def build_cped():
global dict_id, entry_id, term_id, con
entries = cped_data.entries
# Perform corrections and stuff on the entries data.
# -> utf8
for entry in entries:
# Generate phonetic hash, and boost factor. Boost is ignored for CPED.
entry[0] = textfunctions.vel_to_uni(entry[0])
entry[1:1] = [textfunctions.phonhash(entry[0]), 1]
CPEDRow = collections.namedtuple(
'CPEDRow',
'pali phonhash boost defn grammar meaning source inflectgroup inflectinfo baseword basedefn funcstem regular'
)
rows = tuple(CPEDRow._make(entry) for entry in entries)
# Populate dictionary information.
dict_id += 1
author = "A.P.Buddhadatta Mahāthera"
about = ""
details = ""
con.execute('INSERT INTO dicts VALUES(?, ?, ?, ?, ?, ?)',
(dict_id, 'CPED', 'Concise Pali-English Dictionary', author,
about, details))
terms = []
entries = []
search_entries = []
for row in rows:
term_id += 1
entry_id += 1
terms.append(
(term_id, None, row.pali, 1, row.phonhash, row.boost, entry_id))
entries.append(
(entry_id, term_id, None, textfunctions.mangle(row.meaning),
row.meaning, row.grammar, dict_id))
search_entries.append((entry_id, row.meaning))
print(len(terms), len(entries))
# term_id, base_id, term, number, phon_hash, boost, entry_id
con.executemany('INSERT INTO terms_base VALUES(?, ?, ?, ?, ?, ?, ?)',
terms)
# entry_id, term_id, html, text, brief, info, dict_id
con.executemany('INSERT INTO entries_base VALUES (?, ?, ?, ?, ?, ?, ?)',
entries)
def build_cped():
global dict_id, entry_id, term_id, con
entries = cped_data.entries
# Perform corrections and stuff on the entries data.
# -> utf8
for entry in entries:
# Generate phonetic hash, and boost factor. Boost is ignored for CPED.
entry[0] = textfunctions.vel_to_uni(entry[0])
entry[1:1] = [textfunctions.phonhash(entry[0]), 1]
CPEDRow = collections.namedtuple('CPEDRow', 'pali phonhash boost defn grammar meaning source inflectgroup inflectinfo baseword basedefn funcstem regular')
rows = tuple(CPEDRow._make(entry) for entry in entries)
# Populate dictionary information.
dict_id += 1
author = "A.P.Buddhadatta Mahāthera"
about = ""
details = ""
con.execute('INSERT INTO dicts VALUES(?, ?, ?, ?, ?, ?)',
(dict_id, 'CPED', 'Concise Pali-English Dictionary',
author, about, details))
terms = []
entries = []
search_entries = []
for row in rows:
term_id += 1
entry_id += 1
terms.append((term_id, None, row.pali, 1, row.phonhash, row.boost, entry_id))
entries.append((entry_id, term_id, None, textfunctions.mangle(row.meaning), row.meaning, row.grammar, dict_id))
search_entries.append((entry_id, row.meaning))
print(len(terms), len(entries))
# term_id, base_id, term, number, phon_hash, boost, entry_id
con.executemany('INSERT INTO terms_base VALUES(?, ?, ?, ?, ?, ?, ?)', terms)
# entry_id, term_id, html, text, brief, info, dict_id
con.executemany('INSERT INTO entries_base VALUES (?, ?, ?, ?, ?, ?, ?)', entries)
def build_dppn():
global dict_id, entry_id, term_id, con
dict_path = sc.dict_sources_dir / 'sc_dppn.html'
with dict_path.open('r', encoding='utf-8') as f:
dom = lxml.html.fromstring(f.read())
# Perform sanity-correction
items = list(dom.cssselect('meta, person, place, thing'))
for e in items:
if e.getparent() != dom:
dom.append(e)
# Populate dictionary information.
dict_id += 1
metas = dom.cssselect('meta')
author = metas[1].attrib['content']
about = metas[0].attrib['content']
details = metas[2].attrib['content']
con.execute('INSERT INTO dicts VALUES(?, ?, ?, ?, ?, ?)',
(dict_id, 'EBPN', 'Early Buddhism Proper Names', author, about,
details))
# Used to process references in DPPN
refrex = regex.compile(r'(\w+)(?:[.]([ivxml]+|p))?[.](\d+)(?:–(\d+))?')
count = collections.Counter()
def loc(e):
e.attrib['class'] = 'location'
ll = e.text
if ll and ll[0].isdigit():
e.tag = 'a'
ll = ll.replace(' ', '')
e.attrib['href'] = 'http://maps.google.com.au/maps?ll={}'.format(
ll)
e.text = '^ see location '
else:
if ll:
e.tag = 'span'
else:
if e.getnext().tag == "precision":
e.getnext().drop_tree()
e.drop_tree()
tfn = {
'ref': ('a', 'ref'),
'description': None,
'place': ('div', 'place'),
'person': ('div', 'person'),
'thing': ('div', 'thing'),
'location': loc,
'precision': ('span', 'precision'),
'type': ('span', 'type'),
}
for entry in dom.cssselect('person, place, thing'):
entry_id += 1
html = lxml.html.tostring(entry, encoding='utf8').decode()
name_rows = []
for i, e in enumerate(entry.iter('name')):
name = e.text_content().strip()
try:
html_id = e.attrib['id']
except KeyError:
html_id = None
count.update([name])
boost = textfunctions.mc4_boost(len(html), 1000)
if i > 0:
boost = (1 + i + boost) / (2 + i)
term_id += 1
name_rows.append([
term_id, 0, name, count[name],
textfunctions.phonhash(name) if name else None, boost, entry_id
])
e.drop_tree()
alt_names = ", ".join(r[2] for r in name_rows[1:])
for e in entry.iter():
if e.tag in tfn:
value = tfn[e.tag]
if value is None:
e.drop_tag()
elif callable(value):
value(e)
elif len(value) == 2:
e.tag = value[0]
e.attrib['class'] = value[1]
else:
raise NotImplementedError
html = lxml.html.tostring(entry, encoding='utf8').decode()
html = html.replace('</a>', '</a> ').replace('<a',
' <a').replace(' ', ' ')
# Destructively modify the element
refstrs = list(t.text.strip() for t in entry.iter('ref'))
for ref in entry.iter('ref'):
ref.text = " " + ref.text + " "
ref.drop_tag()
paras = [p.text_content() for p in entry.iter('p')]
text = " ".join(paras)
text = regex.sub(" {2,}", " ", text)
if paras:
brief = create_brief(paras[0])
else:
brief = None
# terms:
# term_id, base_id, term, number, phon_hash, boost, entry_id
con.executemany('INSERT INTO terms_base VALUES(?, ?, ?, ?, ?, ?, ?)',
name_rows)
# entries:
# entry_id, term_id, html, text, brief, info, dict_id
con.execute('INSERT INTO entries_base VALUES (?, ?, ?, ?, ?, ?, ?)',
(entry_id, name_rows[0][0], alt_names, html,
textfunctions.mangle(text), entry.tag, dict_id))
# Populate references.
for refstr in refstrs:
m = refrex.match(refstr)
try:
values = [entry_id]
values.extend(m[1:])
con.execute('INSERT INTO refs VALUES (?, ?, ?, ?, ?)', values)
except TypeError:
print("Malformed ref: {}, ignoring.".format(refstr))
except:
print(values)
raise
def build_dppn():
global dict_id, entry_id, term_id, con
dict_path = sc.dict_sources_dir / 'sc_dppn.html'
with dict_path.open('r', encoding='utf-8') as f:
dom = lxml.html.fromstring(f.read())
# Perform sanity-correction
items = list(dom.cssselect('meta, person, place, thing'))
for e in items:
if e.getparent() != dom:
dom.append(e)
# Populate dictionary information.
dict_id += 1
metas = dom.cssselect('meta')
author = metas[1].attrib['content']
about = metas[0].attrib['content']
details = metas[2].attrib['content']
con.execute('INSERT INTO dicts VALUES(?, ?, ?, ?, ?, ?)',
(dict_id, 'EBPN', 'Early Buddhism Proper Names',
author, about, details))
# Used to process references in DPPN
refrex = regex.compile(r'(\w+)(?:[.]([ivxml]+|p))?[.](\d+)(?:–(\d+))?')
count = collections.Counter()
def loc(e):
e.attrib['class'] = 'location'
ll = e.text
if ll and ll[0].isdigit():
e.tag = 'a'
ll = ll.replace(' ', '')
e.attrib['href'] = 'http://maps.google.com.au/maps?ll={}'.format(ll)
e.text = '^ see location '
else:
if ll:
e.tag = 'span'
else:
if e.getnext().tag == "precision":
e.getnext().drop_tree()
e.drop_tree()
tfn = {'ref': ('a', 'ref'),
'description': None,
'place': ('div', 'place'),
'person': ('div', 'person'),
'thing': ('div', 'thing'),
'location': loc,
'precision': ('span', 'precision'),
'type': ('span', 'type'),}
for entry in dom.cssselect('person, place, thing'):
entry_id += 1
html = lxml.html.tostring(entry, encoding='utf8').decode()
name_rows = []
for i, e in enumerate(entry.iter('name')):
name = e.text_content().strip()
try:
html_id = e.attrib['id']
except KeyError:
html_id = None
count.update([name])
boost = textfunctions.mc4_boost(len(html), 1000)
if i > 0:
boost = (1 + i + boost) / (2 + i)
term_id += 1
name_rows.append([
term_id,
0,
name,
count[name],
textfunctions.phonhash(name) if name else None,
boost,
entry_id])
e.drop_tree()
alt_names = ", ".join(r[2] for r in name_rows[1:])
for e in entry.iter():
if e.tag in tfn:
value = tfn[e.tag]
if value is None:
e.drop_tag()
elif callable(value):
value(e)
elif len(value) == 2:
e.tag = value[0]
e.attrib['class'] = value[1]
else:
raise NotImplementedError
html = lxml.html.tostring(entry, encoding='utf8').decode()
html = html.replace('</a>', '</a> ').replace('<a', ' <a').replace(' ', ' ')
# Destructively modify the element
refstrs = list(t.text.strip() for t in entry.iter('ref'))
for ref in entry.iter('ref'):
ref.text = " " + ref.text + " "
ref.drop_tag()
paras = [p.text_content() for p in entry.iter('p')]
text = " ".join(paras)
text = regex.sub(" {2,}", " ", text)
if paras:
brief = create_brief(paras[0])
else:
brief = None
# terms:
# term_id, base_id, term, number, phon_hash, boost, entry_id
con.executemany('INSERT INTO terms_base VALUES(?, ?, ?, ?, ?, ?, ?)',
name_rows)
# entries:
# entry_id, term_id, html, text, brief, info, dict_id
con.execute('INSERT INTO entries_base VALUES (?, ?, ?, ?, ?, ?, ?)',
(entry_id, name_rows[0][0], alt_names, html, textfunctions.mangle(text), entry.tag, dict_id))
# Populate references.
for refstr in refstrs:
m = refrex.match(refstr)
try:
values = [entry_id]
values.extend(m[1:])
con.execute('INSERT INTO refs VALUES (?, ?, ?, ?, ?)', values)
except TypeError:
print("Malformed ref: {}, ignoring.".format(refstr))
except:
print(values)
raise
