def _process_wordclass_set(wordclass_set, id):
varset = VARIANTS_CACHE.find(id=id, wordclass=wordclass_set.wordclass())
if varset:
seen = set([type_unit.sort for type_unit in wordclass_set.types()])
bucket = []
for inflection in INFLECTIONS[wordclass_set.wordclass()]:
for variant_form in varset.variants.get(inflection, []):
if variant_form.sort in seen:
continue
if variant_form.date.end < MINIMUM_DATE:
continue
variant_form.wordclass = inflection
bucket.append(variant_form)
seen.add(variant_form.sort)
for variant_form in bucket:
morphset_node = etree.Element('morphSet', fragment='true')
d1, d2 = variant_form.date.constrain((wordclass_set.date().start,
wordclass_set.date().end))
daterange = DateRange(start=d1, end=d2, hardEnd=True)
morphset_node.append(daterange.to_xml(omitProjected=True))
if variant_form.regional:
morphset_node.set('regional', 'true')
if variant_form.irregular:
morphset_node.set('irregular', 'true')
type_node = etree.SubElement(morphset_node, 'type')
form_node = etree.SubElement(type_node, 'form')
form_node.text = variant_form.form
type_node.append(Wordclass(variant_form.wordclass).to_xml())
wordclass_set.morphset_block().append(morphset_node)
def _variant_date_node(variant_date, block_date):
"""
Return a date-range node for a variant form
"""
# Constrain the variant's dates, so that they don't
# fall outside the limits of the parent entry
start_date, end_date = variant_date.constrain((block_date.start, block_date.projected_end()))
# Create a new date range with the constrained dates
constrained_date_range = DateRange(start=start_date, end=end_date, hardEnd=True)
return constrained_date_range.to_xml(omitProjected=True)
def construct_entry_node(self, entry):
entry_node = etree.Element('e', odoLexid=entry.lexid)
if entry.wordclass_blocks[0].wordclass == 'NP':
entry_node.set('encyclopedic', 'true')
lemma_node = etree.Element('lemma', src=self.dictname)
lemma_node.text = entry.headword
if entry.headword_us:
lemma_node.set('locale', 'uk')
lemma_node2 = etree.Element('lemma', locale='us', src=self.dictname)
lemma_node2.text = entry.headword_us
entry_node.append(lemma_node)
entry_node.append(lemma_node2)
else:
entry_node.append(lemma_node)
for block in entry.wordclass_blocks:
# date node
if entry.date is not None:
daterange = DateRange(start=entry.date,
end=END_DATE,
estimated=False)
if block.wordclass == 'NP':
daterange.is_estimated = True
local_date_node = etree.tostring(daterange.to_xml(omitProjected=True))
elif block.wordclass != 'NP':
local_date_node = DEFAULT_DATE_NODE
else:
local_date_node = None
wordclass_set_node = etree.SubElement(entry_node, 'wordclassSet')
wordclass_set_node.append(Wordclass(block.wordclass).to_xml())
if local_date_node:
wordclass_set_node.append(etree.fromstring(local_date_node))
morphsetblock_node = etree.Element('morphSetBlock')
for morphgroup in block.morphgroups:
morphset_node = etree.SubElement(morphsetblock_node, 'morphSet')
if local_date_node:
morphset_node.append(etree.fromstring(local_date_node))
for unit in morphgroup.morphunits:
type_node = etree.SubElement(morphset_node, 'type')
form_node = etree.SubElement(type_node, 'form')
form_node.text = unit.form
type_node.append(Wordclass(unit.wordclass).to_xml())
wordclass_set_node.append(morphsetblock_node)
wordclass_set_node.append(self.definition_node(block))
wordclass_set_node.append(self.resource_node(block, entry.lexid))
return entry_node
def __init__(self, form, start_date, end_date):
self.original_form = form
self.date = DateRange(start=start_date,
end=end_date,
hardEnd=True)
self.wordclass = None # usually stays undefined
self._lemma_manager = None
self.regional = False # default
self.irregular = False # default
self.has_en_ending = False # default
self.undated = False
self.computed = False
self.structural_id = 0 # default
self.headers = []
self.header_labels = []
self.label = ''
self.grammatical_information = ''
def date(self):
"""
Return a DateRange object for this entry.
"""
# Entries don't have their own date ranges; so we compile this
# from the date ranges of its component wordclasses
try:
return self._date
except AttributeError:
start = min(block.date().start for block in self.wordclass_sets())
end = max(block.date().end for block in self.wordclass_sets())
exact_start = min(block.date().exact('start')
for block in self.wordclass_sets())
exact_end = min(block.date().exact('end')
for block in self.wordclass_sets())
self._date = DateRange(start=start, end=end)
self._date.set_exact('start', exact_start)
self._date.set_exact('end', exact_end)
return self._date
class GelEntry(_GelComponent):
def __init__(self, node, filename):
_GelComponent.__init__(self, node=node, file=filename)
def oed_id(self):
return self.attribute('oedId')
def oed_lexid(self):
return self.attribute('oedLexid')
def tag(self):
return self.attribute('tag')
def wordclass_sets(self):
try:
return self._wordclass_sets
except AttributeError:
self._wordclass_sets = [WordclassSet(n, self.filepath) for n in
self.node.findall('./wordclassSet')]
return self._wordclass_sets
def definition(self, **kwargs):
"""
Return the definition from the first wordclass block.
"""
return self.wordclass_sets()[0].definition(**kwargs)
def primary_wordclass(self):
"""
Return the Penn wordclass from the first wordclass block.
"""
return self.wordclass_sets()[0].wordclass()
def lemma_manager(self):
try:
return self._lemma_manager
except AttributeError:
self._lemma_manager = Lemma(self.node.findtext('./lemma'))
return self._lemma_manager
@property
def lemma(self):
return self.lemma_manager().lemma
@property
def headword(self):
return self.lemma_manager().lemma
@property
def sort(self):
return self.lemma_manager().lexical_sort()
def lemmas(self):
return [Lemma(l.text) for l in self.node.findall('./lemma')]
def us_variant(self):
for l in self.node.findall('./lemma'):
if l.get('locale') == 'us':
return l.text
return None
def frequency(self):
"""
Return the modern frequency for this entry.
"""
# Entries don't have their own frequency tables; so we sum
# the frequencies of its component wordclasses
return sum([block.frequency() for block in self.wordclass_sets()])
def date(self):
"""
Return a DateRange object for this entry.
"""
# Entries don't have their own date ranges; so we compile this
# from the date ranges of its component wordclasses
try:
return self._date
except AttributeError:
start = min(block.date().start for block in self.wordclass_sets())
end = max(block.date().end for block in self.wordclass_sets())
exact_start = min(block.date().exact('start')
for block in self.wordclass_sets())
exact_end = min(block.date().exact('end')
for block in self.wordclass_sets())
self._date = DateRange(start=start, end=end)
self._date.set_exact('start', exact_start)
self._date.set_exact('end', exact_end)
return self._date
def types(self):
types_list = []
for block in self.wordclass_sets():
types_list.extend(block.types())
return types_list
def morphsets(self):
morphset_list = []
for block in self.wordclass_sets():
morphset_list.extend(block.morphsets())
return morphset_list
def oed_entry_type(self):
for block in self.wordclass_sets():
etype = block.oed_entry_type()
if etype is not None:
return etype
return None
class VariantForm(object):
"""
Class representing an OED variant form and its date range.
Arguments to init:
-- form (string)
-- start_date (int)
-- end_date (int)
"""
def __init__(self, form, start_date, end_date):
self.original_form = form
self.date = DateRange(start=start_date,
end=end_date,
hardEnd=True)
self.wordclass = None # usually stays undefined
self._lemma_manager = None
self.regional = False # default
self.irregular = False # default
self.has_en_ending = False # default
self.undated = False
self.computed = False
self.structural_id = 0 # default
self.headers = []
self.header_labels = []
self.label = ''
self.grammatical_information = ''
def __repr__(self):
return '<VariantForm: %s (%d\u2014%d)>' % (self.form,
self.date.start,
self.date.projected_end())
@property
def form(self):
"""
Return the variant form itself (shortcut for self.lemma_manager.lemma).
"""
return self.lemma_manager().lemma
@property
def sort(self):
return self.lemma_manager().lexical_sort()
def lemma_manager(self):
"""
Return a Lemma object based on the variant form.
"""
if self._lemma_manager is None:
form = self.original_form.strip().replace('(', '').replace(')', '')
self._lemma_manager = Lemma(form)
return self._lemma_manager
def reset_form(self, new_lemma):
"""
Replace the Lemma object with a new Lemma object using a new form.
Used by Formslist.detruncate when detruncating the forms list.
"""
self._lemma_manager = Lemma(new_lemma)
return self._lemma_manager
def set_grammatical_information(self, value):
self.grammatical_information = DISTIL_PATTERNS.sub(r'\1', value)
return self.grammatical_information
def is_truncated(self):
"""
Return True if the form is truncated, or False if not.
"""
return self.lemma_manager().is_affix()
def check_en_ending(self, wordclass, headword):
"""
Switch self.has_en_ending to True if this is a verb
ending in -en (and the headword does *not* end in -n).
"""
if (wordclass == 'VB' and
self.date.end <= 1600 and
re.search(r'[eay]n+$', self.form) and
not re.search(r'ne?$', headword)):
self.has_en_ending = True
def merge(self, other):
"""
Merge another VariantForm instance into this one.
Argument is another VariantForm instance.
Used by Formslist.forms_uniq. Since merging will only be done in
cases where the two instances have exactly the same form and exactly
the same grammatical info, we only have to check if the date range
needs to be extended in either direction.
"""
# Switch 'regional' setting to other's regional setting if other
# looks like a more significant form. Ditto 'irregular' setting.
if not self.is_more_important_than(other):
self.regional = other.regional
self.irregular = other.irregular
# If necessary, extend the start date to an earlier start date
self.date.extend_range(other.date)
return
def is_more_important_than(self, other):
if self.date.projected_end() > other.date.projected_end():
return True
if (self.date.projected_end() == other.date.projected_end() and
self.date.span() > other.date.span()):
return True
return False