def get_words_from_kap(node):
r"""Return a list of all the terms in a <kap>. Every term in a
<kap> is an alternative spelling of the same term. This is not
necessarily single words, since ReVo includes entries such as
'brazila nukso'.
<kap><ofc>*</ofc><tld/>o</kap>
<kap>brazil<tld/>arbo, <var><kap>brazila <tld/>arbo</kap></var></kap>
(from nuks.xml)
The heavy lifting is done in flatten_kap, all we do here is
separate out terms and remove extraneous whitespace.
Possible formats encountered:
'foo'
'foo, bar'
'foo,\n bar'
'(n,p)-matrico' (the only term in ReVo with an internal comma)
"""
flat_string = flatten_node(node, skip_tags=['ofc', 'fnt'])
if flat_string == '(n,p)-matrico':
words = ['(n,p)-matrico']
else:
words = flat_string.split(',')
if len(words) > 1:
for i in range(len(words)):
# remove trailing/leading space and awkard newlines
words[i] = clean_string(words[i])
return words
def get_words_from_kap(node):
r"""Return a list of all the terms in a <kap>. Every term in a
<kap> is an alternative spelling of the same term. This is not
necessarily single words, since ReVo includes entries such as
'brazila nukso'.
<kap><ofc>*</ofc><tld/>o</kap>
<kap>brazil<tld/>arbo, <var><kap>brazila <tld/>arbo</kap></var></kap>
(from nuks.xml)
The heavy lifting is done in flatten_kap, all we do here is
separate out terms and remove extraneous whitespace.
Possible formats encountered:
'foo'
'foo, bar'
'foo,\n bar'
'(n,p)-matrico' (the only term in ReVo with an internal comma)
"""
flat_string = flatten_node(node, skip_tags=["ofc", "fnt"])
if flat_string == "(n,p)-matrico":
words = ["(n,p)-matrico"]
else:
words = flat_string.split(",")
if len(words) > 1:
for i in range(len(words)):
# remove trailing/leading space and awkard newlines
words[i] = clean_string(words[i])
return words
def flatten_node(node, skip_tags=None):
"""Return a friendly string representing the contents of this node
and its children. This method is generic although occasionally we
need methods which are specific to a certain node type.
skip_tags specifies node tags for a node which we don't recurse
into (although we will collect its tail, since that is outside).
Some examples:
<rim>
La tuta terminologio pri <tld/>oj, <tld/>-vektoroj kaj -subspacoj
de endomorfio ekzistas ankaŭ
por <frm>(<k>n</k>,<k>n</k>)</frm>-matrico, konvencie
identigita kun la endomorfio, kies matrico rilate al la kanona bazo
de <frm><g>K</g><sup><k>n</k></sup></frm> ĝi estas.
</rim>
(from ajgen.xml)
<ekz>
<ctl>popolo</ctl>, <ctl>foliaro</ctl>, <ctl>herbo</ctl>,
<ctl>armeo</ctl> estas ar<tld/>oj.
</ekz>
(from vort.xml)
<ekz>
<ind>saluton!</ind>
[...]
</ekz>
(from salut.xml)
<klr>(de <ref cel="polino.0o">polinomo</ref>)</klr>
(from radik.xml)
"""
flatten_method = get_flatten_method(node)
flat_string = flatten_method(node)
for child in node.getchildren():
flat_string += _flatten(child, skip_tags)
return clean_string(flat_string)
def get_examples(node):
"""Get all examples from the children of a node. Examples tend to
be in <dif>s, and take the following form:
<ekz>
simpla, kunmetita, dubsenca <tld/>o;
</ekz><ekz>
uzi la ĝustan, konvenan <tld/>on;
</ekz><ekz>
la bildoj elvokitaj de la <tld/>oj;
</ekz><ekz>
<tld/>ordo.
</ekz>
(from vort.xml)
Sometimes (bizarrely) examples spread across several <ekz> nodes:
<ekz>
<tld/>i al si plezuron<fnt>Z</fnt>;
</ekz><ekz>
<tld/>i instruon<fnt>Z</fnt>,
</ekz><ekz>
amikecon<fnt>Z</fnt>,
[...]
(from sercx.xml)
Sometimes only references, which we discard:
<ekz>
<ref tip="sub" cel="bier.0o">biero</ref>,
<ref tip="sub" cel="brand.0o">brando</ref>,
<ref tip="sub" cel="vin.0o">vino</ref>
</ekz>
(from alkoho.xml)
<subsnc mrk="afekt.0o.sxajnigi" ref="afekt.0i.sxajnigi">
<ekz>
kiom aĉas la <tld/>o komplezi al duonvivul'
(from afekt.xml)
"""
raw_examples = []
# examples tend to be on <dif>s
for dif_node in node.findall('dif'):
for ekz_node in dif_node.findall('ekz'):
raw_example = flatten_example(ekz_node)
if raw_example:
raw_examples.append(raw_example)
# but examples can also be on the <snc>/<subsnc> itself
# (or even a <drv>!)
for ekz_node in node.findall('ekz'):
raw_example = flatten_example(ekz_node)
if raw_example:
raw_examples.append(raw_example)
# fix examples spread over multiple <ekz>s by concatenating each
# example that ends with a comma with the next example
examples = []
example_string = ""
for (example, source) in raw_examples:
example_string += ' ' + example
if not example_string.endswith(','):
examples.append((clean_string(example_string), source))
example_string = ""
if example_string != "":
art_node = ekz_node.iterancestors('art').next()
kap_node = art_node.iter('kap').next()
word = get_words_from_kap(kap_node)[0]
print("Warning: example for %s ended with comma: %s" % \
(word, clean_string(example_string)))
return examples
def get_args(self, message, lower=True):
messy_args = message.clean_content[len(f'okuyasu {self.name} '):]
return clean_string(messy_args, lower=lower)
