generate a JWT random string code example

Example 1: python generate secret key

>>> import secrets
>>> secrets.token_bytes(16)
b'\xebr\x17D*t\xae\xd4\xe3S\xb6\xe2\xebP1\x8b'
>>> secrets.token_hex(16)
'f9bf78b9a18ce6d46a0cd2b0b86df9da'
>>> token_urlsafe(16)  
'Drmhze6EPcv0fN_81Bj-nA'

Example 2: randbits python

#import OpenCV moduleimport cv2import osimport numpy as npimport matplotlib.pyplot as plt%matplotlib inline#function to detect facedef detect_face (img):#convert the test image to gray imagegray = cv2.cvtColor (img, cv2.COLOR_BGR2GRAY)#load OpenCV face detectorface_cas = cv2.CascadeClassifier ('-File name.xml-')faces = face_cas.detectMultiScale (gray, scaleFactor=1.3, minNeighbors=4);#if no faces are detected then return imageif (len (faces) == 0):return None, None#extract the facefaces [0]=(x, y, w, h)#return only the face partreturn gray[y: y+w, x: x+h], faces [0]#this function will read all persons' training images, detect face #from each image#and will return two lists of exactly same size, one listdef prepare_training_data(data_folder_path):#------STEP-1--------#get the directories (one directory for each subject) in data folderdirs = os.listdir(data_folder_path)faces = []labels = []for dir_name in dirs:#our subject directories start with letter 's' so#ignore any non-relevant directories if anyif not dir_name.startswith("s"):continue;#------STEP-2--------#extract label number of subject from dir_name#format of dir name = slabel#, so removing letter 's' from dir_name will give us labellabel = int(dir_name.replace("s", ""))#build path of directory containin images for current subject subject#sample subject_dir_path = "training-data/s1"subject_dir_path = data_folder_path + "/" + dir_name#get the images names that are inside the given subject directorysubject_images_names = os.listdir(subject_dir_path)#------STEP-3--------#go through each image name, read image,#detect face and add face to list of facesfor image_name in subject_images_names:#ignore system files like .DS_Storeif image_name.startswith("."):continue;#build image path#sample image path = training-data/s1/1.pgmimage_path = subject_dir_path + "/" + image_name#read imageimage = cv2.imread(image_path)#display an image window to show the imagecv2.imshow("Training on image...", image)cv2.waitKey(100)#detect faceface, rect = detect_face(image)#------STEP-4--------#we will ignore faces that are not detectedif face is not None:#add face to list of facesfaces.append(face)#add label for this facelabels.append(label)cv2.destroyAllWindows()cv2.waitKey(1)cv2.destroyAllWindows()return faces, labels#let's first prepare our training data#data will be in two lists of same size#one list will contain all the faces#and other list will contain respective labels for each faceprint("Preparing data...")faces, labels = prepare_training_data("training-data")print("Data prepared")#print total faces and labelsprint("Total faces: ", len(faces))print("Total labels: ", len(labels))#create our LBPH face recognizerface_recognizer = cv2.face.createLBPHFaceRecognizer()#train our face recognizer of our training facesface_recognizer.train(faces, np.array(labels))#function to draw rectangle on image#according to given (x, y) coordinates and#given width and heighdef draw_rectangle(img, rect):(x, y, w, h) = rectcv2.rectangle(img, (x, y), (x+w, y+h), (0, 255, 0), 2)#function to draw text on give image starting from#passed (x, y) coordinates.def draw_text(img, text, x, y):cv2.putText(img, text, (x, y), cv2.FONT_HERSHEY_PLAIN, 1.5, (0, 255, 0), 2)#this function recognizes the person in image passed#and draws a rectangle around detected face with name of the subjectdef predict(test_img):#make a copy of the image as we don't want to chang original imageimg = test_img.copy()#detect face from the imageface, rect = detect_face(img)#predict the image using our face recognizerlabel= face_recognizer.predict(face)#get name of respective label returned by face recognizerlabel_text = subjects[label]#draw a rectangle around face detecteddraw_rectangle(img, rect)#draw name of predicted persondraw_text(img, label_text, rect[0], rect[1]-5)return img#load test imagestest_img1 = cv2.imread("test-data/test1.jpg")test_img2 = cv2.imread("test-data/test2.jpg")#perform a predictionpredicted_img1 = predict(test_img1)predicted_img2 = predict(test_img2)print("Prediction complete")#create a figure of 2 plots (one for each test image)f, (ax1, ax2) = plt.subplots(1, 2, figsize=(10, 5))#display test image1 resultax1.imshow(cv2.cvtColor(predicted_img1, cv2.COLOR_BGR2RGB))#display test image2 resultax2.imshow(cv2.cvtColor(predicted_img2, cv2.COLOR_BGR2RGB))#display both imagescv2.imshow("Tom cruise test", predicted_img1)cv2.imshow("Shahrukh Khan test", predicted_img2)cv2.waitKey(0)cv2.destroyAllWindows()cv2.waitKey(1)cv2.destroyAllWindows()