/usr/share/pyshared/frog/csiparse2.py is in frog 0.12.17-7.1+b1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 | #!/usr/bin/env python
"""
Generate instances for the pairwise dependency prediction task.
usage: %prog [options] [file...]
-mDIST, --max-dist=DIST: maximum distance between head and dependent
---
:Refinements:
-m: type='int', dest='maxDist'
"""
import fileinput
from itertools import imap, izip
from sentences import sentenceIterator
import common
import deptree
def list_rindex(lst, x):
for i in xrange(len(lst) - 1, -1, -1):
if lst[i] == x:
return i
raise ValueError("")
def parseDist(string):
dist = dict((label.strip(), float(weight))
for label, weight in map(str.split, string.strip().split(",")))
normFactor = sum(dist.itervalues())
for cls, weight in dist.iteritems():
dist[cls] = weight / normFactor
return dist
def parseInstanceLine(line):
instance = line.split()
startIndex = list_rindex(instance, "{")
endIndex = instance.index("}", startIndex)
dist = parseDist(" ".join(instance[startIndex:endIndex]).strip("{}"))
return instance[:startIndex], dist, 0 #float(instance[endIndex+1])
def instanceIterator(stream):
return imap(parseInstanceLine, stream)
def formulateWCSP(sentence, dirInstances,
relInstances, pairInstances,
options):
domains = [None] + [[] for token in sentence]
constraints = []
for dependent in sentence:
dependentId = int(dependent[0])
headId = 0
instance, distribution, distance = pairInstances.next()
cls = instance[-1]
conf = distribution[cls]
if cls != "__":
constraints.append(deptree.HasDependency(dependentId,
headId, cls,
conf))
if cls != "__":
domains[dependentId].append((headId, cls))
for dependent, head in common.pairIterator(sentence, options):
dependentId = int(dependent[0])
headId = int(head[0])
instance, distribution, distance = pairInstances.next()
cls = instance[-1]
conf = distribution[cls]
#conf = sum(distribution[x] for x in distribution if x != "__")
#if distance == 0:
# conf = 1000000
#else:
# conf = 1.0 / distance
#constraints.append(deptree.HasDependency(dependentId,
# headId, cls,
# conf))
if cls != "__":
constraints.append(deptree.HasDependency(dependentId,
headId, cls,
conf))
if cls != "__":
domains[dependentId].append((headId, cls))
if dirInstances:
for token, (instance, distribution, distance) in izip(sentence,
dirInstances):
tokenId = int(token[0])
cls = instance[-1]
if True: #cls != "__":
#constraints.append(deptree.DependencyDirection(
# tokenId,
# getattr(deptree.DependencyDirection, cls),
# distribution[cls]))
for cls in distribution:
constraints.append(deptree.DependencyDirection(
tokenId,
getattr(deptree.DependencyDirection, cls),
distribution[cls]))
#if "__" in distribution:
# constraints.append(deptree.DependencyDirection(
# tokenId, 100, distribution["__"]))
#if "RIGHT" in distribution:
# cls = "RIGHT"
# constraints.append(deptree.DependencyDirection(
# tokenId,
# getattr(deptree.DependencyDirection, cls),
# distribution[cls]))
#if "LEFT" in distribution:
# cls = "LEFT"
# constraints.append(deptree.DependencyDirection(
# tokenId,
# getattr(deptree.DependencyDirection, cls),
# distribution[cls]))
#else:
# constraints.append(deptree.DependencyDirection(
# tokenId, deptree.DependencyDirection.LEFT,
# -distribution[cls]))
# constraints.append(deptree.DependencyDirection(
# tokenId, deptree.DependencyDirection.RIGHT,
# -distribution[cls]))
if relInstances:
for token, (instance, distribution, distance) in izip(sentence,
relInstances):
tokenId = int(token[0])
cls = instance[-1]
distribution = dict((tuple(key.split("|")), value)
for key, value
in distribution.iteritems())
if cls != "__":
for rel in cls.split("|"):
conf = sum(value
for key, value
in distribution.iteritems()
if rel in key)
constraints.append(deptree.HasIncomingRel(tokenId,
rel,
conf))
return domains, constraints
def main(options, args):
dirOutput, relsOutput, pairsOutput = args[:3]
dirStream = open(dirOutput)
relsStream = open(relsOutput)
pairsStream = open(pairsOutput)
dirIterator = instanceIterator(dirStream)
relsIterator = instanceIterator(relsStream)
pairsIterator = instanceIterator(pairsStream)
for sentence in sentenceIterator(fileinput.input(args[3:])):
csp = formulateWCSP(sentence,
dirIterator,
relsIterator,
pairsIterator,
options)
dirStream.close()
relsStream.close()
pairsStream.close()
if __name__ == "__main__":
import cmdline
main(*cmdline.parse())
|