first commit

Author: Lucas Peng

alexarank.py

@@ -0,0 +1,10 @@
+#!/usr/bin/env python
+from urllib import request
+import sys, re
+xml = request.urlopen('http://data.alexa.com/data?cli=10&dat=s&url=%s'%sys.argv[1]).read().decode("utf-8")
+sp = re.search(r'REACH RANK="\d+"', xml).span()
+print(int(xml[sp[0]+12:sp[1]-1]))
+
+# try: rank = int(re.search(r'\d+', xml).groups()[0])
+# except: rank = -1
+# print('Your rank for %s is %d!\n' % (sys.argv[1], rank))

hqtrivia.py

@@ -0,0 +1,163 @@
+import numpy as np
+import cv2
+from PIL import Image
+import sys, re
+import pyocr
+import pyocr.builders
+import webbrowser
+from nltk.tag import pos_tag
+from urllib import request
+#from google import search
+from bs4 import BeautifulSoup
+import string
+import urllib
+class HQTrivia:
+#    cam_size = (1280, 720)
+    cam_size = (1920, 1080)
+    capture_size = (int(700),int(280))
+    answer_size = (int(650),int(350))
+    capture_rect = ((int(cam_size[0]/2-capture_size[0]/2+70),
+                     int(cam_size[1]/2-capture_size[1]/2)),
+                    (int(cam_size[0]/2+capture_size[0]/2+70),
+                     int(cam_size[1]/2+capture_size[1]/2)))
+    answer_rect  = ((int(cam_size[0]/2-answer_size[0]/2+70),
+                     int(cam_size[1]/2-answer_size[1]/2 + 350)),
+                    (int(cam_size[0]/2+answer_size[0]/2+70),
+                     int(cam_size[1]/2+answer_size[1]/2)+ 350))
+
+    def __init__(self):
+        # initialize OCR
+        tools = pyocr.get_available_tools()
+        if len(tools) == 0:
+            print("No OCR tool found")
+            sys.exit(1)
+        self.tool = tools[0]
+        print("Will use tool '%s'" % (self.tool.get_name()))
+        langs = self.tool.get_available_languages()
+        print("Available languages: %s" % ", ".join(langs))
+        self.lang = langs[0]
+        print("Will use lang '%s'" % (self.lang))
+
+    def get_page(self,query):
+        headers = {'User-Agent':'Mozilla/5.0 (Windows NT 6.1; WOW64; rv:23.0) Gecko/20100101 Firefox/23.0'}
+        req = urllib.request.Request('http://www.google.com/search?q='+query,headers=headers)
+        page = urllib.request.urlopen(req)
+        html = page.read()
+        return html
+
+    def alexa_rank(self,url):
+        xml = request.urlopen('http://data.alexa.com/data?cli=10&dat=s&url=%s'%url).read().decode("utf-8")
+        sp = re.search(r'REACH RANK="\d+"', xml).span()
+        return int(xml[sp[0]+12:sp[1]-1])
+
+    def search_answer(self,query,ans):
+        #search whole question
+        query_plus = query.replace(' ','+')
+        #print("query url:", "www.google.com/search?q="+query_plus)
+        webbrowser.get('chrome').open_new_tab("http://www.google.com/search?q="+query_plus)
+
+        #try quote
+        find_quote = False
+        qmark_combine = (("“","\""),("\"","\""),("“","“"),("\"","“"),("“","”"),("\"","”"),("”","“"))
+        for qc in qmark_combine:
+            start_pt = query_plus.find(qc[0])
+            end_pt = query_plus.find(qc[1], start_pt + 1)  # add one to skip the opening "
+            quote = query_plus[start_pt + 1: end_pt]  # add one to get the quote excluding the ""
+            if len(quote)!=0 and start_pt!=-1 and end_pt!=-1:
+                print("find quote!")
+                find_quote = True
+                print(quote)
+                webbrowser.get('chrome').open_new_tab("http://www.google.com/search?q="+quote)
+
+        #search proper nouns
+        translator = str.maketrans('', '', string.punctuation)
+        query = query.translate(translator)
+        tagged_sent = pos_tag(query.split())
+        propernouns = [word for word,pos in tagged_sent if pos == 'NNP']
+        propernouns = list(filter(lambda x: x != "Which", propernouns))
+        if len(propernouns)!=0:
+            if find_quote == False:
+                print("search: "," ".join(propernouns))
+                webbrowser.get('chrome').open_new_tab("http://www.google.com/search?q="+"+".join(propernouns))
+        print('------------------------------------------------------')
+        if len(ans)<=5:
+            for a in ans:
+                print("\n"+a+" "+" ".join(propernouns))
+                try:
+                    html=self.get_page(a+"+"+"+".join(propernouns))
+                    html = str(html)
+                    start_idx = html.find("<h3 class=\"r\"><a href=\"")
+                    end_idx = html.find('\"',start_idx+len("<h3 class=\"r\"><a href=\""))
+                    url = html[start_idx+len("<h3 class=\"r\"><a href=\""):end_idx]
+                    print(url)
+                    print(len(url))
+                except:
+                    pass
+        print('------------------------------------------------------')
+        print('------------------------------------------------------')
+    def main_loop(self,flip=False):
+        cap = cv2.VideoCapture(0)
+        while(True):
+            # Capture frame-by-frame
+            ret, frame = cap.read()
+
+            # Our operations on the frame come here
+            gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+            #print(gray.shape)
+            if flip:
+                gray_flip = cv2.flip(gray,1)
+            else:
+                gray_flip = gray
+            cv2.rectangle(gray_flip, self.capture_rect[0], self.capture_rect[1], (0,0,0))
+            cv2.rectangle(gray_flip, self.answer_rect[0], self.answer_rect[1], (0,0,0))
+            # Display the resulting frame
+            cv2.imshow('frame',gray_flip)
+
+            key = cv2.waitKey(1) & 0xFF
+            if key == ord('c'):
+                print('capture!')
+                cap_p1 = self.capture_rect[0]
+                cap_p2 = self.capture_rect[1]
+                cap_img = gray[cap_p1[1]:cap_p2[1], cap_p1[0]:cap_p2[0]]
+                ans_p1 = self.answer_rect[0]
+                ans_p2 = self.answer_rect[1]
+                ans_img = gray[ans_p1[1]:ans_p2[1], ans_p1[0]:ans_p2[0]]
+                #cap_img = cv2.medianBlur(cap_img,5)
+                #cap_img = cv2.GaussianBlur(cap_img,(3,3),0)
+                ret,cap_img = cv2.threshold(cap_img,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU)
+                #ans_img = cv2.GaussianBlur(ans_img,(5,5),0)
+                #ret,ans_img = cv2.threshold(ans_img,200,255,cv2.THRESH_BINARY)
+                ret,ans_img = cv2.threshold(ans_img,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU)
+                #ans_img = cv2.adaptiveThreshold(ans_img,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,11,2)
+                #cap_img = cv2.adaptiveThreshold(cap_img,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,11,2)
+                cv2.imwrite('test.png',cap_img)
+                cv2.imwrite('test_ans.png',ans_img)
+                txt = self.tool.image_to_string(
+                    Image.fromarray(cap_img),
+                    lang=self.lang,
+                    builder=pyocr.builders.TextBuilder()
+                )
+                anstxt = self.tool.image_to_string(
+                    Image.fromarray(ans_img),
+                    lang=self.lang,
+                    builder=pyocr.builders.TextBuilder()
+                )
+                query = txt.replace('\n',' ').rstrip()
+                print(query)
+                # for b in ans_box:
+                #     print(b.content)
+                anstxt = anstxt.rstrip().split('\n')
+                anstxt = list(filter(lambda x: x!='', anstxt))
+                print(anstxt)
+                self.search_answer(query,anstxt)
+
+            elif key == ord('q'):
+                break
+
+        # When everything done, release the capture
+        cap.release()
+        cv2.destroyAllWindows()
+
+if __name__ == '__main__':
+    hqt = HQTrivia()
+    hqt.main_loop()