def findSourceDoc(self):
root.filename = tkinter.filedialog.askopenfilename(
initialdir=os.path.dirname(os.path.realpath(__file__)),
title="Select file to search",
filetypes=(("txt files", "*.txt"), ("all files", "*.*")))
if root.filename == '':
print("cancelled")
else:
srcCandidate, dupl = FindSourceDoc.run(root.filename,
root,
filter=False,
filamt=0)
try:
ParseXML.run(root.filename)
except:
print("already in annotation")
dicc = g.openResult('output/annotation.csv')
annoSource = dicc.get(ntpath.basename(root.filename[:-4]))[0]
annoDup = dicc.get(ntpath.basename(root.filename[:-4]))[1]
p, r, f = g.allmeasure(srcCandidate, dupl, annoSource, annoDup)
print(p)
print(r)
print(f)
result = tkinter.Toplevel(root)
result.minsize(200, 200)
result.title("Evaluation " + ntpath.basename(root.filename))
tkinter.Label(result, text="Precision ").grid(row=0,
column=0,
padx=3)
tkinter.Label(result, text="Recall ").grid(row=1, column=0, padx=3)
tkinter.Label(result, text="F1 Score ").grid(row=2,
column=0,
padx=3)
tkinter.Label(result, text=round(p, 3)).grid(row=0,
column=1,
padx=3)
tkinter.Label(result, text=round(r, 3)).grid(row=1,
column=1,
padx=3)
tkinter.Label(result, text=round(f, 3)).grid(row=2,
column=1,
padx=3)
def run(filename):
xmldoc = minidom.parse(filename[:-4]+'.xml')
if "simulatedFiles" in filename:
plg = xmldoc.getElementsByTagName('plagiarized')
srcs = []
for x in plg:
print(x.attributes['sourceid'].value)
y = x.attributes['sourceid'].value
if str(y)+'.txt' not in srcs:
srcs.append(str(y)+'.txt')
else:
plg = xmldoc.getElementsByTagName('features')
srcs = []
for x in plg:
print(x.attributes['source_reference'].value)
y = x.attributes['source_reference'].value
if str(y) + '.txt' not in srcs:
srcs.append(str(y) + '.txt')
if 'testdoc2' in filename:
dupesDict = GlobalFunctions.openFiles('duplicate/FirstDuplicates.csv')
else:
dupesDict = GlobalFunctions.openFiles('duplicate/NumericDuplicate.csv')
dupes = []
for dupe in dupesDict:
for x in srcs:
if x in dupesDict.get(dupe) and dupe not in srcs:
dupes.append(dupe)
for x in srcs:
print(dupesDict.get(x))
try:
if len(dupesDict.get(x)) > 1:
for dupe in dupesDict.get(x):
if dupe not in dupes:
dupes.append(dupe)
else:
try:
if dupesDict.get(x)[0] not in dupes:
dupes = dupes + dupesDict.get(x)
except:
continue
except:
print("not in dictionary")
print(dupes)
tupleS = [srcs, dupes]
print(tupleS)
try:
dicc = GlobalFunctions.openResult('output/annotation.csv')
dicc[ntpath.basename(filename[:-4])] = tupleS
except:
dicc = {ntpath.basename(filename[:-4]):tupleS}
with open('output/annotation.csv', 'w', encoding='utf-8', newline='')as csvfile:
fieldname = ['TestDocs', 'Source','Duplicates']
writer = csv.DictWriter(csvfile, fieldnames=fieldname)
writer.writeheader()
for x in dicc:
try:
writer.writerow({'TestDocs':x, 'Source':dicc.get(x)[0],'Duplicates':dicc.get(x)[1]})
except:
writer.writerow({'TestDocs': x, 'Source': dicc.get(x)[0]})
csvfile.close()
def run(filename, root, show = True, filter = True, filamt = 5):
dicc = GlobalFunctions.openFiles('PLs/CompiledPLs/postingList.csv')
df = GlobalFunctions.openFiles('PLs/dictionary/documentFreq.csv')
testTF, queries = TestDocProcessing.queryExtract(filename)
querySources = []
counter = 0
if show:
win = tkinter.Toplevel(root)
win.minsize(200, 200)
win.title("Search Query for "+ntpath.basename(filename))
c=0
for que in queries:
r=0
counter +=1
querySources.append(getSource(testTF,que,dicc,df))
if show:
tkinter.Label(win, text = "Search Query "+str(counter)).grid(row=r, column = c, padx = 3)
r+=1
try:
for q in que:
if show:
tkinter.Label(win, text=q).grid(row=r, column = c, padx= 3)
r +=1
except:
print("unable to call new window")
c += 1
if show:
qwin = tkinter.Toplevel(win)
qwin.minsize(200, 200)
qwin.title("Query Result for " + ntpath.basename(filename))
c=0
counter = 0
for src in querySources:
r=0
counter+=1
tkinter.Label(qwin, text = "Query Result "+str(counter)).grid(row=r, column = c, padx = 3)
tkinter.Label(qwin, text="Percentage " + str(counter)).grid(row=r, column=c+1, padx=3)
r+=1
try:
for s in src:
tkinter.Label(qwin, text = s[0]).grid(row=r, column = c, padx = 3)
tkinter.Label(qwin, text = round(s[1]*100,2)).grid(row=r,column = c+1, padx=3)
r+=1
except:
print("no val")
c+=2
result =[]
for n in querySources:
for x in n:
result.append(x)
result.sort(key=operator.itemgetter(1))
result.reverse()
filters = []
copyResult = result.copy()
result = []
for x in copyResult:
if x[0] not in filters:
filters.append(x[0])
result.append(x)
print(result)
if 'testdoc2' in filename:
dupes = GlobalFunctions.openFiles('duplicate/FirstDuplicates.csv')
else:
dupes = GlobalFunctions.openFiles('duplicate/NumericDuplicate.csv')
duplicate = []
copyRes = result.copy()
result = []
for res in copyRes:
result.append(res[0])
for res in result.copy():
try:
for x in dupes.get(res):
if x not in duplicate and res not in duplicate:
duplicate.append(x)
result.remove(x)
except:
pass
if 'testdoc2' in filename:
dupes = GlobalFunctions.openFiles('duplicate/NumericDuplicate.csv')
else:
dupes = GlobalFunctions.openFiles('duplicate/FirstDuplicates.csv')
for res in result.copy():
try:
for x in dupes.get(res):
if x not in duplicate and res not in duplicate:
duplicate.append(x)
result.remove(x)
except:
pass
print(result)
print(duplicate)
if filter:
result = result[:filamt]
if 'testdoc2' in filename:
dupes = GlobalFunctions.openFiles('duplicate/FirstDuplicates.csv')
else:
dupes = GlobalFunctions.openFiles('duplicate/NumericDuplicate.csv')
duplicate = []
for res in result:
try:
for x in dupes.get(res):
duplicate.append(x)
except:
pass
if show:
fres = tkinter.Toplevel(qwin)
fres.minsize(200, 200)
fres.title("Source Candidate for " + ntpath.basename(filename))
c = 0
r = 0
dr = 0
counter += 1
tkinter.Label(fres, text="Query Result").grid(row=r, column=c, padx=3)
r += 1
try:
for s in result:
tkinter.Label(fres, text=s).grid(row=r, column=c, padx=3)
r += 1
for d in duplicate:
dr+=1
tkinter.Label(fres, text="Duplicate").grid(row=0, column=c+1, padx=3)
tkinter.Label(fres,text = d).grid(row=dr, column=c+1, padx=3)
except:
print("no val")
if not os.path.exists('output/'):
os.makedirs('output/')
copyResult = []
for res in result.copy():
copyResult.append(res)
copyDupe = []
for dupes in duplicate.copy():
copyDupe.append(dupes)
tupleS = [copyResult, copyDupe]
print(tupleS)
print(result)
print(duplicate)
if filter and filamt == 5:
saveResult = 'output/resultsFilter5.csv'
elif filter and filamt ==10:
saveResult = 'output/resultsFilter10.csv'
else:
saveResult = 'output/resultsNoFil.csv'
try:
dicc = GlobalFunctions.openResult(saveResult)
dicc[ntpath.basename(filename[:-4])] = tupleS
except:
dicc = {ntpath.basename(filename[:-4]):tupleS}
if filter:
with open('output/resultsFilter'+str(filamt)+'.csv', 'w', encoding='utf-8', newline='')as csvfile:
fieldname = ['TestDocs', 'Source_Candidate','Duplicates']
writer = csv.DictWriter(csvfile, fieldnames=fieldname)
writer.writeheader()
for x in dicc:
try:
writer.writerow({'TestDocs':x, 'Source_Candidate':dicc.get(x)[0],'Duplicates':dicc.get(x)[1]})
except:
writer.writerow({'TestDocs': x, 'Source_Candidate': dicc.get(x)[0]})
csvfile.close()
else:
with open('output/resultsNoFil.csv', 'w', encoding='utf-8', newline='')as csvfile:
fieldname = ['TestDocs', 'Source_Candidate','Duplicates']
writer = csv.DictWriter(csvfile, fieldnames=fieldname)
writer.writeheader()
for x in dicc:
try:
writer.writerow({'TestDocs':x, 'Source_Candidate':dicc.get(x)[0],'Duplicates':dicc.get(x)[1]})
except:
writer.writerow({'TestDocs': x, 'Source_Candidate': dicc.get(x)[0]})
csvfile.close()
return copyResult,copyDupe
def simul(filteramt, file):
if file == "Simulated":
file = glob.glob("testdoc/simulatedFiles/*.txt")
else:
file = glob.glob("testdoc/artificialFile/*.txt")
#print(file)
#print(filteramt)
precision = []
recall = []
F1 = []
docname = []
count = 0
for fi in file:
count += 1
docname.append(count)
if fi == '':
print('cancelled')
else:
if filteramt == 0:
srcCandidate, dupl = FindSourceDoc.run(fi,
root,
show=False,
filter=False,
filamt=0)
else:
srcCandidate, dupl = FindSourceDoc.run(fi,
root,
show=False,
filamt=filteramt)
try:
ParseXML.run(fi)
except:
print("already in annotation")
dicc = g.openResult('output/annotation.csv')
annoSource = dicc.get(ntpath.basename(fi[:-4]))[0]
annoDup = dicc.get(ntpath.basename(fi[:-4]))[1]
p, r, f = g.allmeasure(srcCandidate, dupl, annoSource, annoDup)
precision.append(p)
recall.append(r)
F1.append(f)
plt.subplot(2, 2, 1)
plt.bar(docname, precision)
plt.title('Precision = ' + str(round(sum(precision) / len(precision), 3)))
plt.ylim(top=1.05)
plt.subplot(2, 2, 2)
plt.bar(docname, recall)
plt.title('Recall = ' + str(round(sum(recall) / len(recall), 3)))
plt.ylim(top=1.05)
plt.subplot(2, 2, 3)
plt.bar(docname, recall)
plt.title('F1 Score = ' + str(round(sum(F1) / len(F1), 3)))
plt.ylim(top=1.05)
plt.subplots_adjust(hspace=0.3)
plt.show()
