def scansion(self, meter=None, conscious=False):
bp = self.bestParse(meter)
# if not bp: return
lowestScore = ""
str_ot = ""
if bp:
str_ot = bp.str_ot()
lowestScore = bp.score()
from tools import makeminlength
self.om(
"\t".join(
[
str(x)
for x in [
makeminlength(str(self), being.linelen),
makeminlength(str(bp) if bp else "", being.linelen),
len(self.allParses(meter)),
lowestScore,
str_ot,
]
]
),
conscious=conscious,
)
def store_stats(self, meter, init):
textname = init.getName()
if not textname:
textname = str(self).replace(" ", "_")
## store stats
if not textname in init.meter_stats["lines"]:
init.meter_stats["lines"][textname] = {}
if not textname in init.meter_stats["positions"]:
init.meter_stats["positions"][textname] = {}
if not textname in init.meter_stats["texts"]:
init.meter_stats["texts"][textname] = {}
if not textname in init.meter_stats["_ot"]:
init.meter_stats["_ot"][textname] = makeminlength("line", being.linelen) + "\tmeter\t" + init.ckeys + "\n"
try:
parsedat = []
for k, v in sorted(self.__bestparse[meter.id].constraintScores.items()):
if not k in init.meter_stats["texts"][textname]:
init.meter_stats["texts"][textname][k] = []
init.meter_stats["texts"][textname][k].append(v)
# parsedat.append(v/len(self.__bestparse.positions)) #???
parsedat.append(v)
linekey = (
str(len(init.meter_stats["lines"][textname]) + 1).zfill(6)
+ "_"
+ str(self.__bestparse[meter.id].posString())
)
init.meter_stats["lines"][textname][linekey] = parsedat
## OT stats
parses = self.__parses[meter.id]
init.meter_stats["_ot"][textname] += (
makeminlength(str(self), being.linelen) + "\t" + parses[0].str_ot() + "\n"
)
if len(parses) > 1:
for parse in parses[1:]:
init.meter_stats["_ot"][textname] += makeminlength("", being.linelen) + "\t" + parse.str_ot() + "\n"
for posn in range(len(self.__bestparse[meter.id].positions)):
pos = self.__bestparse[meter.id].positions[posn]
(posdat, ckeys) = pos.formatConstraints(normalize=True, getKeys=True)
for cnum in range(len(ckeys)):
if not posn in init.meter_stats["positions"][textname]:
init.meter_stats["positions"][textname][posn] = {}
if not ckeys[cnum] in init.meter_stats["positions"][textname][posn]:
init.meter_stats["positions"][textname][posn][ckeys[cnum]] = []
init.meter_stats["positions"][textname][posn][ckeys[cnum]].append(posdat[cnum])
except (IndexError, KeyError, AttributeError) as e:
# print "!! no lines successfully parsed with any meter"
pass
def store_stats(self, meter, init):
textname = init.getName()
if not textname: textname = str(self).replace(" ", "_")
## store stats
if (not textname in init.meter_stats['lines']):
init.meter_stats['lines'][textname] = {}
if (not textname in init.meter_stats['positions']):
init.meter_stats['positions'][textname] = {}
if (not textname in init.meter_stats['texts']):
init.meter_stats['texts'][textname] = {}
if (not textname in init.meter_stats['_ot']):
init.meter_stats['_ot'][textname] = makeminlength(
"line", being.linelen) + "\tmeter\t" + init.ckeys + "\n"
parsedat = []
for k, v in sorted(self.__bestparse[meter].constraintScores.items()):
if (not k in init.meter_stats['texts'][textname]):
init.meter_stats['texts'][textname][k] = []
init.meter_stats['texts'][textname][k].append(v)
#parsedat.append(v/len(self.__bestparse.positions)) #???
parsedat.append(v)
linekey = str(len(init.meter_stats['lines'][textname]) +
1).zfill(6) + "_" + str(
self.__bestparse[meter].posString())
init.meter_stats['lines'][textname][linekey] = parsedat
## OT stats
parses = self.__parses[meter]
init.meter_stats['_ot'][textname] += makeminlength(
str(self), being.linelen) + "\t" + parses[0].str_ot() + "\n"
if len(parses) > 1:
for parse in parses[1:]:
init.meter_stats['_ot'][textname] += makeminlength(
"", being.linelen) + "\t" + parse.str_ot() + "\n"
for posn in range(len(self.__bestparse[meter].positions)):
pos = self.__bestparse[meter].positions[posn]
(posdat, ckeys) = pos.formatConstraints(normalize=True,
getKeys=True)
for cnum in range(len(ckeys)):
if (not posn in init.meter_stats['positions'][textname]):
init.meter_stats['positions'][textname][posn] = {}
if (not ckeys[cnum]
in init.meter_stats['positions'][textname][posn]):
init.meter_stats['positions'][textname][posn][
ckeys[cnum]] = []
init.meter_stats['positions'][textname][posn][
ckeys[cnum]].append(posdat[cnum])
def scansion(self, meter=None, conscious=False):
bp = self.bestParse(meter)
lowestScore = bp.score()
from tools import makeminlength
self.om("\t".join([
str(x) for x in [
makeminlength(str(self), being.linelen),
makeminlength(str(bp), being.linelen),
len(self.allParses(meter)), lowestScore,
bp.str_ot()
]
]),
conscious=conscious)
def scansion(self,meter=None,conscious=False):
bp=self.bestParse(meter)
config=being.config
#if not bp: return
lowestScore=''
str_ot=''
count=''
meterstr=''
if bp:
meterstr=bp.str_meter()
str_ot=bp.str_ot()
lowestScore=bp.score()
count=bp.totalCount
from tools import makeminlength
self.om("\t".join( [str(x) for x in [makeminlength(str(self),config['linelen']), makeminlength(str(bp) if bp else '', config['linelen']),meterstr,len(self.allParses(meter)),count,lowestScore,str_ot] ] ),conscious=conscious)
def printScores(self, scores):
output = ""
for key, value in sorted(((str(k.name),v) for (k,v) in scores.items()),key=lambda lx: -lx[1]):
if not value: break
output += makeminlength("[*"+key+"]:"+str(value),24)
output = output[:-1]
return output
def printScores(self, scores):
output = ""
for key, value in sorted(
((str(k.name), v) for (k, v) in scores.items())):
output += makeminlength("[*" + key + "]:" + str(value), 24)
output = output[:-1]
return output
def store_stats(self,meter,init):
textname=init.getName()
if not textname: textname=str(self).replace(" ","_")
## store stats
if (not textname in init.meter_stats['lines']):
init.meter_stats['lines'][textname]={}
if (not textname in init.meter_stats['positions']):
init.meter_stats['positions'][textname]={}
if (not textname in init.meter_stats['texts']):
init.meter_stats['texts'][textname]={}
if (not textname in init.meter_stats['_ot']):
init.meter_stats['_ot'][textname]=makeminlength("line",being.linelen)+"\tmeter\t"+init.ckeys+"\n"
parsedat=[]
for k,v in sorted(self.__bestparse[meter].constraintScores.items()):
if (not k in init.meter_stats['texts'][textname]):
init.meter_stats['texts'][textname][k]=[]
init.meter_stats['texts'][textname][k].append(v)
#parsedat.append(v/len(self.__bestparse.positions)) #???
parsedat.append(v)
linekey=str(len(init.meter_stats['lines'][textname])+1).zfill(6)+"_"+str(self.__bestparse[meter].posString())
init.meter_stats['lines'][textname][linekey]=parsedat
## OT stats
parses=self.__parses[meter]
init.meter_stats['_ot'][textname]+=makeminlength(str(self),being.linelen)+"\t"+parses[0].str_ot()+"\n"
if len(parses)>1:
for parse in parses[1:]:
init.meter_stats['_ot'][textname]+=makeminlength("",being.linelen)+"\t"+parse.str_ot()+"\n"
for posn in range(len(self.__bestparse[meter].positions)):
pos=self.__bestparse[meter].positions[posn]
(posdat,ckeys)=pos.formatConstraints(normalize=True,getKeys=True)
for cnum in range(len(ckeys)):
if (not posn in init.meter_stats['positions'][textname]):
init.meter_stats['positions'][textname][posn]={}
if (not ckeys[cnum] in init.meter_stats['positions'][textname][posn]):
init.meter_stats['positions'][textname][posn][ckeys[cnum]]=[]
init.meter_stats['positions'][textname][posn][ckeys[cnum]].append(posdat[cnum])
def scansion(self,meter=None,conscious=False):
bp=self.bestParse(meter)
config=being.config
#if not bp: return
lowestScore=''
str_ot=''
count=''
meterstr=''
if bp:
meterstr=bp.str_meter()
str_ot=bp.str_ot()
lowestScore=bp.score()
count=bp.totalCount
from tools import makeminlength
#print makeminlength(unicode(bp),60)
data = [makeminlength(unicode(self),config['linelen']), makeminlength(unicode(bp) if bp else '', config['linelen']),meterstr,len(self.allParses(meter)),count,lowestScore,str_ot]
data_unicode = [unicode(x) for x in data]
self.om("\t".join( data_unicode ),conscious=conscious)
def printScores(self, scores):
output = ""
for key, value in sorted(
((str(k.name), v) for (k, v) in scores.items()),
key=lambda lx: -lx[1]):
if not value: break
output += makeminlength("[*" + key + "]:" + str(value), 24)
output = output[:-1]
return output
def scansion(self,meter=None,conscious=False):
bp=self.bestParse(meter)
lowestScore=bp.score()
from tools import makeminlength
self.om("\t".join( [ str(x) for x in [makeminlength(str(self),being.linelen), makeminlength(str(bp), being.linelen),len(self.allParses(meter)),lowestScore,bp.str_ot()] ] ),conscious=conscious)
def printScores(self, scores):
output = ""
for key, value in sorted(((str(k.name),v) for (k,v) in scores.items())):
output += makeminlength("[*"+key+"]:"+str(value),24)
output = output[:-1]
return output
def grid(self,nspace=10):
GRID_LINES=[]
for LINE in self.lines():
# word line
line_words=[]
for i,wtok in enumerate(LINE):
if wtok.is_punct: continue
line_words+=[makeminlength(wtok.token,nspace)]
line_words = ' '.join(line_words)
# grid lines
import math,numpy as np
grid_lines=[]
max_grid = max([wtok.phrasal_stress for wtok in self.children if wtok.phrasal_stress!=None])
min_grid = min([wtok.phrasal_stress for wtok in self.children if wtok.phrasal_stress!=None])
#for line_num in range(int(math.ceil(min_grid))+1,int(math.ceil(max_grid))+1):
for line_num in range(1,int(math.ceil(max_grid))+1):
grid_line=[]
for i,wtok in enumerate(LINE):
if wtok.is_punct: continue
mark = 'X' if wtok.phrasal_stress!=None and wtok.phrasal_stress<=line_num else ''
grid_line+=[makeminlength(mark,nspace)]
grid_lines+=[' '.join(grid_line)]
# lines data
data_lines=[]
for datak in ['mean','mean_line','phrasal_stress_peak','phrasal_head']:
data_line=[]
for i,wtok in enumerate(LINE):
if wtok.is_punct: continue
v=wtok.feats.get(datak,'')
if v==None:
v=''
elif type(v) in [float,np.float64]:
if np.isnan(v):
v=''
else:
v=round(v,1)
elif v in [True,False]:
v='+' if v else '-'
if datak == 'phrasal_stress_peak':
v2=wtok.feats.get('phrasal_stress_valley')
if v2 and v:
v=v+'/-'
elif v2 and not v:
v='-'
mark = str(v)
data_line+=[makeminlength(mark,nspace)]
datak_name = datak
if datak=='mean': datak_name='stress [sent]'
if datak=='mean_line': datak_name='stress [line]'
if datak=='norm_mean': datak_name='stress [sent norm]'
if datak=='norm_mean_line': datak_name='stress [line norm]'
if datak=='phrasal_stress_peak': datak_name='peak/valley'
if datak=='phrasal_head': datak_name='head/foot'
data_line+=[makeminlength('('+datak_name+')',nspace)]
data_lines+=[' '.join(data_line)]
#grid_lines.insert(0,line_words)
grid_lines.append(line_words)
grid_lines.append('')
grid_lines.extend(data_lines)
maxlength = max([len(l) for l in grid_lines])
hdrftr='#' * maxlength
hdr='STRESS GRID: '+wordtoks2str(LINE)
#grid_lines.insert(0,self.token)
#grid_lines.insert(0,hdr)
#grid_lines.insert(0,hdrftr)
#grid_lines.append(hdrftr)
GRID_LINES+=['\n'.join(grid_lines)]
return '\n\n\n'.join(GRID_LINES)
def grid(self,nspace=10):
GRID_LINES=[]
for LINE in self.lines():
# word line
line_words=[]
for i,wtok in enumerate(LINE):
if wtok.is_punct: continue
line_words+=[makeminlength(wtok.token,nspace)]
line_words = ' '.join(line_words)
# grid lines
import math,numpy as np
grid_lines=[]
max_grid = max([wtok.phrasal_stress for wtok in self.children if wtok.phrasal_stress!=None])
min_grid = min([wtok.phrasal_stress for wtok in self.children if wtok.phrasal_stress!=None])
#for line_num in range(int(math.ceil(min_grid))+1,int(math.ceil(max_grid))+1):
for line_num in range(1,int(math.ceil(max_grid))+1):
grid_line=[]
for i,wtok in enumerate(LINE):
if wtok.is_punct: continue
mark = 'X' if wtok.phrasal_stress!=None and wtok.phrasal_stress<=line_num else ''
grid_line+=[makeminlength(mark,nspace)]
grid_lines+=[' '.join(grid_line)]
# lines data
data_lines=[]
for datak in ['mean','mean_line','phrasal_stress_peak','phrasal_head']:
data_line=[]
for i,wtok in enumerate(LINE):
if wtok.is_punct: continue
v=wtok.feats.get(datak,'')
if v==None:
v=''
elif type(v) in [float,np.float64]:
if np.isnan(v):
v=''
else:
v=round(v,1)
elif v in [True,False]:
v='+' if v else '-'
if datak == 'phrasal_stress_peak':
v2=wtok.feats.get('phrasal_stress_valley')
if v2 and v:
v=v+'/-'
elif v2 and not v:
v='-'
mark = str(v)
data_line+=[makeminlength(mark,nspace)]
datak_name = datak
if datak=='mean': datak_name='stress [sent]'
if datak=='mean_line': datak_name='stress [line]'
if datak=='norm_mean': datak_name='stress [sent norm]'
if datak=='norm_mean_line': datak_name='stress [line norm]'
if datak=='phrasal_stress_peak': datak_name='peak/valley'
if datak=='phrasal_head': datak_name='head/foot'
data_line+=[makeminlength('('+datak_name+')',nspace)]
data_lines+=[' '.join(data_line)]
#grid_lines.insert(0,line_words)
grid_lines.append(line_words)
grid_lines.append('')
grid_lines.extend(data_lines)
maxlength = max([len(l) for l in grid_lines])
hdrftr='#' * maxlength
hdr='STRESS GRID: '+wordtoks2str(LINE)
#grid_lines.insert(0,self.token)
#grid_lines.insert(0,hdr)
#grid_lines.insert(0,hdrftr)
#grid_lines.append(hdrftr)
GRID_LINES+=['\n'.join(grid_lines)]
return '\n\n\n'.join(GRID_LINES)