454 lines
14 KiB
Python
Executable File
454 lines
14 KiB
Python
Executable File
# import textract
|
||
from tqdm import tqdm
|
||
import sys
|
||
import re
|
||
import docx
|
||
import os
|
||
doc = docx.Document()
|
||
from docx.shared import Inches, Cm, Pt
|
||
from docx.enum.text import WD_ALIGN_PARAGRAPH
|
||
from docx.enum.table import WD_TABLE_ALIGNMENT, WD_ALIGN_VERTICAL
|
||
from collections import Counter
|
||
|
||
|
||
from script_detector import script_cat
|
||
|
||
#google
|
||
#os.environ["GOOGLE_APPLICATION_CREDENTIALS"]="gifted-mountain-318504-0a5f94cda0c8.json"
|
||
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = rf"{basePath}/conversion/gifted-mountain-318504-4f001d5f08db.json"
|
||
from google.cloud import translate
|
||
from google.cloud import translate_v2 as Translate
|
||
translate_client = Translate.Client()
|
||
client = translate.TranslationServiceClient()
|
||
project_id = 'authentic-bongo-272808'
|
||
location = "global"
|
||
parent = f"projects/{project_id}/locations/{location}"
|
||
|
||
|
||
slug_pattern= r'[\d]*[.]?[\s]*[IE][NX]T'
|
||
pat = r'[\d]*[\s]*[IE]/[IE][.]?'
|
||
transitions = ['CUT TO:','FADE IN:','FADE OUT:','DISSOLVE TO:','JUMP CUT TO:','JUMP TO:','CUT BACK TO:','INTERCUT WITH:','I/C WITH:','BACK TO:', 'INTERVAL']
|
||
reserved_words = ['MONTAGE','PBS','FADE','FADE','TITLE','SPLIT', 'SCREEN','CUT']
|
||
style = doc.styles['Normal']
|
||
font = style.font
|
||
font.name = 'Courier New'
|
||
font.size = Pt(12)
|
||
|
||
def breaksen(s):
|
||
l =[]
|
||
#if len(s.split())<=256:
|
||
if len(s.split())<=256:
|
||
l.append(s)
|
||
else:
|
||
n = len(s.split())
|
||
for i in range(n//32 + 1):
|
||
l.append(" ".join(s.split()[32*i:32*(i+1)]))
|
||
return l
|
||
|
||
def getRefined(filename1):
|
||
#print("get_refined_called")
|
||
total_scenes = 0
|
||
text = textract.process(filename1, encoding="utf8", errors='ignore')
|
||
filename="file.txt"
|
||
f=open(filename, 'wb')
|
||
f.write(text)
|
||
f.close()
|
||
dialog_coming=False
|
||
f=open(filename, 'r', encoding="utf8", errors='ignore')
|
||
doc11=f.read()
|
||
|
||
f.close()
|
||
f1=open("file1.txt",'w', encoding="utf8", errors='ignore')
|
||
c=0
|
||
flag=False
|
||
for line in doc11.split("\n"):
|
||
if (line.strip().startswith(('INT.','INT ')) or \
|
||
line.strip().startswith(('I/E','E/I')) or \
|
||
line.strip().startswith(('EXT.','EXT ')) or \
|
||
line.strip().startswith('EXT/INT') or \
|
||
line.strip().startswith('INT/EXT') or \
|
||
re.match(slug_pattern,line.strip())):
|
||
|
||
flag=True
|
||
f1.write(line)
|
||
f1.write('\n')
|
||
continue
|
||
else:
|
||
line = line.strip()
|
||
if flag:
|
||
if line.strip()=='\n':
|
||
continue
|
||
if dialog_coming and (line=='\n' or line.strip()==""):
|
||
continue
|
||
if dialog_coming:
|
||
f1.write(line)
|
||
f1.write('\n')
|
||
if re.match(r"\(.*\)",line):
|
||
|
||
continue
|
||
else:
|
||
dialog_coming=False
|
||
continue
|
||
continue
|
||
if line.isupper() and re.fullmatch(r"([A-Z'’]+[\s]*[-]*)*([#]*[\s]*[1-9])*(\(.*\))*",line.strip()):
|
||
f1.write(line)
|
||
f1.write('\n')
|
||
dialog_coming=True
|
||
continue
|
||
if not line=='\n':
|
||
f1.write(line)
|
||
f1.write('\n')
|
||
|
||
|
||
f1.close()
|
||
filename1="file1.txt"
|
||
#file.txt contains the data of file1.txt , no usage as of now may be change the mame of the file.txt to file1.txt
|
||
text = textract.process(filename1, encoding="utf8", errors='ignore')
|
||
|
||
filename="file.txt"
|
||
_, file_extension = os.path.splitext(filename1)
|
||
f=open(filename, 'wb')
|
||
f.write(text)
|
||
f.close()
|
||
|
||
with open(filename, "r") as input:
|
||
input_ = input.read().split('\n\n')
|
||
|
||
refined=[]
|
||
|
||
for line in input_:
|
||
refined.append(line.strip())
|
||
refined=list(filter(lambda a: a != "", refined))
|
||
#print("processing the script")
|
||
|
||
for i in range(len(refined)):
|
||
if not (refined[i].strip().startswith(('INT.','INT ')) or refined[i].strip().startswith(('EXT.','EXT ')) or refined[i].strip().startswith('EXT/INT') or refined[i].strip().startswith('INT/EXT') or refined[i].strip().startswith(('I/E','E/I')) or re.match(slug_pattern,refined[i].strip()) or re.match(pat,refined[i].strip())):
|
||
total_scenes = total_scenes + 1
|
||
continue
|
||
refined=refined[i:]
|
||
break
|
||
# refined.append(line.strip())
|
||
|
||
refined=list(filter(lambda a: a != "", refined))
|
||
return refined,total_scenes
|
||
|
||
def getSlugAndNonSlug(refined):
|
||
sluglines=[]
|
||
without_slug=[]
|
||
for para in refined:
|
||
para=para.strip()
|
||
if para.strip().startswith(('INT.','INT')) or para.strip().startswith(('EXT.','EXT')) or para.strip().startswith('EXT/INT') or para.strip().startswith(('I/E','E/I')) or para.strip().startswith('INT/EXT') or re.match(slug_pattern,para.strip()) or re.match(pat,para.strip()):
|
||
sluglines.append(para)
|
||
continue
|
||
without_slug.append(para)
|
||
return sluglines,without_slug
|
||
|
||
def getSpeakers(without_slug):
|
||
characters=[]
|
||
for para in without_slug:
|
||
lis=para.split('\n')
|
||
i=0
|
||
for item in lis:
|
||
i=i+1
|
||
i=min(i,len(lis)-2)
|
||
if item.isupper() and not(lis[i+1].strip()==""):
|
||
if re.match(r"[A-Z'’]+[\s]*[-]*[A-Z'’]*([#]*[\s]*[1-9])*(\(.*\))*",item):
|
||
tem = item.split("(")[0].strip()
|
||
characters.append(tem.strip())
|
||
else:
|
||
continue
|
||
|
||
characters=list(set(characters))
|
||
characters=list(filter(lambda x: len(x) >0,characters))
|
||
characters = [character for character in characters if set(character.split(" ")).intersection(reserved_words) == set()]
|
||
return characters
|
||
|
||
|
||
def getScenes(refined,total_scenes,characters):
|
||
# To find scenes data structure and prev and next scenes numbers
|
||
i=0
|
||
scene=[]
|
||
dialogues=[]
|
||
speakers=[]
|
||
slugline_dic={}
|
||
prev_dial_speaker=""
|
||
next_dial_speaker=""
|
||
pc=0
|
||
scene_no=0
|
||
actionline=[]
|
||
successor_scene_no=0
|
||
predecessor_scene_no=0
|
||
parenthetical_lis=[]
|
||
|
||
scenes=[]
|
||
speaker=""
|
||
parenthetical='NONE'
|
||
patttern=r'[\d]*[.]?[\s]*[IE][NX]T'
|
||
for line in refined:
|
||
if line.strip().startswith(('INT.','INT')) or line.strip().startswith(('EXT.','EXT')) or line.strip().startswith('EXT/INT') or line.strip().startswith('INT/EXT') or line.strip().startswith(('I/E','E/I')) or re.match(slug_pattern,line.strip()) or re.match(pat,line.strip()):
|
||
scenes.append(scene)
|
||
scene=[]
|
||
i=0
|
||
scene_no+=1
|
||
scene.append(line)
|
||
slugline_dic[scene_no]=line.split("\n")[0].strip('0123456789.- ')
|
||
|
||
else:
|
||
lis=line.split("\n")
|
||
lis=[l.strip() for l in lis]
|
||
word=lis[0]
|
||
if word.split('(')[0].strip() in characters:
|
||
mydic={}
|
||
prev_dial_speaker=speaker
|
||
speaker=word.split('(')[0].strip()
|
||
if len(lis)>1 and re.match(r"\(.*\)",lis[1]):
|
||
pc=pc+1
|
||
parenthetical=lis[1]
|
||
parenthetical=parenthetical.replace("\n","")
|
||
dia=' '.join(lis[2:])
|
||
dia=dia.replace("\n","")
|
||
##renu
|
||
dia=dia.replace("\"", '')
|
||
|
||
else:
|
||
dia=''.join(lis[1:])
|
||
dia=dia.replace("\n","")
|
||
dia=dia.replace("\"", '')
|
||
if not (len(dia)==0 and parenthetical=="NONE"):
|
||
|
||
if i-1 >= 0:
|
||
try:
|
||
prev=main_lis[scene_no-1][i-1]
|
||
except:
|
||
prev=""
|
||
else:
|
||
prev=""
|
||
try:
|
||
next=main_lis[scene_no-1][i+1]
|
||
except:
|
||
next=""
|
||
#prev is previous speaker and next is next speaker of the dialogue
|
||
mydic[speaker]=[parenthetical,scene_no,dia,len(dia),prev,next]
|
||
# print(mydic)
|
||
prev,next="",""
|
||
i=i+1
|
||
speakers.append(speaker)
|
||
parenthetical_lis.append(parenthetical)
|
||
dialogues.append(mydic)
|
||
scene.append(mydic)
|
||
parenthetical="NONE"
|
||
else:
|
||
line=line.replace("\n"," ")
|
||
line=' '.join(line.split())
|
||
if line.strip() in transitions:
|
||
scene.append({'Transition':line.strip()})
|
||
continue
|
||
actionline.append(line)
|
||
scene.append(line.strip())
|
||
|
||
|
||
scenes.append(scene)
|
||
speakers=list(set(speakers))
|
||
scenes=scenes[1:]
|
||
s = []
|
||
for scene in scenes:
|
||
s1=[]
|
||
for ele in scene:
|
||
if type(ele) == type(""):
|
||
s1.extend(ele.split("\n"))
|
||
else:
|
||
s1.append(ele)
|
||
s.append(s1)
|
||
return s,actionline,parenthetical_lis,speakers,dialogues
|
||
|
||
|
||
filename1 = sys.argv[1]
|
||
#print(filename1)
|
||
|
||
refined,total_scenes = getRefined(filename1)
|
||
#print(refined)
|
||
sluglines,without_slug = getSlugAndNonSlug(refined)
|
||
characters = getSpeakers(without_slug)
|
||
scenes,actionline,parenthetical_lis,speakers,dialogues = getScenes(refined,total_scenes,characters)
|
||
#print(scenes)
|
||
|
||
def language_detector(text):
|
||
result = translate_client.translate(text, target_language='hi')
|
||
det_lang = result["detectedSourceLanguage"]
|
||
return det_lang
|
||
|
||
def script_det(text):
|
||
punctuations = '''!()-[]{};:'"\,<>./?@#$%^&*_~'''
|
||
no_punct = ""
|
||
for char in text:
|
||
if char not in punctuations:
|
||
no_punct = char
|
||
break
|
||
script = script_cat(no_punct)[0]
|
||
return script
|
||
|
||
'''
|
||
A. Language of Highest number of full dialogues,
|
||
B. Numbers of dialogues in action line language,
|
||
C. Number of dialogues in other languages)
|
||
'''
|
||
|
||
def A_B_C(dialogue_language, non_dial_src_lang):
|
||
dict1 = dict(Counter(dialogue_language))
|
||
sorted_values = sorted(dict1.values(), reverse=True) # Sort the values
|
||
sorted_dict = {}
|
||
for i in sorted_values:
|
||
for k in dict1.keys():
|
||
if dict1[k] == i:
|
||
sorted_dict[k] = dict1[k]
|
||
sources = list(sorted_dict.keys())
|
||
A = sources[0]
|
||
if len(sources)!=1:
|
||
B = sorted_values[sources.index(non_dial_src_lang)]
|
||
C = sum(sorted_values[2:])
|
||
else:
|
||
B=0
|
||
C=0
|
||
return A, B, C
|
||
|
||
|
||
#print(scenes)
|
||
|
||
dialogue_language = []
|
||
count =0
|
||
for scene in tqdm(scenes[:]):
|
||
#print("scene")
|
||
|
||
for i,line in enumerate(scene):
|
||
if i == 0:
|
||
continue
|
||
if type(line)==type(""):
|
||
#print("here")
|
||
if count==0:
|
||
#print(line)
|
||
non_dial_src_lang = language_detector(line)
|
||
count+=1
|
||
#print("non_dial_src_lang", non_dial_src_lang)
|
||
|
||
else:
|
||
#print("line", line)
|
||
[speaker] = line.keys()
|
||
#print([speaker])
|
||
if speaker == 'Transition':
|
||
continue
|
||
|
||
#print("dial", line[speaker][2])
|
||
dial_src_lang = language_detector(line[speaker][2])
|
||
dialogue_language.append(dial_src_lang)
|
||
|
||
#dial_src_script = script_det(line[speaker][2])
|
||
|
||
# print("non_dial_src_lang", non_dial_src_lang)
|
||
# print("dial_src_lang", dialogue_language)
|
||
|
||
|
||
#print(len(dialogue_language))
|
||
#print(Counter(dialogue_language))
|
||
|
||
A, B, C = A_B_C(dialogue_language, non_dial_src_lang)
|
||
# print("A = {} B = {} C = {}".format(A, B, C))
|
||
|
||
|
||
def dial_each_word_lang1(non_dial_src_lang, dial):
|
||
for word in dial.split():
|
||
if language_detector(word)==non_dial_src_lang:
|
||
print("word", word)
|
||
return "True"
|
||
return "False"
|
||
|
||
def dial_each_word_lang2(non_dial_src_lang, A, dial ):
|
||
for word in dial.split():
|
||
if (language_detector(word)!=non_dial_src_lang) or (language_detector(word)!=A):
|
||
print("in 4")
|
||
print("word", word)
|
||
return "True"
|
||
return "False"
|
||
|
||
def word_with_actionline(scenes):
|
||
|
||
for scene in tqdm(scenes[:]):
|
||
for i,line in enumerate(scene):
|
||
if i == 0:
|
||
continue
|
||
if type(line)==type(""):
|
||
continue
|
||
|
||
else:
|
||
[speaker] = line.keys()
|
||
if speaker == 'Transition':
|
||
continue
|
||
|
||
dial_src_lang = language_detector(line[speaker][2])
|
||
if dial_src_lang==A:
|
||
word_lang_with_actionline = dial_each_word_lang1(non_dial_src_lang, line[speaker][2])
|
||
|
||
if word_lang_with_actionline == "True":
|
||
return word_lang_with_actionline
|
||
|
||
|
||
def word_with_other(scenes):
|
||
|
||
for scene in tqdm(scenes[:]):
|
||
for i,line in enumerate(scene):
|
||
if i == 0:
|
||
continue
|
||
if type(line)==type(""):
|
||
continue
|
||
|
||
else:
|
||
[speaker] = line.keys()
|
||
if speaker == 'Transition':
|
||
continue
|
||
|
||
dial_src_lang = language_detector(line[speaker][2])
|
||
|
||
if dial_src_lang==A:
|
||
word_lang_with_other = dial_each_word_lang2(non_dial_src_lang, A, line[speaker][2])
|
||
|
||
if word_lang_with_other == "True":
|
||
return word_lang_with_other
|
||
|
||
word_lang_with_actionline = word_with_actionline(scenes)
|
||
#print(word_lang_with_actionline)
|
||
|
||
word_lang_with_other = word_with_other(scenes)
|
||
#print(word_lang_with_other)
|
||
|
||
|
||
####
|
||
print("actionline_lanuge", non_dial_src_lang)
|
||
#print("A = {} B = {} C = {}".format(A, B, C))
|
||
print("dial_language", A)
|
||
|
||
if B>0:
|
||
print("UI option3 - yes" )
|
||
else:
|
||
print("UI option3 - no" )
|
||
|
||
if C>0:
|
||
print("UI option4 - yes" )
|
||
else:
|
||
print("UI option4 - no" )
|
||
|
||
if word_lang_with_actionline=="True":
|
||
print("UI option5 - Yes")
|
||
else:
|
||
print("UI_option5 - NO")
|
||
|
||
if word_lang_with_other=="True":
|
||
print("UI option6 - Yes")
|
||
else:
|
||
print("UI option6 - No")
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|