet {id}")
                     return None
 
     # since MUST BE TIMEZONE AWARE
@@ -126,7 +139,7 @@ class Scraper:
         else:
             print(f"grabbing tweets since {since.date()}")
 
-        uid = self.app._get_user_id(username)
+        uid = int(self.app._get_user_id(username))
         print(f"{username} = {uid}")
 
         def add_tweet(tweet: Tweet):
@@ -168,6 +181,8 @@ class Scraper:
                 for e in cur_page:
                     if isinstance(e, Tweet):
                         add_tweet(e)
+                        if e == cur_page[-1]:
+                            print(f"{e.date} (last tweet) < {since.date()} (since) ?")
                     elif isinstance(e, SelfThread):
                         # FIXME: rework when replied_to is fixed (currently populates user_mentions)
                         # latest tweet in thread = og author's reply
@@ -175,9 +190,14 @@ class Scraper:
                             add_tweet(t)
 
                 cur = search.cursor
-            except (UnknownError, RateLimitReached):
-                print("UnknownError occurred, probably rate-limited")
+            except RateLimitReached:
+                print("RateLimitReached occurred getting tweets from user")
                 self.login_wait(uid in talent_lists.privated_accounts)
+            except UnknownError as e:
+                print(f"UnknownError occurred getting tweets from user: {e.message.rstrip()}")
+                print("treating like RateLimitReached...")
+                self.login_wait(uid in talent_lists.privated_accounts)
+            sleep(5) # FIXME: temporary attempt to avoid scraper lock-up
 
         tweets.sort(key=lambda t: t.id)
         return tweets

src/talenttweet.py

@@ -311,9 +311,11 @@ class TalentTweet:
                 rtm_msg(QUOTED_TWEET_MENTIONS_B, quoted_username)
             else:
                 ret += QUOTE_TWEET.format(author_username, quoted_username)
-        elif len(self.mentions) > 0:  # standalone tweet
+        elif len(self.mentions) > 0:  # standalone tweet that mentions other
             ret += TWEET.format(author_username, ", ".join(mention_usernames))
             mention_usernames.clear()
+        elif len(self.rt_mentions) > 0: # reply to non-talent tweet that mentions B
+            rtm_msg(REPLY_TO_MENTION_B, "")
         else:
             raise ValueError(
                 f"TalentTweet {self.tweet_id} has insufficient other parties"

src/ttweetqueue.py

@@ -73,9 +73,9 @@ class TalentTweetQueue:
                 for line in f:
                     if len(line) > 0:
                         ttweet = tt.TalentTweet.deserialize(line)
-                        if ttweet.tweet_id in self.ttweets_dict:
-                            self.ttweets_dict[ttweet.tweet_id] = ttweet
+                        if ttweet.tweet_id not in self.ttweets_dict:
                             print(f"adding unfinished tweet {ttweet.tweet_id}")
+                            self.ttweets_dict[ttweet.tweet_id] = ttweet
         # finished ttweets
         try:
             with open(self.finished_ttweets_path, "r") as f:

src/twapi.py

@@ -130,6 +130,7 @@ class TwAPI:
 
     async def get_ttweet_image_media_id(self, ttweet):
         img = await util.create_ttweet_image(ttweet)
+        print(f"obtaining media id for {img}...")
         media = self.api.media_upload(img)
         return media.media_id
 
@@ -141,6 +142,7 @@ class TwAPI:
         print(
             f"------{ttweet.tweet_id} ({util.get_username_local(ttweet.author_id)})------"
         )
+        print(ttweet)
 
         text = ttweet.announce_text()
         ttweet_url = ttweet.url()
@@ -151,41 +153,41 @@ class TwAPI:
             return False
 
         # main tweet: text + screenshot
-        try:
+        # try:
         print("creating main QRT w/ screenshot...")
         media_ids = [await self.get_ttweet_image_media_id(ttweet)]
         twt_resp = await self.post_tweet(
             text, media_ids=media_ids, quote_tweet_id=ttweet.tweet_id
         )
         print("done")
-        except:
-            print(
-                "error occurred trying to create main tweet, falling back to URL-main + reply screencap format"
-            )
-            traceback.print_exc()
-            try:
-                print("posting main tweet...")
-                twt_resp = await self.post_tweet(text, quote_tweet_id=ttweet.tweet_id)
-                print("done")
-                twt_id = twt_resp.data["id"]
+        # except:
+        #     print(
+        #         "error occurred trying to create main tweet, falling back to URL-main + reply screencap format"
+        #     )
+        #     traceback.print_exc()
+        #     try:
+        #         print("posting main tweet...")
+        #         twt_resp = await self.post_tweet(text, quote_tweet_id=ttweet.tweet_id)
+        #         print("done")
+        #         twt_id = twt_resp.data["id"]
 
-                try:
-                    print("creating reply img...", end="")
-                    media_ids = [await self.get_ttweet_image_media_id(ttweet)]
-                    print("posting reply tweet...", end="")
-                    await self.post_tweet(reply_to_tweet=twt_id, media_ids=media_ids)
-                    print("done")
-                except:
-                    print("Had trouble posting reply image tweet.")
-                print("successfully posted ttweet!")
-            except tweepy.Forbidden as e:
-                if "duplicate content" in e.api_messages[0]:
-                    print(
-                        "Twitter says the TalentTweet is a duplicate; skipping error-free..."
-                    )
-                    return False
-                else:
-                    raise e
+        #         try:
+        #             print("creating reply img...", end="")
+        #             media_ids = [await self.get_ttweet_image_media_id(ttweet)]
+        #             print("posting reply tweet...", end="")
+        #             await self.post_tweet(reply_to_tweet=twt_id, media_ids=media_ids)
+        #             print("done")
+        #         except:
+        #             print("Had trouble posting reply image tweet.")
+        #         print("successfully posted ttweet!")
+        #     except tweepy.Forbidden as e:
+        #         if "duplicate content" in e.api_messages[0]:
+        #             print(
+        #                 "Twitter says the TalentTweet is a duplicate; skipping error-free..."
+        #             )
+        #             return False
+        #         else:
+        #             raise e
         return True
 
     async def post_ttweet_by_id(self, id: int, dry_run=False):

src/util.py

@@ -83,6 +83,7 @@ async def create_ttweet_image(ttweet):
         tc.driver_path = "/usr/bin/chromedriver"
     filename = working_path(file="img.png")
     img = None
+    print(f"Creating image for TalentTweet {ttweet.url()}")
     try:
         os.remove(filename)
     except:
@@ -94,10 +95,12 @@ async def create_ttweet_image(ttweet):
             mode=4,
             night_mode=1,
             show_parent_tweets=True,
+            #parent_tweets_limit=3
         )
         img = fix_aspect_ratio(img)
-    except:
-        print("unable to create tweet image")
+    except Exception as e:
+        print("ERROR: unable to create tweet image")
+        print(e)
         traceback.print_exc()
         return None