Expand lines

Co-authored-by: aider (gemini/gemini-2.5-pro) <aider@aider.chat>

.gitignore

@@ -115,3 +115,4 @@ venv.bak/
 
 secrets.py
 tmp.png
+.aider*

consumable_label.py

@@ -0,0 +1,40 @@
+from PIL import Image, ImageEnhance, ImageFont, ImageDraw
+import os
+import textwrap
+import qrcode
+import urllib.parse
+
+location = os.path.dirname(os.path.realpath(__file__))
+
+def print_consumable_label(item):
+    im = Image.open(location + '/label.png')
+    width, height = im.size
+    draw = ImageDraw.Draw(im)
+
+    #logging.info('Printing consumable label item: %s', item)
+
+    encodeded = urllib.parse.quote(item)
+    url = 'https://my.protospace.ca/out-of-stock?item=' + encodeded
+
+    qr = qrcode.make(url, version=6, box_size=9)
+    im.paste(qr, (840, 325))
+
+    item_size = 150
+
+    w = 9999
+    while w > 1200:
+        item_size -= 5
+        font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', item_size)
+        w, h = draw.textsize(item, font=font)
+
+    x, y = (width - w) / 2, ((height - h) / 2) - 140
+    draw.text((x, y), item, font=font, fill='black')
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', 100)
+    draw.text((100, 410), 'Out of stock?', font=font, fill='black')
+    draw.text((150, 560), 'Scan here:', font=font, fill='black')
+
+    im.save('tmp.png')
+
+
+print_consumable_label('Brown paper towel')

forum_label.py

@@ -0,0 +1,92 @@
+from PIL import Image, ImageEnhance, ImageFont, ImageDraw
+import os
+import textwrap
+import qrcode
+import urllib.parse
+
+location = os.path.dirname(os.path.realpath(__file__))
+
+def print_forum_label(thread):
+    im = Image.open(location + '/label.png')
+    width, height = im.size
+    draw = ImageDraw.Draw(im)
+
+    #logging.info('Printing forum thread: %s', thread['title'])
+
+    url = 'https://forum.protospace.ca/t/{}/'.format(thread['id'])
+
+    qr = qrcode.make(url, version=6, box_size=9)
+    im.paste(qr, (840, 150))
+
+    item_size = 150
+
+    w = 9999
+    while w > 1200:
+        item_size -= 5
+        font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', item_size)
+        w, h = draw.textsize(url, font=font)
+
+    x, y = (width - w) / 2, ((height - h) / 2) + 300
+    draw.text((x, y), url, font=font, fill='black')
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf', 80)
+    draw.text((120, 150), 'Forum Thread', font=font, fill='black')
+
+    text = thread['title']
+
+    MARGIN = 50
+    MAX_W, MAX_H, PAD = 900 - (MARGIN*2), 450 - (MARGIN*2), 5
+
+    def fit_text(text, font_size):
+        font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', font_size)
+
+        for cols in range(100, 1, -4):
+            print('trying size', font_size, 'cols', cols)
+
+            paragraph = textwrap.wrap(text, width=cols, break_long_words=False)
+
+            total_h = -PAD
+            total_w = 0
+
+            for line in paragraph:
+                w, h = draw.textsize(line, font=font)
+                if w > total_w:
+                    total_w = w
+                total_h += h + PAD
+
+            if total_w <= MAX_W and total_h < MAX_H:
+                return True, paragraph, total_h
+
+        return False, [], 0
+
+    font_size_range = [1, 500]
+
+    # Thanks to Alex (UDIA) for the binary search algorithm
+    while abs(font_size_range[0] - font_size_range[1]) > 1:
+        font_size = sum(font_size_range) // 2
+        image_fit, check_para, check_h = fit_text(text, font_size)
+        if image_fit:
+            print('Does fit')
+            font_size_range = [font_size, font_size_range[1]]
+            good_size = font_size
+            paragraph = check_para
+            total_h = check_h
+        else:
+            print('Does not fit')
+            font_size_range = [font_size_range[0], font_size]
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', good_size)
+    offset = height - MAX_H - MARGIN
+    start_h = -100 + offset
+
+    current_h = start_h
+    for line in paragraph:
+        w, h = draw.textsize(line, font=font)
+        x, y = (MAX_W - w) / 2, current_h
+        draw.text((x+MARGIN, y), line, font=font, fill='black')
+        current_h += h + PAD
+
+    im.save('tmp.png')
+
+
+print_forum_label(dict(id=10197, title='Pitch: A wild split-flap display appeared!'))

requirements.txt

@@ -1,18 +1,18 @@
-annotated-types==0.5.0
 certifi==2022.6.15
 charset-normalizer==2.1.1
 idna==3.3
-inflect==7.0.0
+inflect==6.0.0
-jaraco.itertools==6.4.1
+jaraco.itertools==6.2.1
-more-itertools==9.1.0
+more-itertools==8.14.0
+paho-mqtt==2.1.0
 Pillow==9.2.0
-pydantic==2.0.3
+pydantic==1.10.2
-pydantic_core==2.3.0
 pyserial==3.5
 pytz==2022.2.1
+qrcode==8.1
 requests==2.28.1
 six==1.16.0
-typing_extensions==4.7.1
+typing-extensions==4.4.0
 urllib3==1.26.12
 git+https://git.tannercollin.com/tanner/wolframalpha.git
 x256==0.0.3

secrets.py.example

@@ -1,11 +1,8 @@
 wa_api_key = ''
 
-# Get these from your network inspector after sending a message to character.ai
+openai_key = ''
-character_ai_cookies = {
+
-    '_legacy_auth0.whatever.is.authenticated': 'true',
+MQTT_WRITER_PASSWORD = ''
-    'auth0.whatever.is.authenticated': 'true',
+
-    'messages': '',
+FORUM_SEARCH_API_KEY = ''
-    'csrftoken': '',
+
-    'sessionid': '',
-}
-character_ai_token = ''

utils.py

@@ -0,0 +1,681 @@
+#!/usr/bin/env python
+
+import os, logging
+import warnings
+warnings.filterwarnings("ignore", category=DeprecationWarning)
+
+import requests
+import pytz
+import re
+import os
+import time
+import json
+import textwrap
+import random
+import qrcode
+import urllib.parse
+import unicodedata
+from PIL import Image, ImageEnhance, ImageFont, ImageDraw
+from datetime import datetime, timezone, timedelta
+import paho.mqtt.publish as publish
+
+try:
+    import secrets
+    wa_api_key = secrets.wa_api_key
+except:
+    wa_api_key = None
+
+try:
+    import secrets
+    openai_key = secrets.openai_key
+    FORUM_SEARCH_API_KEY = secrets.FORUM_SEARCH_API_KEY
+    MQTT_WRITER_PASSWORD = secrets.MQTT_WRITER_PASSWORD
+except:
+    openai_key = None
+    FORUM_SEARCH_API_KEY = None
+    MQTT_WRITER_PASSWORD = None
+
+
+TIMEZONE_CALGARY = pytz.timezone('America/Edmonton')
+
+DRUGWARS_LOCATION = '/home/pi/protovac/env/bin/drugwars'
+NETHACK_LOCATION = '/usr/games/nethack'
+MORIA_LOCATION = '/usr/games/moria'
+_2048_LOCATION = '/home/pi/2048-cli/2048'
+FROTZ_LOCATION = '/usr/games/frotz'
+HITCHHIKERS_LOCATION = '/home/pi/frotz/hhgg.z3'
+SUDOKU_LOCATION = '/usr/games/nudoku'
+
+HAS_DRUGWARS = os.path.isfile(DRUGWARS_LOCATION)
+HAS_NETHACK = os.path.isfile(NETHACK_LOCATION)
+HAS_MORIA = os.path.isfile(MORIA_LOCATION)
+HAS_2048 = os.path.isfile(_2048_LOCATION)
+HAS_FROTZ = os.path.isfile(FROTZ_LOCATION)
+HAS_HITCHHIKERS = os.path.isfile(HITCHHIKERS_LOCATION)
+HAS_SUDOKU = os.path.isfile(SUDOKU_LOCATION)
+
+
+location = os.path.dirname(os.path.realpath(__file__))
+
+with open(location + '/info.txt') as f:
+    PROTO_INFO = f.read()
+
+for num, line in enumerate(PROTO_INFO.split('\n')):
+    try:
+        line.encode('ascii')
+    except UnicodeEncodeError:
+        print('non-ascii found in line:', num+1)
+        raise
+
+with open(location + '/lastquestion.txt') as f:
+    LAST_QUESTION = f.read()
+
+
+def format_date(datestr):
+    if not datestr: return 'None'
+
+    d = datetime.strptime(datestr, '%Y-%m-%dT%H:%M:%SZ').replace(tzinfo=pytz.UTC)
+    d = d.astimezone(TIMEZONE_CALGARY)
+    return d.strftime('%a %b %-d, %Y  %-I:%M %p')
+
+def normalize_to_ascii(s):
+    return unicodedata.normalize('NFKD', s).encode('ascii', 'ignore').decode('ascii')
+
+def truncate_string(s, max_length):
+    return s[:max_length-3] + '...' if len(s) > max_length else s
+
+def sign_send(to_send):
+    try:
+        logging.info('Sending to sign: %s', to_send)
+        data = dict(sign=to_send, on_behalf_of='protovac')
+        r = requests.post('https://api.my.protospace.ca/stats/sign/', data=data, timeout=5)
+        r.raise_for_status()
+        return 'Success!'
+    except BaseException as e:
+        logging.exception(e)
+        return 'Error'
+
+def protovac_sign_color(color):
+    try:
+        logging.info('Sending color to protovac sign: %s', color)
+        data = dict(on=True, bri=255, seg=[dict(col=[color, [0,0,0]])])
+        r = requests.post('http://10.139.251.5/json', json=data, timeout=3)
+        r.raise_for_status()
+        return 'Success!'
+    except BaseException as e:
+        logging.exception(e)
+        return 'Error'
+
+def protovac_sign_effect(effect):
+    try:
+        logging.info('Sending effect to protovac sign: %s', effect)
+        data = dict(on=True, bri=255, seg=[dict(fx=effect)])
+        r = requests.post('http://10.139.251.5/json', json=data, timeout=3)
+        r.raise_for_status()
+        return 'Success!'
+    except BaseException as e:
+        logging.exception(e)
+        return 'Error'
+
+def fetch_stats():
+    try:
+        logging.info('Fetching status...')
+        r = requests.get('https://api.my.protospace.ca/stats/', timeout=5)
+        r.raise_for_status()
+        return r.json()
+    except BaseException as e:
+        logging.exception(e)
+        return 'Error'
+
+def fetch_classes():
+    try:
+        logging.info('Fetching classes...')
+        r = requests.get('https://api.my.protospace.ca/sessions/', timeout=5)
+        r.raise_for_status()
+        return r.json()['results']
+    except BaseException as e:
+        logging.exception(e)
+        return 'Error'
+
+def fetch_protocoin():
+    try:
+        logging.info('Fetching protocoin...')
+        r = requests.get('https://api.my.protospace.ca/protocoin/transactions/', timeout=5)
+        r.raise_for_status()
+        return r.json()
+    except BaseException as e:
+        logging.exception(e)
+        return 'Error'
+
+def mqtt_publish(topic, message):
+    if not MQTT_WRITER_PASSWORD:
+        return False
+
+    try:
+        publish.single(
+            topic,
+            str(message),
+            hostname='172.17.17.181',
+            port=1883,
+            client_id='protovac',
+            keepalive=5,  # timeout
+        )
+    except BaseException as e:
+        logging.error('Problem sending MQTT message: ' + str(e))
+
+QUOTES = [
+    'THEY MADE ME WEAR THIS',
+    'ASK ME ABOUT TOAST',
+    'ASK ME ABOUT BIKESHEDDING',
+    'ASK ME ABOUT VETTING',
+    'ASK ME ABOUT MAGNETS',
+    'ASK ME ABOUT SPACE',
+    'ASK ME ABOUT COUNTING',
+    'EXPERT WITNESS',
+    'AS SEEN ON TV',
+    'CONTAINS MEAT',
+    'PROTOCOIN ECONOMIST',
+    'EXPERT ON ALIENS',
+    'EXPERT ON WARP DRIVES',
+    'CHIEF OF STARFLEET OPERATIONS',
+    'ALIEN DOCTOR',
+    'NASA ASTROLOGIST',
+    'PINBALL WIZARD',
+    'JEDI KNIGHT',
+    'GHOSTBUSTER',
+    'DOUBLE AGENT',
+    'POKEMON TRAINER',
+    'POKEMON GYM LEADER',
+    'ASSISTANT TO THE REGIONAL MANAGER',
+    'BOUNTY HUNTER',
+    'I\'M NOT A DOCTOR',
+    'SPACE PIRATE',
+    'BATTERIES NOT INCLUDED',
+    'QUANTUM MECHANIC',
+    'PROTO SPACEX PILOT',
+    'EARTHBENDER',
+    'AIRBENDER',
+    'WATERBENDER',
+    'FIREBENDER',
+    '01001000 01101001',
+    'CURRENT EBAY BID: $8.51',
+    'MADE YOU LOOK!',
+    '(OR SIMILAR PRODUCT)',
+    'BATTERY MAY EXPLODE OR LEAK',
+    'CONNECT GROUND WIRE TO AVOID SHOCK',
+    'COOK THROROUGHLY',
+    'CURRENT AT TIME OF PRINTING',
+    'DO NOT BLEACH',
+    'DO NOT LEAVE UNATTENDED',
+    'DO NOT REMOVE TAG UNDER PENALTY OF LAW',
+    'DROP IN ANY MAILBOX',
+    'EDITED FOR TELEVISION',
+    'FOR A LIMITED TIME ONLY',
+    'FOR INDOOR OR OUTDOOR USE ONLY',
+    'KEEP AWAY FROM FIRE OR FLAMES',
+    'KEEP AWAY FROM SUNLIGHT',
+    'MADE FROM 100% RECYCLED ELECTRONS',
+    'LIFEGUARD ON DUTY',
+    'NOT DISHWASHER SAFE',
+    'NOT TO BE COMBINED WITH OTHER RADIOISOTOPES',
+    'NOT TO BE USED AS A PERSONAL FLOTATION DEVICE',
+    'PEEL FROM PAPER BACKING BEFORE EATING',
+    'STORE IN A COOL, DRY PLACE',
+    'VOID WHERE PROHIBITED',
+    'THE FUTURE IS NOW',
+    'MASTER OF DISGUISE',
+    'YOUR PERSONAL TIME TRAVEL GUIDE',
+    'OFFICIAL TASTE TESTER',
+    'INTERGALACTIC AMBASSADOR',
+    'VIRTUAL REALITY PIONEER',
+    'PARANORMAL INVESTIGATOR',
+    'UNDERCOVER SUPERHERO',
+    'THE COSMIC CHEF',
+    'THE ROBOT WHISPERER',
+    'THE DREAM WEAVER',
+    'USE AT YOUR OWN RISK',
+    'RESULTS MAY VARY',
+    'READ INSTRUCTIONS CAREFULLY',
+    'KEEP OUT OF REACH OF PETS',
+    'USE ONLY AS DIRECTED',
+    'NOT INTENDED FOR MEDICAL USE',
+    'DO NOT USE IF SEAL IS BROKEN',
+    'PRODUCT SOLD AS-IS',
+    'NO WARRANTIES, EXPRESS OR IMPLIED',
+    'USE CAUTION WHEN HANDLING',
+    'CRASH OVERRIDE',
+    'ACID BURN',
+    'CEREAL KILLER',
+    'ZERO COOL',
+    'ASK ME HOW I SAVED 15% ON MY CAR INSURANCE',
+    'TELL YOUR CAT I SAID PSP PSP PSP',
+    'I\'M BEGGING YOU, DON\'T SAY HI',
+    'BACK 2 BACK HOTDOG EATING WORLD CHAMPION',
+    'I ATE A BROWNIE ONCE',
+    'THIS ISN\'T ACTUALLY MY NAME',
+    'I OFTEN WONDER HOW I EVEN GOT HERE',
+    'YOU READ THIS, NOW WE MUST DUEL',
+    'DAVE\'S ARCH NEMESIS',
+    'EXCEPTIONALLY MID',
+    'VOTE ME 4 PREZADENT',
+    'PREVALENT IN OTHER WAYS',
+    'SCINTILLATING CHATOYANCY',
+    'TRAIN EXPERT',
+    'ASSISTANT MANAGER',
+    'WOW, ANOTHER TAG LINE',
+    'I SURVIVED VETTING AND ALL I GOT WAS THIS LOUSY NAME-TAG',
+    'NOT GUEST',
+    'WHEN WAS THE LAST TIME YOU TOOK OUT A GARBAGE?',
+    'YOU HAD ME AT BATMAN',
+    'DAD?',
+    'LEAD PROJECT UN-FINISHER',
+    'NOODLE CONNOISSEUR',
+    'GARTH\'S #1 FAN',
+    'UNVETTABLE',
+    'B-',
+    'CLOWN CAR DRIVER',
+    'IS A NAME TAG ON BREAD CONSIDERED A SANDWICH?',
+    'IS A NAME TAG FOLDED IN HALF CONSIDERED A TACO?',
+    'DISHWASHER SAFE',
+    'WET NOODLE',
+    'I\'M HERE FOR THE SIMULATION',
+    'BIRDS ARE NOT REAL',
+]
+random.shuffle(QUOTES)
+
+quote_count = 0
+assigned_quotes = {}
+
+def print_nametag(name, guest=False):
+    global quote_count
+    quote = ''
+
+    if guest:
+        quote_size = 120
+        quote = 'GUEST'
+        logging.info('Printing guest nametag for: %s', name)
+    else:
+        quote_size = 80
+        name_lookup = name.lower()[:4]
+        if name_lookup in assigned_quotes:
+            quote = assigned_quotes[name_lookup]
+        else:
+            quote = QUOTES[quote_count % len(QUOTES)]
+            quote_count += 1
+            assigned_quotes[name_lookup] = quote
+        logging.info('Printing member nametag for: %s, quote: %s', name, quote)
+
+    name_size = 305
+
+    im = Image.open(location + '/label.png')
+    width, height = im.size
+    draw = ImageDraw.Draw(im)
+
+    w = 9999
+    while w > 1084:
+        name_size -= 5
+        font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', name_size)
+        w, h = draw.textsize(name, font=font)
+
+    x, y = (width - w) / 2, ((height - h) / 2) - 20
+    draw.text((x, y), name, font=font, fill='black')
+
+    w = 9999
+    while w > 1200:
+        quote_size -= 5
+        font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', quote_size)
+        w, h = draw.textsize(quote, font=font)
+
+    x, y = (width - w) / 2, height - h - 30
+    draw.text((x, y), quote, font=font, fill='black')
+
+    im.save('tmp.png')
+    os.system('lp -d dymo tmp.png > /dev/null 2>&1')
+
+
+def print_tool_label(wiki_num):
+    im = Image.open(location + '/blank.png')
+    w1, h1 = im.size
+
+    logging.info('Printing tool label for ID: %s', wiki_num)
+
+    draw = ImageDraw.Draw(im)
+
+    params = {'id': str(wiki_num), 'size': '4'}
+    res = requests.get('https://labels.protospace.ca/', stream=True, params=params, timeout=5)
+    res.raise_for_status()
+
+    label = Image.open(res.raw)
+
+    new_size = (1280, 640)
+    label = label.resize(new_size, Image.ANTIALIAS)
+
+    w2, h2 = label.size
+
+    x, y = int((w1 - w2) / 2), int((h1 - h2) / 2)
+
+    im.paste(label, (x, y))
+
+    im.save('tmp.png')
+    os.system('lp -d dymo tmp.png > /dev/null 2>&1')
+
+
+def print_sheet_label(name, contact):
+    def get_date():
+        d = datetime.now(tz=timezone.utc)
+        d = d.astimezone(TIMEZONE_CALGARY)
+        return d.strftime('%b %-d, %Y')
+
+    def get_expiry_date():
+        d = datetime.now(tz=timezone.utc) + timedelta(days=90)
+        d = d.astimezone(TIMEZONE_CALGARY)
+        return d.strftime('%b %-d, %Y')
+
+    logging.info('Printing sheet label for: %s, contact: %s', name, contact)
+
+    name_size = 85
+    contact_size = 65
+    date_size = 65
+
+    im = Image.open(location + '/label.png')
+
+    draw = ImageDraw.Draw(im)
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', name_size)
+    draw.text((20, 300), name, font=font, fill='black')
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', contact_size)
+    draw.text((20, 425), contact, font=font, fill='black')
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', date_size)
+    date_line = 'Printed: ' + get_date()
+    draw.text((20, 590), date_line, font=font, fill='black')
+    date_line = 'EXPIRES: ' + get_expiry_date()
+    draw.text((20, 680), date_line, font=font, fill='black')
+
+    im.save('tmp.png')
+    os.system('lp -d dymo tmp.png > /dev/null 2>&1')
+
+
+def _fit_text(text, font_size, draw, max_w, max_h):
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', font_size)
+    pad = 5
+
+    for cols in range(100, 1, -4):
+        paragraph = textwrap.wrap(text, width=cols, break_long_words=False)
+        total_h = -pad
+        total_w = 0
+
+        for line in paragraph:
+            w, h = draw.textsize(line, font=font)
+            if w > total_w:
+                total_w = w
+            total_h += h + pad
+
+        if total_w <= max_w and total_h < max_h:
+            return True, paragraph, total_h
+    return False, [], 0
+
+
+def print_generic_label(text):
+    MARGIN = 50
+    MAX_W, MAX_H, PAD = 1285 - (MARGIN*2), 635 - (MARGIN*2), 5
+
+    logging.info('Printing generic label: %s', text)
+
+    im = Image.open(location + '/label.png')
+    width, height = im.size
+    draw = ImageDraw.Draw(im)
+
+    font_size_range = [1, 500]
+
+    # Thanks to Alex (UDIA) for the binary search algorithm
+    while abs(font_size_range[0] - font_size_range[1]) > 1:
+        font_size = sum(font_size_range) // 2
+        image_fit, check_para, check_h = _fit_text(text, font_size, draw, MAX_W, MAX_H)
+        if image_fit:
+            font_size_range = [font_size, font_size_range[1]]
+            good_size = font_size
+            paragraph = check_para
+            total_h = check_h
+        else:
+            font_size_range = [font_size_range[0], font_size]
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', good_size)
+    offset = height - MAX_H - MARGIN
+    start_h = (MAX_H - total_h) // 2 + offset
+
+    current_h = start_h
+    for line in paragraph:
+        w, h = draw.textsize(line, font=font)
+        x, y = (MAX_W - w) / 2, current_h
+        draw.text((x+MARGIN, y), line, font=font, fill='black')
+        current_h += h + PAD
+
+
+    im.save('tmp.png')
+    os.system('lp -d dymo tmp.png > /dev/null 2>&1')
+
+
+def print_consumable_label(item):
+    im = Image.open(location + '/label.png')
+    width, height = im.size
+    draw = ImageDraw.Draw(im)
+
+    logging.info('Printing consumable label item: %s', item)
+
+    encodeded = urllib.parse.quote(item)
+    url = 'https://my.protospace.ca/out-of-stock?item=' + encodeded
+
+    qr = qrcode.make(url, version=6, box_size=9)
+    im.paste(qr, (840, 325))
+
+    item_size = 150
+
+    w = 9999
+    while w > 1200:
+        item_size -= 5
+        font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', item_size)
+        w, h = draw.textsize(item, font=font)
+
+    x, y = (width - w) / 2, ((height - h) / 2) - 140
+    draw.text((x, y), item, font=font, fill='black')
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', 100)
+    draw.text((100, 410), 'Out of stock?', font=font, fill='black')
+    draw.text((150, 560), 'Scan here:', font=font, fill='black')
+
+    im.save('tmp.png')
+    os.system('lp -d dymo tmp.png > /dev/null 2>&1')
+
+
+def print_forum_label(thread):
+    im = Image.open(location + '/label.png')
+    width, height = im.size
+    draw = ImageDraw.Draw(im)
+
+    logging.info('Printing forum thread ID: %s, title: %s', thread['id'], thread['title'])
+
+    url = 'https://forum.protospace.ca/t/{}/'.format(thread['id'])
+
+    qr = qrcode.make(url, version=6, box_size=9)
+    im.paste(qr, (840, 150))
+
+    item_size = 150
+
+    w = 9999
+    while w > 1200:
+        item_size -= 5
+        font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', item_size)
+        w, h = draw.textsize(url, font=font)
+
+    x, y = (width - w) / 2, ((height - h) / 2) + 300
+    draw.text((x, y), url, font=font, fill='black')
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf', 80)
+    draw.text((120, 150), 'Forum Thread', font=font, fill='black')
+
+    text = thread['title']
+
+    MARGIN = 50
+    MAX_W, MAX_H, PAD = 900 - (MARGIN*2), 450 - (MARGIN*2), 5
+
+    font_size_range = [1, 500]
+
+    # Thanks to Alex (UDIA) for the binary search algorithm
+    while abs(font_size_range[0] - font_size_range[1]) > 1:
+        font_size = sum(font_size_range) // 2
+        image_fit, check_para, check_h = _fit_text(text, font_size, draw, MAX_W, MAX_H)
+        if image_fit:
+            font_size_range = [font_size, font_size_range[1]]
+            good_size = font_size
+            paragraph = check_para
+            total_h = check_h
+        else:
+            font_size_range = [font_size_range[0], font_size]
+
+    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', good_size)
+    offset = height - MAX_H - MARGIN
+    start_h = -100 + offset
+
+    current_h = start_h
+    for line in paragraph:
+        w, h = draw.textsize(line, font=font)
+        x, y = (MAX_W - w) / 2, current_h
+        draw.text((x+MARGIN, y), line, font=font, fill='black')
+        current_h += h + PAD
+
+    im.save('tmp.png')
+    os.system('lp -d dymo tmp.png > /dev/null 2>&1')
+
+
+def search_forum_thread(search):
+    params = dict(q=search + ' in:title')
+    headers = {'Api-Key': FORUM_SEARCH_API_KEY, 'Api-Username': 'system'}
+
+    results = []
+
+    r = requests.get('https://forum.protospace.ca/search.json', params=params, headers=headers, timeout=5)
+    r.raise_for_status()
+    r = r.json()
+    for topic in r.get('topics', []):
+        results.append(dict(
+            id=topic['id'],
+            title=normalize_to_ascii(topic['title']),
+        ))
+    return sorted(results, key=lambda x: x['id'], reverse=True)[:7]
+
+
+def message_protovac(thread):
+    try:
+        logging.info('Message to Protovac: %s', thread[-1]['content'])
+
+        data = dict(
+            messages=thread,
+            model='gpt-3.5-turbo',
+            temperature=0.5,
+            user='Protovac',
+            max_tokens=1000,
+        )
+        headers = {'Authorization': 'Bearer ' + openai_key}
+
+        res = None
+        try:
+            res = requests.post('https://api.openai.com/v1/chat/completions', json=data, headers=headers, timeout=4)
+        except requests.ReadTimeout:
+            logging.info('Got timeout, modifying prompt...')
+            data['messages'][-1]['content'] = 'Be terse in your response to this: ' + data['messages'][-1]['content']
+            res = requests.post('https://api.openai.com/v1/chat/completions', json=data, headers=headers, timeout=20)
+        res.raise_for_status()
+        res = res.json()
+        gpt_reply = res['choices'][0]['message']
+
+        logging.info('Message reply: %s', gpt_reply['content'])
+
+        return gpt_reply
+    except BaseException as e:
+        logging.exception(e)
+        return dict(role='assistant', content='INSUFFICIENT DATA FOR A MEANINGFUL ANSWER.')
+
+if wa_api_key:
+    import wolframalpha
+    wa_client = wolframalpha.Client(wa_api_key)
+else:
+    wa_client = None
+
+def think_send(query):
+    if not wa_client:
+        return 'WOLFRAM|ALPHA API KEY NOT CONFIGURED'
+
+    result = ''
+    try:
+        res = wa_client.query(query, timeout=10)
+    except BaseException as e:
+        logging.error('Error hitting W|A API: {} - {}\n'.format(e.__class__.__name__, e))
+        return 'Network error'
+
+    if 'didyoumeans' in res:
+        try:
+            guess = res['didyoumeans']['didyoumean']['#text']
+        except TypeError:
+            guess = res['didyoumeans']['didyoumean'][0]['#text']
+        next_result = think_send(guess)
+        result += 'Confused, using \'' + guess + '\'\n' + next_result
+    elif 'pod' in res:
+        pods = res['pod'] if isinstance(res['pod'], list) else [res['pod']]
+        for pod in pods:
+            title = pod['@title']
+            subpods = pod['subpod'] if isinstance(pod['subpod'], list) else [pod['subpod']]
+            plaintexts = []
+
+            for subpod in subpods:
+                if subpod['plaintext']:
+                    plaintexts.append(subpod['plaintext'])
+
+            plaintext = '; '.join(plaintexts)
+
+            if any([x in title.lower() for x in ['input', 'conversion', 'corresponding', 'comparison', 'interpretation']]):
+                pass
+            elif 'definition' in title.lower():
+                if plaintext[0] == '1':
+                    definition = plaintext.split('\n')[0].split(' | ', 1)[1]
+                else:
+                    definition = plaintext
+                result += 'Definition: ' + definition + '\n'
+            elif 'result' in title.lower():
+                if re.match(r'^\d+/\d+$', plaintext):
+                    plaintext += '\n' + think_send(plaintext + '.0')
+                if 'base' in query.lower() and '_' in plaintext:
+                    plaintext = '(Base conversion) "' + plaintext + '"'
+                if '(irreducible)' in plaintext and '/' in plaintext:
+                    result = think_send(query + '.0')
+                    break
+                else:
+                    result += 'Result: ' + plaintext + '\n'
+                    break
+            elif plaintext:
+                result += title + ': ' + plaintext + '\n'
+                break
+    else:
+        result = 'Error'
+
+    result = result.strip()
+
+    if len(result) > 500:
+        result = result[:500] + '... truncated.'
+    elif len(result) == 0:
+        result = 'Error'
+
+    result = result.replace('Wolfram|Alpha', 'Protovac')
+    result = result.replace('Stephen Wolfram', 'Tanner') # lol
+    result = result.replace('and his team', '')
+
+    for word in ['according to', 'asked', 'although', 'approximately']:
+        idx = result.lower().find('('+word)
+        if idx > 0:
+            result = result[:idx-1]
+
+    if result == 'Error':
+        result = 'INSUFFICIENT DATA FOR A MEANINGFUL ANSWER.'
+
+    return result