def _get_jel_text(entr):
'''Generate and return a JEL string from entry object
'entr'. The first line (text before the first "\n"
character) is removed since it contains nformation
that will vary between objects read from a database
and created by parsing input text.'''
jeltxt = fmtjel.entr(entr)
return jeltxt.partition('\n')[2]
def main(args, opts):
# The following call creates a database "cursor" that will
# be used for subsequent database operations. It also, as
# a side-effect, create a global variable in module 'jdb'
# named 'KW' which contains data read from all the keyword
# database tables (tables with names matchingthe pattern
# "kw*". We read this data once at program startup to avoid
# multiple hi-cost trips to the database later.
try:
cur = jdb.dbOpen(opts.database, **jdb.dbopts(opts))
except jdb.dbapi.OperationalError as e:
print("Error, unable to connect to database, do you need -u or -p?\n",
str(e),
file=sys.stderr)
sys.exit(1)
enc = opts.encoding or sys.stdout.encoding or 'utf-8'
jdb.reset_encoding(sys.stdout, encoding=enc)
# Get the command line options and convert them into a sql
# statement that will find the desired entries.
sql, sqlargs = opts2sql(args, opts)
if opts.debug:
print(("%s %s" % (sql, repr(sqlargs))))
# Retrieve the entries from the database. 'entrs' will be
# set to a list on entry objects. 'raw' is set to dictionary,
# keyed by table name, and with values consisting of all the
# rows retrieved from that table.
entrs, raw = jdb.entrList(cur, sql, sqlargs, ret_tuple=True)
# Any xrefs in the retrieved entry objects contain contain only
# the entry id numbers of the referenced entries. We want to be
# able to show the refernced entriy's kanji, glosses, etc so we
# call "augment_xrefs" to get this extra information. Same for
# any reverse refrerences.
jdb.augment_xrefs(cur, raw['xref'])
jdb.augment_xrefs(cur, raw['xrer'], rev=1)
jdb.add_xsens_lists(raw['xref'])
jdb.mark_seq_xrefs(cur, raw['xref'])
# Now all we have to do is print the entries.
first = True
for e in entrs:
# Format the entry for printing, according to the
# kind of out put the user requested.
if opts.jel: txt = fmtjel.entr(e)
else: txt = fmt.entr(e)
# Print the formatted entry using the requested encoding
# and inserting a blank line between entries.
if not first: print()
print(txt)
first = False
if len(entrs) == 0: print("No entries found")
def roundtrip(cur, intxt):
jellex.lexreset(Lexer, intxt)
entr = Parser.parse(intxt, lexer=Lexer)
entr.src = 1
jelparse.resolv_xrefs(cur, entr)
for s in entr._sens:
jdb.augment_xrefs(cur, getattr(s, '_xref', []))
for s in entr._sens:
jdb.add_xsens_lists(getattr(s, '_xref', []))
for s in entr._sens:
jdb.mark_seq_xrefs(cur, getattr(s, '_xref', []))
outtxt = fmtjel.entr(entr, nohdr=True)
return outtxt
def _get_text_from_database(seq, src):
cur = jdb.dbOpen('jmdict')
KW = jdb.KW
sql = "SELECT id FROM entr WHERE seq=%s AND src=%s"
elist = jdb.entrList(cur, sql, [seq, src])
if not elist:
print("Entry %s not found" % seq)
return
entr = elist[0]
for s in entr._sens:
jdb.augment_xrefs(cur, getattr(s, '_xref', []))
txt = fmtjel.entr(entr)
txt = txt.partition('\n')[2]
return txt
def dotest(_,
testid,
xmlfn=None,
jelfn=None,
dir='data/fmtjel',
enc='utf_8_sig'):
if xmlfn is None: xmlfn = os.path.join(dir, testid + '.xml')
if jelfn is None: jelfn = os.path.join(dir, testid + '.jel')
expected = readfile(jelfn, enc)
xmlu = readfile(xmlfn, enc)
xml8 = xmlu #xmlu.encode ('utf-8')
elist = Jmparser.parse_entry(xml8)
got = fmtjel.entr(elist[0], nohdr=True)
msg = "\nExpected:\n%s\nGot:\n%s" % (expected, got)
_.assertEqual(expected, got, msg)
def roundtrip(cur, intxt):
# Since hg-180523-6b1a12 we use '\f' to separate the kanji, reading
# and senses sections in JEL text used as input to jelparse()
# rather than '\n' which was previously used. To avoid changing
# all the test data that still uses '\n', we call secsepfix() to
# replace the first two '\n's in the test data with '\f's to make
# suitable for parsing.
intxt = secsepfix(intxt)
jellex.lexreset(Lexer, intxt)
entr = Parser.parse(intxt, lexer=Lexer)
entr.src = 1
jelparse.resolv_xrefs(cur, entr)
for s in entr._sens:
jdb.augment_xrefs(cur, getattr(s, '_xref', []))
for s in entr._sens:
jdb.add_xsens_lists(getattr(s, '_xref', []))
for s in entr._sens:
jdb.mark_seq_xrefs(cur, getattr(s, '_xref', []))
outtxt = fmtjel.entr(entr, nohdr=True)
return outtxt
def _interactive(cur, lexer, parser):
cnt = 0
instr = ''
while 1:
instr = _getinptext()
if not instr: break
jellex.lexreset(lexer, instr)
try:
result = parser.parse(instr, lexer=lexer, debug=opts.debug)
except jelparse.ParseError as e:
if not e.loc: msg = e.args[0]
else: msg = "%s\n%s" % (e.args[0], e.loc)
print(msg)
continue
try:
jelparse.resolv_xrefs(cur, result)
except ValueError:
print(e)
s = fmtjel.entr(result)
print(s)
def test_002(_):
e1 = Entr(id=100, src=1, seq=1000010, stat=2, unap=False)
e1._kanj = [Kanj(txt='手紙', ), Kanj(txt='切手')]
e1._rdng = [Rdng(txt='てがみ'), Rdng(txt='あとで'), Rdng(txt='きって')]
r = Restr()
e1._rdng[0]._restr.append(r)
e1._kanj[1]._restr.append(r)
r = Restr()
e1._rdng[1]._restr.append(r)
e1._kanj[0]._restr.append(r)
r = Restr()
e1._rdng[2]._restr.append(r)
e1._kanj[0]._restr.append(r)
r = Restr()
e1._rdng[2]._restr.append(r)
e1._kanj[1]._restr.append(r)
expect = 'jmdict 1000010 A {100}\n' \
'手紙；切手\n' \
'てがみ[手紙]；あとで[切手]；きって[nokanji]\n'
jeltxt = fmtjel.entr(e1)
msg = "\nA:\n%s\nB:\n%s" % (expect, jeltxt)
_.assertEqual(expect, jeltxt, msg)
def check(self, seq):
global Cur, KW
# Read expected text, remove any unicode BOM or trailing whitespace
# that may have been added when editing.
expected = open("data/fmtjel/" + str(seq) + ".txt",
encoding='utf-8').read().rstrip()
if expected[0] == '\ufeff': expected = expected[1:]
# Read the entry from the database. Be sure to get from the right
# corpus and get only the currently active entry. Assert that we
# received excatly one entry.
sql = "SELECT id FROM entr WHERE src=1 AND seq=%s AND stat=2 AND NOT unap"
entrs, data = jdb.entrList(Cur, sql, (seq, ), ret_tuple=True)
self.assertEqual(1, len(entrs))
# Add the annotations needed for dislaying xrefs in condensed form.
jdb.augment_xrefs(Cur, data['xref'])
jdb.augment_xrefs(Cur, data['xrer'], rev=True)
fmtjel.markup_xrefs(Cur, data['xref'])
# Test fmtjel by having it convert the entry to JEL.
resulttxt = fmtjel.entr(entrs[0]).splitlines(True)
# Confirm that the received text matched the expected text.
if resulttxt: resulttxt = ''.join(resulttxt[1:])
self.assert_(10 < len(resulttxt))
msg = "\nExpected:\n%s\nGot:\n%s" % (expected, resulttxt)
self.assertEqual(expected, resulttxt, msg)
def main (args, opts):
jdb.reset_encoding (sys.stdout, 'utf-8')
errs = []
try: form, svc, dbg, cur, sid, sess, parms, cfg = jmcgi.parseform()
except Exception as e: jmcgi.err_page ([str (e)])
entries = jmcgi.get_entrs (cur, form.getlist ('e'),
form.getlist ('q'), errs)
if errs: jmcgi.err_page (errs)
# Add a .SEQKR attribute to each entry in 'entries' that
# gives the kanji and reading of the newest (most recently
# edited) entry that has the same sequence number.
seqkr_decorate (entries)
# Sort the entries. The sorting order will group entries
# with the same sequence number (.src,.seq) together and
# each of those groups will be ordered by the kanji/reading
# of the newest (most recently edited) entry in the group.
# (The kanji and/or readings of an entry are sometimes changed
# and this order will keep the changed entries together with
# their pre-changed versions, while maintaining an overall
# ordering by kanji/reading.) Within each group having the
# same sequence number, entries are sorted in descending order
# by the timestamp of the most recent history; that is, from
# the most recently edited entry to the least recently edited
# one.
entries.sort (key=lambda e: (
e.SEQKR[0], e.SEQKR[1],
e.src, e.seq, # In case different seqs have same SEQKR.
# e._hist[*].dt is a datatime.datetime instance.
-(e._hist[-1].dt.timestamp() if e._hist else 0),
-e.id))
for e in entries:
for s in e._sens:
if hasattr (s, '_xref'): jdb.augment_xrefs (cur, s._xref)
if hasattr (s, '_xrer'): jdb.augment_xrefs (cur, s._xrer, 1)
if hasattr (e, '_snd'): jdb.augment_snds (cur, e._snd)
cur.close()
disp = form.getfirst ('disp')
if disp == 'xml':
etxts = [fmtxml.entr (e) for e in entries]
elif disp == 'jm':
etxts = [fmtxml.entr (e, compat='jmdict') for e in entries]
elif disp == 'jmne':
etxts = [fmtxml.entr (e, compat='jmnedict') for e in entries]
elif disp == 'jel':
etxts = [fmtjel.entr (e) for e in entries]
elif disp == 'ed':
etxts = [xslfmt.entr (e) for e in entries]
else:
etxts = ['' for e in entries]
jmcgi.htmlprep (entries)
jmcgi.add_encodings (entries) # For kanjidic entries.
if disp == 'ed': etxts = [jmcgi.txt2html (x) for x in etxts]
jmcgi.add_filtered_xrefs (entries, rem_unap=True)
if errs: jmcgi.err_page (errs)
jmcgi.jinja_page ('entr.jinja',
entries=list(zip(entries, etxts)), disp=disp,
svc=svc, dbg=dbg, sid=sid, session=sess, cfg=cfg,
parms=parms, this_page='entr.py')
def test_001(_):
e1 = Entr(id=100, src=1, seq=1000010, stat=2, unap=False)
expect = 'jmdict 1000010 A {100}\n\n\n'
jeltxt = fmtjel.entr(e1)
_.assertEqual(expect, jeltxt)
