final.py

css_code = open("code/css.txt","r").read()
html_code = open("code/html.txt","r").read()
java_code = open("code/java.txt","r").read()
python_code = open("code/python.txt","r").read()
js_code = open("code/js.txt","r").read()

docs = []
labels = []

for r in css_code.split('\n'):
    docs.append(r)
    labels.append('css')

for r in html_code.split('\n'):
    docs.append(r)
    labels.append('html')

for r in java_code.split('\n'):
    docs.append(r)
    labels.append('java')

for r in python_code.split('\n'):
    docs.append(r)
    labels.append('python')

# Splitting data into training and test sets
from sklearn.model_selection import train_test_split

docs_train, docs_test, labels_train, labels_test = train_test_split(docs,
                                                                    labels,
                                                                    test_size = 0.25,
                                                                    random_state = 42)


# Tokenizing text with scikit-learn
from sklearn.feature_extraction.text import CountVectorizer

count_vect = CountVectorizer()
X_train_counts = count_vect.fit_transform(docs_train)
print("Shape CountVectorizer on date: ",X_train_counts.shape)

from sklearn.feature_extraction.text import TfidfTransformer

tf_transformer = TfidfTransformer(use_idf=False).fit(X_train_counts)
X_train_tfidf = tf_transformer.transform(X_train_counts)
print("Shape of TfidfTransformer on count_vect: ",X_train_tfidf.shape)

from sklearn.svm import SVC
from sklearn.model_selection import GridSearchCV

# parameters = {'C': [1, 10, 100, 1000], 'gamma': [0.001, 0.0001, 0.00001], 'class_weight':['balanced', None], 'kernel':['linear', 'rbf']}
# classifier_lsvc =  SVC()
# clf_LSVC = GridSearchCV(classifier_lsvc, parameters).fit(X_train_tfidf, labels_train)
# print("Best parameters set found on development set:", clf_LSVC.best_estimator_)

# Saving classifier in pickle
import pickle
# Saving
# save_classifier = open("svm.pickle","wb")
# pickle.dump(clf_LSVC, save_classifier)
# save_classifier.close()
# # loading
classifier_f = open("svm.pickle", "rb")
clf_LSVC = pickle.load(classifier_f)
classifier_f.close()

# Data transformation
X_new_counts = count_vect.transform(docs_test)
X_new_tfidf = tf_transformer.transform(X_new_counts)

testing1_code = ['<h1>']
testing2_code = ['import java.util.']

x_java_testing_counts = count_vect.transform(testing1_code)
x_java_testing_tfidf = tf_transformer.transform(x_java_testing_counts)

x_html_testing_counts = count_vect.transform(testing2_code)
x_html_testing_tfidf = tf_transformer.transform(x_html_testing_counts)

# Testing and Accuracy
from sklearn.metrics import accuracy_score

pred = clf_LSVC.predict(X_new_tfidf)
acc = accuracy_score(pred, labels_test)
print("Accuracy LinearSVC: ", acc*100)

pred_html = clf_LSVC.predict(x_java_testing_tfidf)
pred_java = clf_LSVC.predict(x_html_testing_tfidf)
print("True pred shud be html, but:",pred_html)
print("True pred shud be java, but:",pred_java)