Added CoffeeHouse ALG

Makefile

@@ -13,9 +13,13 @@ clean_dltc:
 clean_nlpfr:
 	rm -rf nlpfr/build nlpfr/dist nlpfr/nltk.egg-info
 
+clean_alg:
+	rm -rf alg/build alg/dist alg/alg.egg-info
+
 clean:
 	make clean_apt clean_stopwords clean_tokenizer clean_nlpfr
 	make clean_dltc
+	make clean_alg
 
 # ======================================================================================================================
 
@@ -38,9 +42,13 @@ build_nlpfr:
 build_dltc:
 	cd dltc; python3 setup.py build; python3 setup.py sdist
 
+build_alg:
+	cd alg; python3 setup.py build; python3 setup.py dist
+
 build:
 	make build_nlpfr
 	make build_dltc
+	make build_alg
 
 # ======================================================================================================================
 
@@ -63,5 +71,10 @@ install_nlpfr:
 install_dltc:
 	cd dltc; python3 setup.py install
 
+install_alg:
+	cd alg; python3 setup.py install
+
 install:
-	make install_nlpfr install_dltc
+	make install_nlpfr
+	make install_dltc
+	make install_alg

alg/README.md

@@ -0,0 +1,24 @@
+# CoffeeHouse ALG
+
+CoffeeHouse ALG is a algorithmia API wrapper
+
+## Image Tagger
+
+Predicts the content and appropriate tags for an image
+
+```python
+from coffeehouse_alg.utilities import Utilities
+from coffeehouse_alg.image_tagger import ImageTagger
+
+url = "https://cdn4.telesco.pe/file/ETCYVng0VOEczJuIjUbsgzC4a7EOi8wMfwfFBAss_OlKZd822FNiF5Xy5V7MpY0nUvHcBCzDt1MexuBrawuit0WL5HGa0-O5bZXYsBwA5P4QAFKM6VyH0k1G9VbnCn-QiFJ5ipClcP1azPykgcRqDz4d-g_3IHWegBY3wHBD8BHm1T589XhkXbfCl_HN6TJSZ1LIYnoASnN1FF6YDV92UOhMc0-c2FTQkoxx-gGt4A7Vwy821x4-TAd6AlNI_q5NO4NKe047oRlJfm_MCtEsojYGP5tyJKB-xZYm_zRtzriwC0GSJh2zoY_RCv4L-Uudq3GGQ0_8A2OrHCmCNj0AGA.jpg"
+
+# Process the image
+image_tagger = ImageTagger()
+image_content = Utilities.process_img_from_url(url)
+
+image_tagger.predict_tags(image_content)
+# {'result': {'character': [], 'copyright': [], 'general': [{'photo': 0.995330810546875}, {'cosplay': 0.28029680252075195}, {'solo': 0.2679383456707001}], 'rating': [{'safe': 0.7592381834983826}, {'questionable': 0.21827033162117004}, {'explicit': 0.024775557219982147}]}, 'metadata': {'content_type': 'json', 'duration': 0.298899313}}
+
+image_tagger.predict_inception(image_content)
+# {'result': {'tags': [{'class': 'Persian cat', 'confidence': 0.6589894890785217}, {'class': 'Egyptian cat', 'confidence': 0.009713547304272652}, {'class': 'handkerchief, hankie, hanky, hankey', 'confidence': 0.00886216014623642}, {'class': 'Angora, Angora rabbit', 'confidence': 0.008085943758487701}, {'class': 'teddy, teddy bear', 'confidence': 0.00662141153588891}]}, 'metadata': {'content_type': 'json', 'duration': 0.594030507}}
+```

alg/coffeehouse_alg/__init__.py

@@ -0,0 +1,14 @@
+from . import utilities
+from .utilities import *
+
+from . import image_tagger
+from image_tagger import *
+
+from . import exceptions
+from exceptions import *
+
+__all__ = ['utilities',
+           'image_tagger',
+           'exceptions',
+           'Utilities',
+           'CoffeeHouseAlgException']

alg/coffeehouse_alg/exceptions.py

@@ -0,0 +1,14 @@
+__all__ = ["CoffeeHouseAlgException"]
+
+
+class CoffeeHouseAlgException(Exception):
+    """
+    Exception raised by API errors.
+    The exception message is set to the server's response.
+    """
+
+    def __init__(self, type, message, content):
+        self.type = type
+        self.content = content
+        self.message = "{0} ERROR: {1}".format(type, message)
+        super().__init__(self.message or content)

alg/coffeehouse_alg/image_tagger/__init__.py

@@ -0,0 +1,4 @@
+from . import image_tagger
+from .image_tagger import *
+
+__all__ = ['image_tagger', 'ImageTagger']

alg/coffeehouse_alg/image_tagger/image_tagger.py

@@ -0,0 +1,85 @@
+import json
+import requests
+
+from coffeehouse_alg.exceptions import CoffeeHouseAlgException
+from coffeehouse_alg.utilities import Utilities
+
+
+class ImageTagger:
+
+    def __init__(self, access_key=None,
+                 illustration_endpoint="https://api.algorithmia.com/v1/web/algo/demo/IllustrationTaggerDemo/0.1.0",
+                 inception_endpoint="https://api.algorithmia.com/v1/web/algo/demo/InceptionNetDemo/0.1.0",
+                 js_source="https://demos.algorithmia.com/image-tagger/public/js/main.js"):
+        """
+        Public constructor
+
+        :param endpoint: The main endpoint for Illustration Tagger
+        :param js_source: The javascript source code containing the API Key
+        """
+        self.access_key = access_key
+        self.illustration_endpoint = illustration_endpoint
+        self.inception_endpoint = inception_endpoint
+        self.js_source = js_source
+
+        if access_key is None:
+            self.access_key = Utilities.get_access_key(self.js_source)
+
+    def predict_tags(self, image_content, results=12):
+        """
+        Processes the bare-bone request and returns the JSON results
+
+        :param image_content:
+        :param results:
+        :return:
+        """
+        headers = {
+            "Accept": "application/json, text/javascript",
+            "Authorization": self.access_key,
+            "Content-Type": "application/json",
+            "Origin": "https://demos.algorithmia.com",
+            "Referer": "https://demos.algorithmia.com/image-tagger",
+            "Sec-Fetch-Dest": "empty",
+            "Sec-Fetch-Mode": "cors",
+            "Sec-Fetch-Site": "same-site",
+        }
+        payload = {
+            "image": image_content,
+            "numResults": results
+        }
+        headers = Utilities.add_client_headers(headers)
+
+        response = requests.post(self.illustration_endpoint, data=json.dumps(payload), headers=headers)
+        j_response = json.loads(response.text)
+
+        if 'error' in j_response:
+            raise CoffeeHouseAlgException('SERVER', j_response['error']['message'], response)
+
+        return j_response
+
+    def predict_inception(self, image_content):
+        """
+        Processes the bare-bone request and returns the JSON results
+
+        :param image_content:
+        :return:
+        """
+        headers = {
+            "Accept": "application/json, text/javascript",
+            "Authorization": self.access_key,
+            "Content-Type": "application/json",
+            "Origin": "https://demos.algorithmia.com",
+            "Referer": "https://demos.algorithmia.com/image-tagger",
+            "Sec-Fetch-Dest": "empty",
+            "Sec-Fetch-Mode": "cors",
+            "Sec-Fetch-Site": "same-site",
+        }
+        headers = Utilities.add_client_headers(headers)
+
+        response = requests.post(self.inception_endpoint, data=json.dumps(image_content), headers=headers)
+        j_response = json.loads(response.text)
+
+        if 'error' in j_response:
+            raise CoffeeHouseAlgException('SERVER', j_response['error']['message'], response)
+
+        return j_response

alg/coffeehouse_alg/utilities.py

@@ -0,0 +1,58 @@
+import requests
+import mimetypes
+import base64
+import re
+
+
+class Utilities:
+
+    @staticmethod
+    def get_cookies(http_endpoint):
+        """
+        Makes the initial request to get all the required cookies
+
+        :return: dict value of all the returned cookies
+        """
+        session = requests.Session()
+        session.get(http_endpoint)
+        return session.cookies.get_dict()
+
+    @staticmethod
+    def get_access_key(http_endpoint):
+        """
+        Sends a request to a Javascript content file and parses the contents
+        to extract the access key (Simple)
+
+        :param http_endpoint:
+        :return:
+        """
+        response = requests.get(http_endpoint)
+        access_key = re.findall('\'([^"]*)\'', response.text.splitlines()[1])[0]
+        return "Simple {0}".format(access_key)
+
+    @staticmethod
+    def process_img_from_url(url):
+        """
+        Downloads the image from the URl and processes it into a proper
+        base64 encoded image with the mime type information included
+
+        :param url:
+        :return:
+        """
+        mimetype = mimetypes.guess_type(url)[0]
+        image_content = requests.get(url).content
+        image_b64content = base64.b64encode(image_content).decode('utf8')
+        return "data:%s;base64,%s" % (mimetype, image_b64content)
+
+    @staticmethod
+    def add_client_headers(headers):
+        """
+        Adds the required client headers
+        :param headers:
+        :return:
+        """
+        headers['Accept-Encoding'] = 'gzip, deflate, br'
+        headers['Accept-Language'] = 'en'
+        headers['User-Agent'] = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) ' \
+                                'Chrome/80.0.3987.122 Safari/537.36 '
+        return headers

alg/requirements.txt

@@ -0,0 +1 @@
+requests

alg/setup.py

@@ -0,0 +1,23 @@
+from setuptools import setup, find_packages
+
+setup(
+    name='coffeehouse_alg',
+    version='1.0.0',
+    description='CoffeeHouse algorithmia wrapper',
+    url='https://github.com/Intellivoid/CoffeeHouse-ALG',
+    author='Zi Xing Narrakas',
+    author_email='netkas@intellivoid.info',
+    classifiers=[
+        #   3 - Alpha
+        #   4 - Beta
+        #   5 - Production/Stable
+        'Development Status :: 4 - Beta',
+        'Topic :: API Wrapper',
+        'Programming Language :: Python :: 3',
+    ],
+    keywords='multi-label classification nlp neural networks deep learning api wrapper',
+    packages=find_packages(exclude=['tests']),
+    install_requires=[
+        'requests'
+    ]
+)

alg/tests/example.py

@@ -0,0 +1,11 @@
+from coffeehouse_alg.utilities import Utilities
+from coffeehouse_alg.image_tagger import ImageTagger
+
+
+url = "https://cdn4.telesco.pe/file/ETCYVng0VOEczJuIjUbsgzC4a7EOi8wMfwfFBAss_OlKZd822FNiF5Xy5V7MpY0nUvHcBCzDt1MexuBrawuit0WL5HGa0-O5bZXYsBwA5P4QAFKM6VyH0k1G9VbnCn-QiFJ5ipClcP1azPykgcRqDz4d-g_3IHWegBY3wHBD8BHm1T589XhkXbfCl_HN6TJSZ1LIYnoASnN1FF6YDV92UOhMc0-c2FTQkoxx-gGt4A7Vwy821x4-TAd6AlNI_q5NO4NKe047oRlJfm_MCtEsojYGP5tyJKB-xZYm_zRtzriwC0GSJh2zoY_RCv4L-Uudq3GGQ0_8A2OrHCmCNj0AGA.jpg"
+
+image_tagger = ImageTagger()
+image_content = Utilities.process_img_from_url(url)
+
+print(image_tagger.predict_tags(image_content))
+print(image_tagger.predict_inception(image_content))