Skip to content

Commit

Permalink
Add files via upload
Browse files Browse the repository at this point in the history 
mainly fixing a characters in scene issue
  • Loading branch information
@StephanosPSteer
StephanosPSteer authored Apr 26, 2023
1 parent 1834f73 commit 652a890
Show file tree
Hide file tree
Showing 3 changed files with 85 additions and 37 deletions.
20 changes: 14 additions & 6 deletions Aubrushli.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -89,6 +89,11 @@ def createshotlist(self, numcharacters, outpath, action_df, df, shot_vals_df, se
(action_df['original_line'].astype(int) <= endline)]

filreact_df = filact_df.reset_index(drop=True)
if 'charsinscene' not in filreact_df:
print('not')
filreact_df.loc[:, 'charsinscene'] = numcharacters
else:
print('ya')
# shot number you create depending on shots

#the headers for the shotlist
Expand Down Expand Up @@ -146,16 +151,13 @@ def createshotlist(self, numcharacters, outpath, action_df, df, shot_vals_df, se
finshot ={}
# do I want drones and extreme wides, probably not indoors
dronesetc = 0
if sending_format==1:
if 'EXT.' in currscenename or currscenename.startswith("."):
#if sending_format==1:
if 'EXT.' in currscenename or currscenename.startswith("."):
dronesetc=1 #thetext in row['element_text']:
# choose shots depending on characters in scene
#print(f_scene_name)
#print(dronesetc)
# this check is all about what shots are needed as per number of characters
filshot_df = shot_vals_df[(shot_vals_df['min_chars'] <=numcharacters) & (shot_vals_df['dronesetc'] <= dronesetc) & (shot_vals_df['dialog'] <= len(currdialog_df))]
finshot_df = filshot_df.reset_index(drop=True)
finshot_df = finshot_df.drop_duplicates()


#elif current_type == 'Action':
#if current_type == '':
Expand All @@ -181,6 +183,12 @@ def createshotlist(self, numcharacters, outpath, action_df, df, shot_vals_df, se
#print(mydesc)
#print(eddesc)
#THESE ARE THE SHOTS

# this check is all about what shots are needed as per number of characters
numcharacters = theaction['charsinscene']
filshot_df = shot_vals_df[(shot_vals_df['min_chars'] <=numcharacters) & (shot_vals_df['dronesetc'] <= dronesetc) & (shot_vals_df['dialog'] <= len(currdialog_df))]
finshot_df = filshot_df.reset_index(drop=True)
finshot_df = finshot_df.drop_duplicates()
for index, shotrow in finshot_df.iterrows():
#otsok = 0
#values = index + 1, currscneno, f_scene_name, shotrow['shot_size'], shotrow['shot_type'], shotrow['AngleOrigin'], 'STATIC or PAN', shotrow['lens'] \
Expand Down
Binary file modified aubrushli.db
View file
Binary file not shown.
102 changes: 71 additions & 31 deletions fdx2aub.py
View file
Original file line number Diff line number Diff line change
@@ -1,4 +1,10 @@
import xml.etree.ElementTree as ET
# tree = ET.parse(r'D:\pythonEnvs\stable_LLM\Big_Fish.fdx')
# root = tree.getroot()

# for child in root:
# print(child.tag, child.attrib)

import sys
import sqlite3
import pandas as pd
Expand All @@ -21,7 +27,7 @@ def createbigcsv(self, csvpath, screenplayid):

sql = "SELECT linenumber, line, linetype, scenenum, length, pagenum, charsinscene FROM fdxlines where screenplayid = ? order by linenumber"
castcsvdf = pd.read_sql(sql, connection, params=(screenplayid,))
print(castcsvdf)
#print(castcsvdf)
#connection.commit()
castcsvdf.to_csv(csvpath, index=False)

Expand All @@ -35,7 +41,7 @@ def writelinetodb(self, connection, screenplayID, line, linenum, linetype, scene


def create_cast_list(self,castlistpath, screenplayid):
print(screenplayid)
#print(screenplayid)

current_dir = os.path.dirname(os.path.abspath(__file__))

Expand All @@ -47,7 +53,7 @@ def create_cast_list(self,castlistpath, screenplayid):

sql = "select line as character, count(line) as lines from fdxlines where linetype = 'character' and screenplayid = ? group by line order by lines desc"
castcsvdf = pd.read_sql(sql, connection, params=(screenplayid,))
print(castcsvdf)
#print(castcsvdf)
#connection.commit()
castcsvdf.to_csv(castlistpath, index=False)

Expand All @@ -60,6 +66,17 @@ def fdx2dbaub(self, input_file_name, shotsdef, outpath):
connection.execute('PRAGMA journal_mode = WAL')
connection.execute('PRAGMA cache_size = -8192')
connection.commit()






#output_file_name = r"D:\pythonEnvs\stable_LLM\test.txt"
#input_file_name = r'G:\Aubrushli_images_current\Big_Fish.fdx'
#shotsdef = r'G:\Aubrushli_images_current\shot_vals.csv'
#outpath = 'G:/Aubrushli_images_current/fdxshotlists/'
#file = open(output_file_name,"w")



Expand Down Expand Up @@ -123,16 +140,17 @@ def fdx2dbaub(self, input_file_name, shotsdef, outpath):
for index, text_element in enumerate(paragraph.findall('Text')):
if (text_element.text is not None):
linetype =''
if "(" in text_element.text or ")" in text_element.text:
if "(" in text_element.text or ")" in text_element.text or "cont’d" in text_element.text:
#charsinscene= charsinscene +1
linetype = 'parenthetical'
else:
#print(text_element.text, linenum)
charsinscene= charsinscene +1
linetype = 'character'
linenum = linenum + 1
# typenum = typenum +1
scenetypenum= scenetypenum + 1
self.writelinetodb(connection, screenplayID, text_element.text.upper(), linenum, linetype, scenenum, scenetypenum, chars=charsinscene )
self.writelinetodb(connection, screenplayID, text_element.text.upper().strip(), linenum, linetype, scenenum, scenetypenum, chars=charsinscene )

elif paragraph.attrib['Type'] == "Parenthetical":
if hasattr(paragraph.find('Text'), 'text'):
Expand All @@ -154,49 +172,64 @@ def fdx2dbaub(self, input_file_name, shotsdef, outpath):
self.writelinetodb(connection, screenplayID, text_element.text, linenum, 'dialogue', scenenum, scenetypenum )
elif paragraph.attrib['Type'] == "Scene Heading":
off= paragraph.get('Number')


scene_properties = paragraph.find('SceneProperties').attrib


# Access the values of the SceneProperties attributes
# Access the values of the SceneProperties attributes
length = scene_properties['Length']
page = scene_properties['Page']




#print(child.attrib)
#print(child.attrib)
if hasattr(paragraph.find('Text'), 'text'):
thetext = ''
for index, text_element in enumerate(paragraph.findall('Text')):
if (text_element.text is not None):
# print(leng)
# print(page)
charsinscene=0
linenum = linenum + 1
thetext = thetext + ' ' + text_element.text.upper()
charsinscene=0
linenum = linenum + 1
# typenum = typenum +1
scenenum = scenenum + 1
scenetypenum=0
self.writelinetodb(connection, screenplayID, text_element.text.upper() , linenum, 'scene', scenenum, scenetypenum, off, length, page, chars=charsinscene)
scenenum = scenenum + 1
scenetypenum=0
self.writelinetodb(connection, screenplayID, thetext , linenum, 'scene', scenenum, scenetypenum, off, length, page, chars=charsinscene)


#*************************steph need a characters in scene update here to distinct characters*************************************
cursor = connection.cursor()
sql = "CREATE TEMP TABLE distchars as select distinct line, scenenum from fdxlines where linetype = 'character' group by scenenum, line order by scenenum"
sql1 ="create temp table finchars as select count(line) thechars, scenenum from distchars group by scenenum"
sql2 = "update fdxlines set charsinscene = f.thechars from (select thechars, scenenum from finchars) as f where f.scenenum = fdxlines.scenenum"
sql3 = "INSERT INTO characters (screenplayID, charactername) select distinct screenplayid, trim(line) from fdxlines where linetype = 'character'"
cursor.execute(sql)
cursor.execute(sql1)
cursor.execute(sql2)
cursor.execute(sql3)
connection.commit()
cursor = connection.cursor()
sql4 = "SELECT characterID, screenplayID, charactername FROM characters where screenplayid =? "
params = (screenplayID,) # create a tuple of parameters
cursor.execute(sql4, params) # pass the tuple to cursor.execute()
for row in cursor:
print(row[1], row[2])
cursor1 = connection.cursor()
sql5= "INSERT INTO characteractions (characterID, actionID, charactername, theline) select ? , fdxlineID , ? , line from fdxlines where screenplayid =? and linetype = 'action' and (line like '% ' || ? || '%' or line like ? || '%') "
params1 = (row[0], row[2], screenplayID, row[2], row[2])
cursor1.execute(sql5, params1)
#for newrow in cursor1:
# print(newrow)
#b= row
cursor = connection.cursor()
sql6= "update fdxlines set actchars = c.actchars from (select count(actionid) as actchars, actionid FROM characteractions group by actionid, theline) as c where c.actionid = fdxlines.fdxlineid"
cursor.execute(sql6)
connection.commit()




# need to loop round each scene in db for shots call aubrushli.py




# need a complete DF of the fdxlines table in the same format as fountain
# need to put an id in fdxlines to match to the screenplay you need.
sql = 'SELECT line, linetype, linenumber, scenenum, charsinscene from fdxlines where screenplayID=?'
sql = 'SELECT line, linetype, linenumber, scenenum, case when actchars is null then charsinscene else actchars end as charsinscene from fdxlines where screenplayID=?'
df = pd.read_sql(sql, connection, params=(screenplayID,))

df = df.rename(columns={'line': 'element_text', 'linetype': 'element_type', 'linenumber': 'original_line'})
Expand All @@ -219,10 +252,10 @@ def fdx2dbaub(self, input_file_name, shotsdef, outpath):
scenedf = scenedf.reset_index(drop=True)
currscenenamedf = scenedf[(scenedf['element_type']== 'scene')]
currscenenamedf = currscenenamedf.reset_index(drop=True)
print(currscenenamedf)
#print(currscenenamedf)
if len(currscenenamedf) >0:
curscenename = currscenenamedf['element_text'][0]
print(curscenename)
#print(curscenename)
charsinscene = scenedf['charsinscene'][0]
currsceneno = scenedf['scenenum'][0]

Expand All @@ -235,17 +268,24 @@ def fdx2dbaub(self, input_file_name, shotsdef, outpath):
endline = scenedf['original_line'].max()
combined=[]

#print(scene[0])
print(scenedf)
print(actionsdf)
print(startline)
print(endline)
print(charsinscene)
# #print(scene[0])
# print(scenedf)
# print(actionsdf)
# print(startline)
# print(endline)
# print(charsinscene)

aubshot = Aubrushli()
aubshot.createshotlist( charsinscene, outpath, actionsdf, df, shot_vals_df, 2, startline, endline, currsceneno, curscenename, combined, 'Big_FDX_Fish')

#createshotlist( numcharacters, outpath, action_df, df, shot_vals_df, sending_format=1, startline=f_start_scene_line, endline=f_next_scene_line, currscene=currscneno, currscenename=f_scene_name, combined=combine_chars):





# also need a start and end line for the scene
# use the sending format param to allow extra

connection.close()

0 comments on commit 652a890

Please sign in to comment.