def DivideCV(self):
"""Divide the word among C and V groups to fill the sylBounds list.
Here, and here alone, we need to catch e-with-grave-accent to count it
as not only a vowel but syllabic ('an aged man' vs. 'aged beef'). Other
special characters might be useful to recognize, but won't make the
same syllabic difference.
"""
unicodeVowels = u"[ae\N{LATIN SMALL LETTER E WITH GRAVE}iouy]+"
uniConsonants = u"[^ae\N{LATIN SMALL LETTER E WITH GRAVE}iouy]+"
firstvowel = sre.search(unicodeVowels, self.wd).start()
for v in sre.finditer(unicodeVowels, self.wd):
lastvowel = v.end() # replaced for each group, last sticks
disyllabicvowels = self.sylvowels.search(v.group())
if disyllabicvowels:
self.sylBounds.append(v.start() + disyllabicvowels.start() + 1)
for cc in sre.finditer(uniConsonants, self.wd):
if cc.start() < firstvowel or cc.end() >= lastvowel: continue
numcons = len(cc.group())
if numcons < 3: pos = cc.end() - 1 # before single C or betw. 2
elif numcons > 3: pos = cc.end() - 2 # before penult C
else: # 3 consonants, divide 1/2 or 2/1?
cg = cc.group() # our CCC cluster
if cg[-3] == cg[-2] or self.splitLeftPairs.search(cg):
pos = cc.end() - 2 # divide 1/2
else: pos = cc.end() - 1 # divide 2/1
if not self.wd[pos-1].isalpha() and not self.wd[pos].isalpha():
self.sylBounds.append(pos-1)
else: self.sylBounds.append(pos)
def DivideCV(self):
"""Divide the word among C and V groups to fill the sylBounds list.
Here, and here alone, we need to catch e-with-grave-accent to count it
as not only a vowel but syllabic ('an aged man' vs. 'aged beef'). Other
special characters might be useful to recognize, but won't make the
same syllabic difference.
"""
unicodeVowels = u"[ae\N{LATIN SMALL LETTER E WITH GRAVE}iouy]+"
uniConsonants = u"[^ae\N{LATIN SMALL LETTER E WITH GRAVE}iouy]+"
firstvowel = sre.search(unicodeVowels, self.wd).start()
for v in sre.finditer(unicodeVowels, self.wd):
lastvowel = v.end() # replaced for each group, last sticks
disyllabicvowels = self.sylvowels.search(v.group())
if disyllabicvowels:
self.sylBounds.append(v.start() + disyllabicvowels.start() + 1)
for cc in sre.finditer(uniConsonants, self.wd):
if cc.start() < firstvowel or cc.end() >= lastvowel: continue
numcons = len(cc.group())
if numcons < 3: pos = cc.end() - 1 # before single C or betw. 2
elif numcons > 3: pos = cc.end() - 2 # before penult C
else: # 3 consonants, divide 1/2 or 2/1?
cg = cc.group() # our CCC cluster
if cg[-3] == cg[-2] or self.splitLeftPairs.search(cg):
pos = cc.end() - 2 # divide 1/2
else: pos = cc.end() - 1 # divide 2/1
if not self.wd[pos-1].isalpha() and not self.wd[pos].isalpha():
self.sylBounds.append(pos-1)
else: self.sylBounds.append(pos)
def TestLengthAndDice(self, logger):
normlen = (self.LD.data['lfeet'] - len(self.LD.data['footlist'])) * 2
if self.LD.data['lastfoot']: normlen -= 2
start = self.LD.data['midremain'][0]
end = self.LD.data['midremain'][1]
currlen = end - start
logger.ExpFootDivision(currlen, normlen)
marks = self.P.GetMarks()
if currlen == normlen:
for (footname, sylinx) \
in footfinder(footDict, marks, 2, start, end):
if footname: self.LD.appendFoot(footname)
else: return self.P.GetScanString(), False
if sylinx < end: # not at end of line or where lastfoot starts
self.P.AddFootDivMark(sylinx
+ self.LD.data['midremain'][0])
elif currlen < normlen:
candidate = marks.find('x//', start, end)
if candidate % 2 != 0: # 'x//' at odd pos hopelessly messy
logger.Explain("\nFAIL! no good position "
+ "for single-stress foot")
return self.P.GetScanString(), False
candidate += 2 # point directly at defective foot
for (footname, sylinx) \
in footfinder(footDict, marks, 2, start, candidate):
if footname: self.LD.appendFoot(footname)
else: return self.P.GetScanString(), False
self.P.AddFootDivMark(sylinx)
self.LD.appendFoot('defective')
self.P.AddFootDivMark(candidate+1)
for (footname, sylinx) \
in footfinder(footDict, marks, 2, candidate+1, end):
if footname: self.LD.appendFoot(footname)
else: return self.P.GetScanString(), False
if sylinx < end: self.P.AddFootDivMark(sylinx)
else: # anapest(s)
need = currlen - normlen
candidates = [p.start()+1 for p in sre.finditer(r'(?=/xx/)',marks)]
if len(candidates) < need:
morecands = [p.start() for p in sre.finditer(r'(?=xx/)',marks)]
candidates += morecands
while start < end:
if need and (start in candidates):
foot = marks[start:start+3]
if footDict.has_key(foot):
self.LD.appendFoot(footDict[foot])
else: return self.P.GetScanString(), False
start += 3
need -= 1
else:
foot = marks[start:start+2]
if footDict.has_key(foot):
self.LD.appendFoot(footDict[foot])
else: return self.P.GetScanString(), False
start += 2
if start < end: self.P.AddFootDivMark(start)
if self.LD.data['lastfoot']:
self.LD.appendFoot(self.LD.data['lastfoot'])
return self.P.GetScanString(), True
def depend(self, filename):
if filename != '-':
fp = open(filename, "r")
body = fp.read()
fp.close()
else:
body = sys.stdin.read()
pass
files = [filename]
for ssi in re.finditer("<!--#(\w+)\s+(.*)\s*-->", body):
element = ssi.expand("\\1")
attrs = self.__parse_attribute(ssi.expand("\\2"))
if element == "include":
l = self.depend(os.path.normpath(self.document_root + attrs["virtual"]))
files.append(l)
elif element == "set":
value = self.__interpolate_variable(attrs["value"])
self.variables.update({attrs["var"]: value})
elif element == "echo":
pass
else:
print >>sys.stderr, "Unsupported SSI element", element, "in", body[ssi.start():ssi.end()]
raise AttributeButton
pass
return (files)
def __interpolate_variables(self, name):
s = ""
_last = 0
for l in re.finditer("\${(\w+)}|\$(\w+)", name):
s += name[_last:l.start()]
v = filter(lambda x: x != None, l.groups())
v = v[0]
s += self.variables[v]
_last = l.end()
pass
s += name[_last:]
return (s)
def parse(self, filename):
if filename != '-':
fp = open(filename, "r")
body = fp.read()
fp.close()
else:
body = sys.stdin.read()
pass
_last = 0
output = ""
#matches = re.findall("<!--#(\w+)\s+(.*)\s*-->", body)
#for ssi in matches:
for ssi in re.finditer("<!--#(\w+)\s+(.*)\s*-->", body):
output += body[_last:ssi.start()]
element = ssi.expand("\\1")
attrs = self.__parse_attribute(ssi.expand("\\2"))
if element == "include":
output += self.parse(os.path.normpath(self.document_root + attrs["virtual"]))
elif element == "set":
value = self.__interpolate_variables(attrs["value"])
self.variables.update({attrs["var"]: value})
elif element == "echo":
output += self.__interpolate_variables(attrs["var"])
else:
print >>sys.stderr, "Unsupported SSI element", element, "in", body[ssi.start():ssi.end()]
raise AttributeError
_last = ssi.end()
pass
output += body[_last:]
return (output)
def DoAlgorithm(self, whichAlgorithm, scansion):
"""Run through whole iambic scansion, either algorithm, silently.
Called by top-level DeduceParameters to figure metron. After, if iambic,
called by ChooseAlgorithm to try each approach. If global-to-module
linelength not set, figure it (not reliably!) and set it for step-by-steps.
Note that when call is from DeduceParameters, lfeetset is *false*.
"""
## whichAlgorithm = 1 # TESTTESTTEST
if not lineDat['lfeetset']:
if len(scansion) // 2 >= 2:
linefeet = len(scansion) // 2 # crude! but can't find reliable improvement
lineDat['lfeet'] = linefeet # was not set before; set for all following
else: return ([], [])
else: linefeet = lineDat['lfeet']
footlist = []
if whichAlgorithm == 1:
normlen = linefeet * 2
currlen = len(scansion)
if (currlen > (normlen + 1)) and (scansion[-4:] in ('x/xx', 'xx/x')):
lastfoot = footDict[scansion[-4:]]
linefeet -= 1 # only in local copy!
scansion = scansion[:-4]
elif currlen >= normlen and scansion[-3:] in ('x/x', '//x'):
lastfoot = footDict[scansion[-3:]]
linefeet -= 1
scansion = scansion[:-3]
else: lastfoot = ''
normlen = linefeet * 2 # unnec. if no special feet, but
currlen = len(scansion)
if currlen <= normlen and scansion[:4] in ('/x/x', '/xxx'):
footlist.append('defective')
linefeet -= 1
scansion = scansion[1:]
# end of special-first-last-feet section
if currlen == normlen: # simple disyllables
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0, len(scansion)):
if footname: footlist.append(footname)
else: return ([], [])
elif currlen < normlen: # defective somewhere
candidate = scansion.find('x//')
if candidate % 2 != 0: return ([], [])
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0, candidate):
if footname: footlist.append(footname)
else: return ([], [])
footlist.append('defective')
for (footname, sylinx) in footfinder(footDict, scansion, 2,
candidate+1, len(scansion)):
if footname: footlist.append(footname)
else: return ([], [])
else: # anapests
need = currlen - normlen
candidates = []
for p in sre.finditer(r'(?=/xx/)', scansion):
candidates.append(p.start() + 1)
if len(candidates) < need:
for p in sre.finditer(r'(?=xx/)', scansion): candidates.append(p.start())
i = 0
while i < currlen:
if i in candidates:
footlist.append('anapest')
i += 3
else:
if footDict.has_key(scansion[i:i+2]):
footlist.append(footDict[scansion[i:i+2]])
else: return ([], [])
i += 2
if lastfoot: footlist.append(lastfoot)
else: # algorithm 2
(startoflongest, longest) = longestMatch(self.possIambRE, scansion)
if startoflongest is None: return ([], [])
if startoflongest % 2 == 0: # divide head into disyllables
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0, startoflongest):
if footname: footlist.append(footname)
else: return ([], [])
elif scansion[:2] == '/x': # headless, I guess
# could use better test from alg 1: sc[:4] in ('/xxx', '/x/x'), IFF 3+ syls!!
footlist.append('defective')
for (footname, sylinx) in footfinder(footDict, scansion, 2, 1, startoflongest):
if footname: footlist.append(footname)
else: return ([], [])
else:
anap = scansion.find('xx/', 0, startoflongest)
if anap == -1: return ([],[])
for (footname, longest) in footfinder(footDict, scansion, 2, 0, anap):
if footname: footlist.append(footname)
else: return ([], [])
footlist.append('anapest')
for (footname, sylinx) in footfinder(footDict, scansion, 2, anap + 3,
startoflongest):
if footname: footlist.append(footname)
else: return ([], [])
for (footname, sylinx) in footfinder(footDict, scansion, 2,
startoflongest, startoflongest + longest):
if footname: footlist.append(footname)
else: return ([], [])
# to divide tail, check e-s endings, divide in pairs (anaps??)
scansion = scansion[startoflongest + longest:]
if len(scansion) > 0:
lastfoot = ''
if scansion[-1] == 'x' and len(scansion) > 2:
if footDict.has_key(scansion[-3:]):
lastfoot = footDict[scansion[-3:]]
scansion = scansion[:-3]
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0, len(scansion)):
if footname: footlist.append(footname)
else: return ([], [])
if lastfoot: footlist.append(lastfoot)
for inx, f in enumerate(footlist):
if f == 'pyrrhic':
if inx == len(footlist) - 1 or footlist[inx+1] != 'spondee':
f = '(iamb)'
test = self.P.FeetAtPunctBounds(footlist)
return (footlist, test)
def TestLengthAndDice(self, logger):
normlen = (self.LD.data['lfeet'] - len(self.LD.data['footlist'])) * 2
if self.LD.data['lastfoot']: normlen -= 2
start = self.LD.data['midremain'][0]
end = self.LD.data['midremain'][1]
currlen = end - start
logger.ExpFootDivision(currlen, normlen)
marks = self.P.GetMarks()
if currlen == normlen:
for (footname, sylinx) \
in footfinder(footDict, marks, 2, start, end):
if footname: self.LD.appendFoot(footname)
else: return self.P.GetScanString(), False
if sylinx < end: # not at end of line or where lastfoot starts
self.P.AddFootDivMark(sylinx +
self.LD.data['midremain'][0])
elif currlen < normlen:
candidate = marks.find('x//', start, end)
if candidate % 2 != 0: # 'x//' at odd pos hopelessly messy
logger.Explain("\nFAIL! no good position " +
"for single-stress foot")
return self.P.GetScanString(), False
candidate += 2 # point directly at defective foot
for (footname, sylinx) \
in footfinder(footDict, marks, 2, start, candidate):
if footname: self.LD.appendFoot(footname)
else: return self.P.GetScanString(), False
self.P.AddFootDivMark(sylinx)
self.LD.appendFoot('defective')
self.P.AddFootDivMark(candidate + 1)
for (footname, sylinx) \
in footfinder(footDict, marks, 2, candidate+1, end):
if footname: self.LD.appendFoot(footname)
else: return self.P.GetScanString(), False
if sylinx < end: self.P.AddFootDivMark(sylinx)
else: # anapest(s)
need = currlen - normlen
candidates = [
p.start() + 1 for p in sre.finditer(r'(?=/xx/)', marks)
]
if len(candidates) < need:
morecands = [
p.start() for p in sre.finditer(r'(?=xx/)', marks)
]
candidates += morecands
while start < end:
if need and (start in candidates):
foot = marks[start:start + 3]
if footDict.has_key(foot):
self.LD.appendFoot(footDict[foot])
else:
return self.P.GetScanString(), False
start += 3
need -= 1
else:
foot = marks[start:start + 2]
if footDict.has_key(foot):
self.LD.appendFoot(footDict[foot])
else:
return self.P.GetScanString(), False
start += 2
if start < end: self.P.AddFootDivMark(start)
if self.LD.data['lastfoot']:
self.LD.appendFoot(self.LD.data['lastfoot'])
return self.P.GetScanString(), True
def DoAlgorithm(self, whichAlgorithm, scansion):
"""Run through whole iambic scansion, either algorithm, silently.
Called by top-level DeduceParameters to figure metron.
After, if iambic, called by ChooseAlgorithm to try each approach.
If global-to-module linelength not set, figure it (not reliably!)
and set it for step-by-steps. Note that when call is from
DeduceParameters, lfeetset is *false*.
"""
## whichAlgorithm = 1 # TESTTESTTEST
if not self.LD.data['lfeetset']:
if len(scansion) // 2 >= 2:
# crude! but can't find reliable improvement
linefeet = len(scansion) // 2
# was not set before; set for all following
self.LD.data['lfeet'] = linefeet
else:
return ([], [])
else:
linefeet = self.LD.data['lfeet']
footlist = []
if whichAlgorithm == 1:
normlen = linefeet * 2
currlen = len(scansion)
if ((currlen > (normlen + 1))
and (scansion[-4:] in ('x/xx', 'xx/x'))):
lastfoot = footDict[scansion[-4:]]
linefeet -= 1 # only in local copy!
scansion = scansion[:-4]
elif currlen >= normlen and scansion[-3:] in ('x/x', '//x'):
lastfoot = footDict[scansion[-3:]]
linefeet -= 1
scansion = scansion[:-3]
else:
lastfoot = ''
normlen = linefeet * 2 # unnec. if no special feet, but
currlen = len(scansion)
if currlen <= normlen and scansion[:4] in ('/x/x', '/xxx'):
footlist.append('defective')
linefeet -= 1
scansion = scansion[1:]
# end of special-first-last-feet section
if currlen == normlen: # simple disyllables
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0,
len(scansion)):
if footname: footlist.append(footname)
else: return ([], [])
elif currlen < normlen: # defective somewhere
candidate = scansion.find('x//')
if candidate % 2 != 0: return ([], [])
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0,
candidate):
if footname: footlist.append(footname)
else: return ([], [])
footlist.append('defective')
for (footname, sylinx) in footfinder(footDict, scansion, 2,
candidate + 1,
len(scansion)):
if footname: footlist.append(footname)
else: return ([], [])
else: # anapests
need = currlen - normlen
candidates = []
for p in sre.finditer(r'(?=/xx/)', scansion):
candidates.append(p.start() + 1)
if len(candidates) < need:
for p in sre.finditer(r'(?=xx/)', scansion):
candidates.append(p.start())
i = 0
while i < currlen:
if i in candidates:
footlist.append('anapest')
i += 3
else:
if footDict.has_key(scansion[i:i + 2]):
footlist.append(footDict[scansion[i:i + 2]])
else:
return ([], [])
i += 2
if lastfoot: footlist.append(lastfoot)
else: # algorithm 2
(startoflongest, longest) = longestMatch(self.possIambRE, scansion)
if startoflongest is None: return ([], [])
if startoflongest % 2 == 0: # divide head into disyllables
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0,
startoflongest):
if footname: footlist.append(footname)
else: return ([], [])
elif scansion[:2] == '/x': # headless, I guess
# could use better test from alg 1: sc[:4] in ('/xxx', '/x/x'), IFF 3+ syls!!
footlist.append('defective')
for (footname, sylinx) in footfinder(footDict, scansion, 2, 1,
startoflongest):
if footname: footlist.append(footname)
else: return ([], [])
else:
anap = scansion.find('xx/', 0, startoflongest)
if anap == -1: return ([], [])
for (footname, longest) in footfinder(footDict, scansion, 2, 0,
anap):
if footname: footlist.append(footname)
else: return ([], [])
footlist.append('anapest')
for (footname, sylinx) in footfinder(footDict, scansion, 2,
anap + 3, startoflongest):
if footname: footlist.append(footname)
else: return ([], [])
for (footname, sylinx) in footfinder(footDict, scansion, 2,
startoflongest,
startoflongest + longest):
if footname: footlist.append(footname)
else: return ([], [])
# to divide tail, check e-s endings, divide in pairs (anaps??)
scansion = scansion[startoflongest + longest:]
if len(scansion) > 0:
lastfoot = ''
if scansion[-1] == 'x' and len(scansion) > 2:
if footDict.has_key(scansion[-3:]):
lastfoot = footDict[scansion[-3:]]
scansion = scansion[:-3]
for (footname, sylinx) in footfinder(footDict, scansion, 2, 0,
len(scansion)):
if footname: footlist.append(footname)
else: return ([], [])
if lastfoot: footlist.append(lastfoot)
for inx, f in enumerate(footlist):
if f == 'pyrrhic':
if inx == len(footlist) - 1 or footlist[inx + 1] != 'spondee':
f = '(iamb)'
test = self.P.FeetAtPunctBounds(footlist)
return (footlist, test)