def expand(word, flags, flush_stdout):
flag_set = flags.split(" ", 1)
main_flag = flag_set[0]
if len(flag_set) > 1:
extra = flag_set[1]
else:
extra = ""
modifiers = get_modifiers(extra, flags, word)
if main_flag[0] == "/":
inflection_flag = main_flag[1:]
sfx_lines = expand_suffixes(word, inflection_flag, modifiers, extra)
sfx_lines = adjust_affix_tags(sfx_lines, main_flag, flags, modifiers)
else:
sfx_lines = [word + " " + word + " " + flags]
sfx_lines = [ line.replace("v_zna", "v_zn1") if "adj:m:v_rod/v_zna" in line else line for line in sfx_lines ]
sfx_lines = affix.expand_alts(sfx_lines, "//") # TODO: change this to some single-char splitter?
sfx_lines = affix.expand_alts(sfx_lines, "/")
if "/adj" in flags:
out_lines = []
for line in sfx_lines:
if "v_zn1" in line:
if "^noun" in flags or "<" in flags:
line = line.replace("v_zn1", "v_zna")
else:
line = line.replace("v_zn1", "v_zna:ranim")
elif "v_zn2" in line:
if "^noun" in flags or "<" in flags:
line = line.replace("v_zn2", "v_zna")
else:
line = line.replace("v_zn2", "v_zna:rinanim")
out_lines.append(line)
sfx_lines = out_lines
if main_flag[0] != "/":
sfx_lines = util.expand_nv(sfx_lines)
sfx_lines = modify(sfx_lines, modifiers)
if "\\" in flags:
for i in range(0, len(sfx_lines)):
sfx_lines[i] = sfx_lines[i] + ":compb"
words = post_expand(sfx_lines, flags)
return words
def preprocess(line):
if line.count(" /") > 1:
parts = line.split(" ")
line1 = parts[:2] + parts[3:]
line2 = parts[:1] + parts[2:]
lines = [" ".join(line1), " ".join(line2)]
else:
lines = affix.expand_alts([line], "|")
out_lines = []
for line in lines:
out_lines.extend(preprocess2(line))
return out_lines
def expand_line(line, flush_stdout):
global main_word
global main_flag
global last_adv
lines = preprocess(line)
out_lines = []
for line in lines:
sub_lines = []
# +cs
if "\\ +" in line:
line, *sub_lines = line.split("\\")
line = line.rstrip()
if " :" in line or not " /" in line:
line += ":compb"
else:
line += " :compb"
# print(" \\+", line, file=sys.stderr)
# main_word = line
# sublines = expand_subposition(main_word, line)
# out_lines.extend( sublines )
# word lemma tags
elif word_lemma_re.search(line):
if "/" in line:
exp_lines = affix.expand_alts([line], "//") # TODO: change this to some single-char splitter?
try:
exp_lines = affix.expand_alts(exp_lines, "/")
except:
print("Failed to expand", exp_lines, file=sys.stderr)
raise
else:
exp_lines = [ line ]
if ":nv" in line and not "v_" in line:
exp_lines = util.expand_nv(exp_lines)
out_lines.extend( exp_lines )
continue
# word tags
# word /flags [mods] [tags]
try:
word, flags = line.split(" ", 1)
except:
print("Failed to find flags in", line, file=sys.stderr)
raise
main_word = word
inflected_lines = expand(word, flags, flush_stdout)
if sub_lines:
idx = 0
for sub_line in sub_lines:
if flags.startswith("adv:"):
extra_flags = flags[3:].replace(":compb", "")
# util.dbg("sub_lines: %s, %s", flags, extra_flags)
elif " :" in flags or flags.startswith(":"):
extra_flags = re_search("(^| )(:[^ ]+)", flags).group(2).replace(":compb", "")
# util.dbg("===", extra_flags)
else:
extra_flags = ""
if " adv" in line:
sublines = expand_subposition_adv_main(main_word, sub_line, extra_flags)
else:
sublines = expand_subposition(main_word, sub_line, extra_flags, idx)
out_lines.extend( sublines )
if ".adv" in line and "/adj" in line:
for inflected_line in inflected_lines:
if " adv" in inflected_line:
last_adv = inflected_line.split()[0]
cs_lines = expand_subposition_adv(last_adv, sub_line, extra_flags)
out_lines.extend(cs_lines)
break
# print(".adv", last_adv, file=sys.stderr)
idx += 1
out_lines.extend( inflected_lines )
for l in inflected_lines:
if not l.strip():
raise Exception("empty liner", inflected_lines)
return post_process(out_lines)