'a': 'adj',

'ad': 'adv',

'd': 'adv',

'u': 'v aux',

'r': 'pron',

'p': 'prep',

'prep': 'prep',

'c': 'conj',

't': 'n',

'm': 'n'

if pos_zh[0] == 'n':

return pos_en == 'n' or pos_en == 'gerund'

elif pos_zh == 'v':

return len(pos_en) and pos_en[0] == 'v'

u'我': 'I',

u'我们': 'we',

u'她': 'she',

u'她们': 'they',

u'他': 'he',

u'他们': 'they',

u'它们': 'they',

# it/you don't matter

# make sure the subject pronoun is in subject form

# heuristic: if it's the first word or the previous word is punctuation

# or conjunction, it's considered a subject

if word[1] == 'r' and word[0] in subject_pronoun:

if idx == 0 or sentence[idx-1][1] in ['x', 'c']:

return (subject_pronoun[word[0]], 'pron')

# handle special case: <digits>/m 日/m

if word[1] == 'm':

if DIGITS_PATTERN.match(word[0]):

if idx+1 < len(sentence) and sentence[idx+1][0] == u'日':

# return proper date string

return (translate_date(int(word[0])), 'n')

else:

# return digits directly

return (word[0], 'n')

elif word[0] == u'日':

# symmetric case

if idx > 0 and DIGITS_PATTERN.match(sentence[i-1][0]):

return ('', '')

# construct a list of translations with the same pos as word

same_pos_translations = filter(lambda t: match_pos(word[1], t[1]), translations)

ng = NGram()

if len(same_pos_translations) > 0:

max_unigram_trans = max(same_pos_translations, key=lambda t: ng.get(t[0]))

return max_unigram_trans

"""find the clause that the word at idx belongs to"""

start = 0

end = len(sentence)

for i, word in enumerate(sentence):

if word[1] == 'x':

if i > idx:

end = i

break

else:

start = i + 1

clause, clause_start = find_clause(sentence, idx)

word, pos_zh = sentence[idx]

if pos_zh == 'x':

# punctuation

return punctuation.get(word, word)

if pos_zh == 'eng':

return word

# remove '到' after verbs

if idx > 0 and word == u'到' and pos_zh in ['u', 'p'] and sentence[idx - 1][1] == 'v':

return ''

# translate normal word

translations = dictionary[word]

# handle multiple equivalence, e.g. 'express; indicate'

translations = map(

lambda t: (t[0][:t[0].index(';')], t[1]) if ';' in t[0] else t,

translations)

trans, pos_en = select_translation(sentence, idx, (word, pos_zh), translations)

if pos_zh == 'v':

# 被 <v> -> be <v>.pp

# 被 <n> <v> -> be <v>.pp by <n>

if idx > 0 and sentence[idx - 1][0] == u'被' and sentence[idx - 1][1] == 'p' or \

idx > 1 and sentence[idx - 2][0] == u'被' and sentence[idx - 2][1] == 'p' and sentence[idx - 1][1] == 'n':

be = 'is' # default to 'is'

if u'我' in clause[:idx - clause_start]:

be = 'am'

elif u'你' in clause[:idx - clause_start] or u'们' in clause[:idx - clause_start]:

be = 'are'

trans = transform_word(trans, 'pp')

if sentence[idx - 1][1] == 'n':

translated[idx - 2] = ''

# swap <n> and <v>

noun = translated[idx - 1]

translated[idx - 1] = '%s %s by' % (be, trans)

trans = noun

else:

translated[idx - 1] = ''

trans = '%s %s' % (be, trans)

# <v> <ul>|<ug> -> past tense

elif idx < len(sentence) - 1 and sentence[idx + 1][1] in ['ul', 'ug']:

trans = transform_word(trans, 'past')

# <v> <u> <ul> -> past tense

elif idx < len(sentence) - 2 and sentence[idx + 1][1] == 'u' and sentence[idx + 2][1] == 'ul':

trans = transform_word(trans, 'past')

# <tps> <adv>? <v> -> third-person singular form of <v>

elif idx > 0 and sentence[idx - 1][0] in [u'他', u'她', u'它'] or \

idx > 1 and sentence[idx - 2][0] in [u'他', u'她', u'它'] and sentence[idx - 1][1] in ['ad', 'd']:

trans = transform_word(trans, 'tps')

# remove <ul>, <ug> in <v> <u>? <ul>|<ug>

if pos_zh in ['ul', 'ug']:

if idx > 0 and sentence[idx - 1][1] == 'v' or \

idx > 1 and sentence[idx - 2][1] == 'v' and sentence[idx - 1][1] == 'u':

trans = ''

# remove <uz> in <v> <uz>

# (although <uz> sometimes implies the state of doing something)

if pos_zh == 'uz':

if idx > 0 and sentence[idx - 1][1] == 'v':

trans = ''

# remove <uv> in <ad>|<d> <uv>

if pos_zh == 'uv':

if idx > 0 and sentence[idx - 1][1] in ['ad', 'd']:

trans = ''

#10. remove <adj> 的 <n>

if idx > 0 and word == u'的' and pos_zh == 'uj' and sentence[idx - 1][1] == 'a':

return ''

#11. <n1> 的 <n2> -> <n1>’s <n2>

if idx > 0 and word == u'的' and pos_zh == 'uj' and sentence[idx - 1][1][0] == 'n':

return "'s"

#12. 我/r 的/uj -> my

if idx > 0 and word == u'我' and sentence[idx + 1][0] == u'的':

del sentence[idx + 1]

return "my"

# 你/r 的/uj -> my

if idx > 0 and word == u'你' and sentence[idx + 1][0] == u'的':

del sentence[idx + 1]

return "your"

# 他/r 的/uj -> my

if idx > 0 and word == u'他' and sentence[idx + 1][0] == u'的':

del sentence[idx + 1]

return "his"

# 她/r 的/uj -> my

if idx > 0 and word == u'她' and sentence[idx + 1][0] == u'的':

del sentence[idx + 1]

return "her"

#13. 还/d 会/v ->还会

if idx > 0 and word == u'还' and sentence[idx + 1][0] == u'会':

del sentence[idx + 1]

return "also"