def get_goose(link):
r = requests.get(link)
clean_url = re.findall('url=https?://abstrusegoose.com/.*"', r.text)
url = clean_url[0].lstrip('url=').rstrip('"')
r = requests.get(url)
logging.warning(r.url)
try:
urls_strips = re.findall('https?://abstrusegoose.com/strips/.*png',
r.text)
url = urls_strips[0]
i = fetch_image(url)
except Exception:
i = None
src = f'[Source]({r.url}).'
try:
soup = BeautifulSoup(r.text, 'html5lib')
txt = soup.find("div", {"id": "blog_text"})
txt = html.unescape(
str(txt).lstrip('<div id="blog_text"><p>\n<p>').rstrip(
'</p></div></p>\n'))
txt = str(lxml_html.fromstring(txt).text_content())
for item in ('r/>', 'align="center">', 'a href="', '" target="_blank"',
're>', 'm>', 'strong>', '/p>'):
txt = txt.replace(item, ' ')
txt = txt.replace('r/>', '') + "\n" + src
except Exception:
txt = src
try:
title = soup.find("h1", {"class": "storytitle"})
title = title.text
except Exception:
title = ''
text = title + '\n' + txt
return i, text
def get_phd(link):
r = requests.get(link)
soup = BeautifulSoup(r.text, 'html5lib')
images = soup.findAll('img', {"id": "comic", "name": "comic"})
url = images[0]["src"]
i = fetch_image(url)
txt = f' [Source]({r.url}).'
return i, txt
def get_exo(link):
r = requests.get(link)
logging.warning(r.url)
urls = re.findall('https?://www.exocomics.com/wp-content/uploads/.*jpg',
r.text)
i = fetch_image(urls[0])
txt = f' [Source]({r.url}).'
return i, txt
def post():
"""
(7) Make a prediction from a URL provided via POST request.
"""
url = request.json.get('url')
img = utils.fetch_image(url)
result = utils.predict_from_image(CLF, img)
return jsonify(result)
def get_tom_gauld(link):
r = requests.get(link)
logging.warning(r.url)
soup = BeautifulSoup(r.text, 'html.parser')
images = soup.findAll('img')
urls = re.findall('https?://64.media.tumblr.com/.*jpg', str(images[0]))
src = f' [Source]({r.url}).'
i = fetch_image(urls[0])
try:
txt = str(images[0]).split("src")[0].lstrip('<img alt=') + src
except Exception:
txt = src
return i, txt
def form():
"""
(4b) Make a prediction from a URL given via a GET form.
"""
url = request.args.get('url')
if url:
img = utils.fetch_image(url)
result = utils.predict_from_image(CLF, img)
result['url'] = url # If we add this back, we can display it.
else:
result = {}
return render_template('form.html', result=result)
def get_from_gocomics(link):
r = requests.get(link)
logging.warning(r.url)
page = BeautifulSoup(r.text, "html5lib")
links = []
for img in page.findAll("div",
{'class': "comic__image js-comic-swipe-target"}):
links.append(img)
urls = re.findall('src="https://assets\.amuniversal\.com.*srcset',
str(links[0]))
url = urls[0].lstrip('src="').rstrip('" srcset"')
logging.warning(url)
i = fetch_image(url)
txt = f"[Source.]({r.url})"
return i, txt
def get_xkcd(link):
r = requests.get(link)
logging.warning(r.url)
src = f'[Source]({r.url}).'
try:
soup = BeautifulSoup(r.text, 'html5lib')
img_url = soup.find_all("img")
text = img_url[2]['title']
title = img_url[2]['alt']
url = 'https:' + (img_url[2]['srcset'].split(' '))[0]
i = fetch_image(url)
txt = title + '\n' + text + '\n' + src
except Exception:
i = None
txt = src
return i, txt
def api():
"""
(8) Make a prediction from a base64-encoded image via POST request.
If accessing the web API from code, you may not have a URL to pass to
the service, and there is no form for doing a file upload. So we need
a way to pass the image as data. There are lots of ways to do this; one
way is to encode as base64.
"""
data = request.json.get('image')
if data.startswith('http'):
img = utils.fetch_image(data)
else:
img = Image.open(BytesIO(base64.b64decode(data)))
result = utils.predict_from_image(CLF, img)
return jsonify(result)
def get_dilbert(link, latest=False):
if latest:
date = str(datetime.date.today())
else:
date = get_random_date(1989, 4, 16)
full_link = link + date
r = requests.get(full_link)
soup = BeautifulSoup(r.text, 'html5lib')
images = soup.findAll('img')
url = str(images[2]['src'])
i = fetch_image(url)
src = f' [Source]({full_link}).'
try:
txt = str(images[2]['alt']) + src
except Exception:
txt = None
return i, txt
def get_smbc(link):
r = requests.get(link)
urls = re.findall('https?://www.smbc-comics.com/comics/.*p?n?gi?f?',
r.text)
logging.warning(r.url)
src = f' [Source]({r.url}).'
try:
txt = re.findall('img\stitle=.*id="cc-comic"', r.text)
txt = txt[0].lstrip('img title="').split('"')[0]
txt = html.unescape(txt) + src
except Exception:
txt = src
if not urls:
logging.warning(link)
return None, None
i = fetch_image(urls[0])
return i, txt
def predict():
"""
(3) Make a prediction from a URL given via GET request.
Using a URL means we can still just accept a string as an arg.
There's still no human interface.
"""
url = request.args.get('url')
img = utils.fetch_image(url)
result = utils.predict_from_image(CLF, img)
# Deal with not getting a URL.
# if url:
# img = utils.fetch_image(url)
# result = utils.predict_from_image(CLF, img)
# else:
# result = 'Please provide a URL'
return jsonify(result)
def get_new_yorker_cartoon(link):
r = urlopen(link)
page = BeautifulSoup(r, "html5lib")
logging.warning(r.url)
links = []
for link in page.findAll(
"picture",
{"class": "responsive-cartoon__image responsive-image"}):
links.append(link)
res = []
for item in links:
caption = re.findall('<img alt=.*" class="responsive-image__image', str(item))[0].\
lstrip('<img alt="').\
rstrip(' class="responsive-image__image')
url = re.findall('class="responsive-image__image" sizes="100vw".* srcset=', str(item))[0].\
lstrip('class="responsive-image__image" sizes="100vw" src="').\
rstrip('" srcset=')
logging.warning(url)
i = fetch_image(url)
txt = html.unescape(caption)
res.append((i, txt))
return [item for item in res[:2]]
def get_poorlydrawnlines(link):
r = requests.get(link)
logging.warning(r.url)
try:
if r.url.endswith('comic/'):
soup = BeautifulSoup(r.text, 'html5lib')
list_of_comics = []
for link in soup.find_all('a'):
if ('png' or 'jpg') in str(link.get('href')):
list_of_comics.append(link.get('href'))
url = list_of_comics[0]
else:
urls = re.findall(
'https?://www.poorlydrawnlines.com/wp-content/uploads/.*j?p?n?g',
r.text)
url = urls[0].split()[0]
if url.endswith('><img'):
url = url.rstrip('><img')
except Exception:
return None, None
i = fetch_image(url)
txt = f' [Source]({r.url}).'
return i, txt
