spaCy/spacy/syntax/arc_eager.pyx

704 lines
23 KiB
Cython
Raw Normal View History

# cython: profile=True, cdivision=True, infer_types=True
2017-10-27 20:45:57 +03:00
from cpython.ref cimport Py_INCREF
2015-06-09 02:41:09 +03:00
from cymem.cymem cimport Pool
from libc.stdint cimport int32_t
from collections import defaultdict, Counter
import json
2017-04-15 14:05:15 +03:00
from ..typedefs cimport hash_t, attr_t
from ..strings cimport hash_string
from ..structs cimport TokenC
from ..tokens.doc cimport Doc, set_children_from_heads
from .stateclass cimport StateClass
2017-10-27 20:45:57 +03:00
from ._state cimport StateC
2017-04-15 14:05:15 +03:00
from .transition_system cimport move_cost_func_t, label_cost_func_t
from ..gold.example cimport Example
from ..errors import Errors
from .nonproj import is_nonproj_tree
from . import nonproj
2015-06-09 02:41:09 +03:00
# Calculate cost as gold/not gold. We don't use scalar value anyway.
cdef int BINARY_COSTS = 1
cdef weight_t MIN_SCORE = -90000
cdef attr_t SUBTOK_LABEL = hash_string(u'subtok')
DEF NON_MONOTONIC = True
DEF USE_BREAK = True
2014-12-16 14:44:43 +03:00
# Break transition from here
# http://www.aclweb.org/anthology/P13-1074
2014-12-16 14:44:43 +03:00
cdef enum:
SHIFT
REDUCE
LEFT
RIGHT
2015-05-11 17:12:03 +03:00
BREAK
2015-05-11 17:12:03 +03:00
2014-12-16 14:44:43 +03:00
N_MOVES
2015-05-11 17:12:03 +03:00
MOVE_NAMES = [None] * N_MOVES
MOVE_NAMES[SHIFT] = 'S'
MOVE_NAMES[REDUCE] = 'D'
MOVE_NAMES[LEFT] = 'L'
MOVE_NAMES[RIGHT] = 'R'
MOVE_NAMES[BREAK] = 'B'
2015-01-27 19:18:43 +03:00
2020-06-14 18:22:14 +03:00
cdef enum:
HEAD_IN_STACK = 0
2020-06-14 18:22:14 +03:00
HEAD_IN_BUFFER
HEAD_UNKNOWN
cdef struct GoldParseStateC:
char* state_bits
attr_t* labels
int32_t* heads
int32_t* n_kids_in_buffer
int32_t* n_kids_in_stack
2020-06-14 18:22:14 +03:00
int32_t length
int32_t stride
cdef GoldParseStateC create_gold_state(Pool mem, StateClass stcls, Example example) except *:
cdef GoldParseStateC gs
gs.length = example.x.length
gs.stride = 1
gs.state_bits = <char*>mem.alloc(gs.length, sizeof(gs.state_bits[0]))
gs.labels = <attr_t*>mem.alloc(gs.length, sizeof(gs.labels[0]))
gs.heads = <int32_t*>mem.alloc(gs.length, sizeof(gs.heads[0]))
gs.n_kids_in_buffer = <int32_t*>mem.alloc(gs.length, sizeof(gs.n_kids_in_buffer[0]))
gs.n_kids_in_stack = <int32_t*>mem.alloc(gs.length, sizeof(gs.n_kids_in_stack[0]))
cand_to_gold = example.alignment.cand_to_gold
2020-06-20 22:49:46 +03:00
gold_to_cand = example.alignment.cand_to_gold
cdef TokenC ref_tok
for cand_i in range(example.x.length):
gold_i = cand_to_gold[cand_i]
2020-06-20 22:49:46 +03:00
if gold_i is not None: # Alignment found
ref_tok = example.y.c[gold_i]
2020-06-21 02:01:09 +03:00
gold_head = gold_to_cand[gold_i + ref_tok.head]
2020-06-20 22:49:46 +03:00
if gold_head is not None:
gs.heads[cand_i] = gold_head
gs.labels[cand_i] = ref_tok.dep
gs.state_bits[cand_i] = set_state_flag(
gs.state_bits[cand_i],
HEAD_UNKNOWN,
0
)
else:
gs.state_bits[cand_i] = set_state_flag(
gs.state_bits[cand_i],
HEAD_UNKNOWN,
1
)
else:
gs.state_bits[cand_i] = set_state_flag(
gs.state_bits[cand_i],
HEAD_UNKNOWN,
1
)
stack_words = set()
for i in range(stcls.stack_depth()):
s_i = stcls.S(i)
2020-06-21 02:01:09 +03:00
head = gs.heads[s_i]
gs.n_kids_in_stack[head] += 1
stack_words.add(s_i)
buffer_words = set()
for i in range(stcls.buffer_length()):
b_i = stcls.B(i)
2020-06-21 02:01:09 +03:00
head = gs.heads[b_i]
gs.n_kids_in_buffer[head] += 1
buffer_words.add(b_i)
for i in range(gs.length):
2020-06-21 02:01:09 +03:00
head = gs.heads[i]
if head in stack_words:
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_STACK,
1
)
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_BUFFER,
0
)
elif head in buffer_words:
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_STACK,
0
)
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_BUFFER,
1
)
return gs
2020-06-14 18:22:14 +03:00
2020-06-19 01:11:59 +03:00
2020-06-21 02:01:09 +03:00
cdef void update_gold_state(GoldParseStateC* gs, StateClass stcls) except *:
for i in range(gs.length):
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_BUFFER,
0
)
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_STACK,
0
)
gs.n_kids_in_stack[i] = 0
gs.n_kids_in_buffer[i] = 0
stack_words = set()
for i in range(stcls.stack_depth()):
s_i = stcls.S(i)
head = gs.heads[s_i]
gs.n_kids_in_stack[head] += 1
stack_words.add(s_i)
buffer_words = set()
for i in range(stcls.buffer_length()):
b_i = stcls.B(i)
head = gs.heads[b_i]
gs.n_kids_in_buffer[head] += 1
buffer_words.add(b_i)
for i in range(gs.length):
head = gs.heads[i]
if head in stack_words:
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_STACK,
1
)
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_BUFFER,
0
)
elif head in buffer_words:
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_STACK,
0
)
gs.state_bits[i] = set_state_flag(
gs.state_bits[i],
HEAD_IN_BUFFER,
1
)
2020-06-19 03:29:16 +03:00
cdef class ArcEagerGold:
2020-06-19 01:11:59 +03:00
cdef GoldParseStateC c
2020-06-20 22:49:46 +03:00
cdef Pool mem
2020-06-19 03:29:16 +03:00
def __init__(self, ArcEager moves, StateClass stcls, Example example):
2020-06-19 01:11:59 +03:00
self.mem = Pool()
self.c = create_gold_state(self.mem, stcls, example)
2020-06-21 02:01:09 +03:00
def update(self, StateClass stcls):
update_gold_state(&self.c, stcls)
2020-06-19 01:11:59 +03:00
cdef int check_state_gold(char state_bits, char flag) nogil:
2020-06-14 18:22:14 +03:00
cdef char one = 1
return state_bits & (one << flag)
cdef int set_state_flag(char state_bits, char flag, int value) nogil:
cdef char one = 1
if value:
return state_bits | (one << flag)
else:
return state_bits & ~(one << flag)
cdef int is_head_in_stack(const GoldParseStateC* gold, int i) nogil:
return check_state_gold(gold.state_bits[i], HEAD_IN_STACK)
2020-06-14 18:22:14 +03:00
cdef int is_head_in_buffer(const GoldParseStateC* gold, int i) nogil:
2020-06-14 18:22:14 +03:00
return check_state_gold(gold.state_bits[i], HEAD_IN_BUFFER)
cdef int is_head_unknown(const GoldParseStateC* gold, int i) nogil:
2020-06-14 18:22:14 +03:00
return check_state_gold(gold.state_bits[i], HEAD_UNKNOWN)
# Helper functions for the arc-eager oracle
cdef weight_t push_cost(StateClass stcls, const void* _gold, int target) nogil:
gold = <const GoldParseStateC*>_gold
cdef weight_t cost = 0
if is_head_in_stack(gold, target):
2020-06-14 18:22:14 +03:00
cost += 1
2020-06-21 02:04:02 +03:00
cost += gold.n_kids_in_stack[target]
2020-06-14 18:22:14 +03:00
if Break.is_valid(stcls.c, 0) and Break.move_cost(stcls, gold) == 0:
cost += 1
return cost
cdef weight_t pop_cost(StateClass stcls, const void* _gold, int target) nogil:
gold = <const GoldParseStateC*>_gold
cdef weight_t cost = 0
if is_head_in_buffer(gold, target):
2020-06-14 18:22:14 +03:00
cost += 1
cost += gold[0].n_kids_in_buffer[target]
2017-05-30 21:37:24 +03:00
if Break.is_valid(stcls.c, 0) and Break.move_cost(stcls, gold) == 0:
cost += 1
return cost
cdef weight_t arc_cost(StateClass stcls, const void* _gold, int head, int child) nogil:
gold = <const GoldParseStateC*>_gold
if arc_is_gold(gold, head, child):
return 0
elif stcls.H(child) == gold.heads[child]:
return 1
# Head in buffer
2017-05-30 21:37:24 +03:00
elif gold.heads[child] >= stcls.B(0) and stcls.B(1) != 0:
return 1
else:
return 0
2020-06-14 18:22:14 +03:00
cdef bint arc_is_gold(const GoldParseStateC* gold, int head, int child) nogil:
if is_head_unknown(gold, child):
return True
elif gold.heads[child] == head:
return True
else:
return False
2020-06-14 18:22:14 +03:00
cdef bint label_is_gold(const GoldParseStateC* gold, int head, int child, attr_t label) nogil:
if is_head_unknown(gold, child):
return True
2017-05-30 21:37:24 +03:00
elif label == 0:
return True
elif gold.labels[child] == label:
return True
else:
return False
2020-06-14 18:22:14 +03:00
cdef bint _is_gold_root(const GoldParseStateC* gold, int word) nogil:
return gold.heads[word] == word or is_head_unknown(gold, word)
2020-06-14 18:22:14 +03:00
cdef class Shift:
@staticmethod
2017-05-28 15:06:40 +03:00
cdef bint is_valid(const StateC* st, attr_t label) nogil:
sent_start = st._sent[st.B_(0).l_edge].sent_start
return st.buffer_length() >= 2 and not st.shifted[st.B(0)] and sent_start != 1
2015-06-09 02:41:09 +03:00
@staticmethod
2017-05-28 15:06:40 +03:00
cdef int transition(StateC* st, attr_t label) nogil:
st.push()
st.fast_forward()
2014-12-16 14:44:43 +03:00
@staticmethod
2020-06-14 18:22:14 +03:00
cdef weight_t cost(StateClass st, const void* _gold, attr_t label) nogil:
gold = <const GoldParseStateC*>_gold
return Shift.move_cost(st, gold) + Shift.label_cost(st, gold, label)
@staticmethod
cdef inline weight_t move_cost(StateClass s, const void* _gold) nogil:
gold = <const GoldParseStateC*>_gold
return push_cost(s, gold, s.B(0))
@staticmethod
cdef inline weight_t label_cost(StateClass s, const void* _gold, attr_t label) nogil:
return 0
cdef class Reduce:
2015-06-09 02:41:09 +03:00
@staticmethod
2017-05-28 15:06:40 +03:00
cdef bint is_valid(const StateC* st, attr_t label) nogil:
return st.stack_depth() >= 2
2015-06-09 02:41:09 +03:00
@staticmethod
2017-05-28 15:06:40 +03:00
cdef int transition(StateC* st, attr_t label) nogil:
if st.has_head(st.S(0)):
st.pop()
else:
st.unshift()
st.fast_forward()
@staticmethod
2020-06-14 18:22:14 +03:00
cdef weight_t cost(StateClass s, const void* _gold, attr_t label) nogil:
gold = <const GoldParseStateC*>_gold
return Reduce.move_cost(s, gold) + Reduce.label_cost(s, gold, label)
@staticmethod
cdef inline weight_t move_cost(StateClass st, const void* _gold) nogil:
gold = <const GoldParseStateC*>_gold
2020-06-14 18:22:14 +03:00
s0 = st.S(0)
cost = pop_cost(st, gold, s0)
return_to_buffer = not st.has_head(s0)
if return_to_buffer:
# Decrement cost for the arcs we save, as we'll be putting this
# back to the buffer
if is_head_in_stack(gold, s0):
2020-06-14 18:22:14 +03:00
cost -= 1
cost -= gold.n_kids_in_stack[s0]
2017-05-30 21:37:24 +03:00
if Break.is_valid(st.c, 0) and Break.move_cost(st, gold) == 0:
cost -= 1
return cost
@staticmethod
cdef inline weight_t label_cost(StateClass s, const void* gold, attr_t label) nogil:
return 0
cdef class LeftArc:
2015-06-09 02:41:09 +03:00
@staticmethod
2017-05-28 15:06:40 +03:00
cdef bint is_valid(const StateC* st, attr_t label) nogil:
if label == SUBTOK_LABEL and st.S(0) != (st.B(0)-1):
2018-04-08 15:50:21 +03:00
return 0
sent_start = st._sent[st.B_(0).l_edge].sent_start
return sent_start != 1
2015-06-09 02:41:09 +03:00
@staticmethod
2017-05-28 15:06:40 +03:00
cdef int transition(StateC* st, attr_t label) nogil:
st.add_arc(st.B(0), st.S(0), label)
st.pop()
st.fast_forward()
@staticmethod
cdef inline weight_t cost(StateClass s, const void* _gold, attr_t label) nogil:
2020-06-14 18:22:14 +03:00
gold = <const GoldParseStateC*>_gold
2020-06-21 02:01:09 +03:00
return LeftArc.move_cost(s, gold) + LeftArc.label_cost(s, gold, label)
@staticmethod
2020-06-21 02:01:09 +03:00
cdef inline weight_t move_cost(StateClass s, const GoldParseStateC* gold) nogil:
cdef weight_t cost = 0
s0 = s.S(0)
b0 = s.B(0)
if arc_is_gold(gold, b0, s0):
# Have a negative cost if we 'recover' from the wrong dependency
return 0 if not s.has_head(s0) else -1
else:
2020-06-21 02:01:09 +03:00
# Account for deps we might lose between S0 and stack
if not s.has_head(s0):
cost += gold.n_kids_in_stack[s0]
if is_head_in_buffer(gold, s0):
cost += 1
return cost + pop_cost(s, gold, s.S(0)) + arc_cost(s, gold, s.B(0), s.S(0))
@staticmethod
2020-06-21 02:01:09 +03:00
cdef inline weight_t label_cost(StateClass s, const GoldParseStateC* gold, attr_t label) nogil:
return arc_is_gold(gold, s.B(0), s.S(0)) and not label_is_gold(gold, s.B(0), s.S(0), label)
cdef class RightArc:
@staticmethod
cdef bint is_valid(const StateC* st, attr_t label) nogil:
# If there's (perhaps partial) parse pre-set, don't allow cycle.
if label == SUBTOK_LABEL and st.S(0) != (st.B(0)-1):
return 0
sent_start = st._sent[st.B_(0).l_edge].sent_start
return sent_start != 1 and st.H(st.S(0)) != st.B(0)
@staticmethod
cdef int transition(StateC* st, attr_t label) nogil:
st.add_arc(st.S(0), st.B(0), label)
st.push()
st.fast_forward()
@staticmethod
cdef inline weight_t cost(StateClass s, const void* _gold, attr_t label) nogil:
gold = <const GoldParseStateC*>_gold
return RightArc.move_cost(s, gold) + RightArc.label_cost(s, gold, label)
@staticmethod
cdef inline weight_t move_cost(StateClass s, const void* _gold) nogil:
gold = <const GoldParseStateC*>_gold
if arc_is_gold(gold, s.S(0), s.B(0)):
return 0
elif s.c.shifted[s.B(0)]:
return push_cost(s, gold, s.B(0))
else:
return push_cost(s, gold, s.B(0)) + arc_cost(s, gold, s.S(0), s.B(0))
@staticmethod
cdef weight_t label_cost(StateClass s, const void* _gold, attr_t label) nogil:
gold = <const GoldParseStateC*>_gold
return arc_is_gold(gold, s.S(0), s.B(0)) and not label_is_gold(gold, s.S(0), s.B(0), label)
cdef class Break:
2015-06-09 02:41:09 +03:00
@staticmethod
2017-05-28 15:06:40 +03:00
cdef bint is_valid(const StateC* st, attr_t label) nogil:
2015-06-09 02:41:09 +03:00
cdef int i
if not USE_BREAK:
return False
elif st.at_break():
2015-06-09 02:41:09 +03:00
return False
elif st.stack_depth() < 1:
return False
elif st.B_(0).l_edge < 0:
return False
2017-10-09 01:02:45 +03:00
elif st._sent[st.B_(0).l_edge].sent_start < 0:
return False
2015-06-09 02:41:09 +03:00
else:
2017-08-26 03:50:55 +03:00
return True
2015-06-09 02:41:09 +03:00
@staticmethod
2017-05-28 15:06:40 +03:00
cdef int transition(StateC* st, attr_t label) nogil:
st.set_break(st.B_(0).l_edge)
st.fast_forward()
@staticmethod
2020-06-14 18:22:14 +03:00
cdef weight_t cost(StateClass s, const void* _gold, attr_t label) nogil:
gold = <const GoldParseStateC*>_gold
return Break.move_cost(s, gold) + Break.label_cost(s, gold, label)
@staticmethod
cdef inline weight_t move_cost(StateClass s, const void* _gold) nogil:
gold = <const GoldParseStateC*>_gold
cdef weight_t cost = 0
cdef int i, j, S_i, B_i
for i in range(s.stack_depth()):
S_i = s.S(i)
2020-06-14 18:22:14 +03:00
cost += gold.n_kids_in_buffer[S_i]
if is_head_in_buffer(gold, S_i):
2020-06-14 18:22:14 +03:00
cost += 1
# Check for sentence boundary --- if it's here, we can't have any deps
# between stack and buffer, so rest of action is irrelevant.
s0_root = _get_root(s.S(0), gold)
b0_root = _get_root(s.B(0), gold)
if s0_root != b0_root or s0_root == -1 or b0_root == -1:
return cost
else:
return cost + 1
2015-08-09 00:31:54 +03:00
@staticmethod
cdef inline weight_t label_cost(StateClass s, const void* gold, attr_t label) nogil:
return 0
2020-06-14 18:22:14 +03:00
cdef int _get_root(int word, const GoldParseStateC* gold) nogil:
if is_head_unknown(gold, word):
return -1
2020-06-14 18:22:14 +03:00
while gold.heads[word] != word and word >= 0:
word = gold.heads[word]
if is_head_unknown(gold, word):
2020-06-14 18:22:14 +03:00
return -1
else:
return word
2015-08-09 00:31:54 +03:00
2017-03-11 20:12:01 +03:00
cdef void* _init_state(Pool mem, int length, void* tokens) except NULL:
2017-11-14 04:11:40 +03:00
st = new StateC(<const TokenC*>tokens, length)
for i in range(st.length):
if st._sent[i].dep == 0:
st._sent[i].l_edge = i
st._sent[i].r_edge = i
st._sent[i].head = 0
st._sent[i].dep = 0
st._sent[i].l_kids = 0
st._sent[i].r_kids = 0
2017-03-11 20:12:01 +03:00
st.fast_forward()
return <void*>st
cdef int _del_state(Pool mem, void* state, void* x) except -1:
cdef StateC* st = <StateC*>state
del st
cdef class ArcEager(TransitionSystem):
2017-03-11 20:12:01 +03:00
def __init__(self, *args, **kwargs):
TransitionSystem.__init__(self, *args, **kwargs)
@classmethod
def get_actions(cls, **kwargs):
min_freq = kwargs.get('min_freq', None)
actions = defaultdict(lambda: Counter())
actions[SHIFT][''] = 1
actions[REDUCE][''] = 1
for label in kwargs.get('left_labels', []):
actions[LEFT][label] = 1
actions[SHIFT][label] = 1
for label in kwargs.get('right_labels', []):
actions[RIGHT][label] = 1
actions[REDUCE][label] = 1
for example in kwargs.get('gold_parses', []):
2020-06-19 12:31:01 +03:00
heads, labels = nonproj.projectivize(example.get_aligned("HEAD"),
example.get_aligned("DEP"))
for child, head, label in zip(example.get_aligned("ID"), heads, labels):
Restructure Example with merged sents as default (#4632) * Switch to train_dataset() function in train CLI * Fixes for pipe() methods in pipeline components * Don't clobber `examples` variable with `as_example` in pipe() methods * Remove unnecessary traversals of `examples` * Update Parser.pipe() for Examples * Add `as_examples` kwarg to `pipe()` with implementation to return `Example`s * Accept `Doc` or `Example` in `pipe()` with `_get_doc()` (copied from `Pipe`) * Fixes to Example implementation in spacy.gold * Move `make_projective` from an attribute of Example to an argument of `Example.get_gold_parses()` * Head of 0 are not treated as unset * Unset heads are set to self rather than `None` (which causes problems while projectivizing) * Check for `Doc` (not just not `None`) when creating GoldParses for pre-merged example * Don't clobber `examples` variable in `iter_gold_docs()` * Add/modify gold tests for handling projectivity * In JSON roundtrip compare results from `dev_dataset` rather than `train_dataset` to avoid projectivization (and other potential modifications) * Add test for projective train vs. nonprojective dev versions of the same `Doc` * Handle ignore_misaligned as arg rather than attr Move `ignore_misaligned` from an attribute of `Example` to an argument to `Example.get_gold_parses()`, which makes it parallel to `make_projective`. Add test with old and new align that checks whether `ignore_misaligned` errors are raised as expected (only for new align). * Remove unused attrs from gold.pxd Remove `ignore_misaligned` and `make_projective` from `gold.pxd` * Restructure Example with merged sents as default An `Example` now includes a single `TokenAnnotation` that includes all the information from one `Doc` (=JSON `paragraph`). If required, the individual sentences can be returned as a list of examples with `Example.split_sents()` with no raw text available. * Input/output a single `Example.token_annotation` * Add `sent_starts` to `TokenAnnotation` to handle sentence boundaries * Replace `Example.merge_sents()` with `Example.split_sents()` * Modify components to use a single `Example.token_annotation` * Pipeline components * conllu2json converter * Rework/rename `add_token_annotation()` and `add_doc_annotation()` to `set_token_annotation()` and `set_doc_annotation()`, functions that set rather then appending/extending. * Rename `morphology` to `morphs` in `TokenAnnotation` and `GoldParse` * Add getters to `TokenAnnotation` to supply default values when a given attribute is not available * `Example.get_gold_parses()` in `spacy.gold._make_golds()` is only applied on single examples, so the `GoldParse` is returned saved in the provided `Example` rather than creating a new `Example` with no other internal annotation * Update tests for API changes and `merge_sents()` vs. `split_sents()` * Refer to Example.goldparse in iter_gold_docs() Use `Example.goldparse` in `iter_gold_docs()` instead of `Example.gold` because a `None` `GoldParse` is generated with ignore_misaligned and generating it on-the-fly can raise an unwanted AlignmentError * Fix make_orth_variants() Fix bug in make_orth_variants() related to conversion from multiple to one TokenAnnotation per Example. * Add basic test for make_orth_variants() * Replace try/except with conditionals * Replace default morph value with set
2019-11-25 18:03:28 +03:00
if label.upper() == 'ROOT' :
label = 'ROOT'
if head == child:
actions[BREAK][label] += 1
elif head < child:
actions[RIGHT][label] += 1
actions[REDUCE][''] += 1
elif head > child:
actions[LEFT][label] += 1
actions[SHIFT][''] += 1
if min_freq is not None:
for action, label_freqs in actions.items():
for label, freq in list(label_freqs.items()):
if freq < min_freq:
label_freqs.pop(label)
# Ensure these actions are present
actions[BREAK].setdefault('ROOT', 0)
if kwargs.get("learn_tokens") is True:
actions[RIGHT].setdefault('subtok', 0)
actions[LEFT].setdefault('subtok', 0)
# Used for backoff
actions[RIGHT].setdefault('dep', 0)
actions[LEFT].setdefault('dep', 0)
return actions
@property
def action_types(self):
return (SHIFT, REDUCE, LEFT, RIGHT, BREAK)
def get_cost(self, StateClass state, Example gold, action):
2020-06-14 18:22:14 +03:00
raise NotImplementedError
def transition(self, StateClass state, action):
cdef Transition t = self.lookup_transition(action)
t.do(state.c, t.label)
return state
2020-06-14 18:22:14 +03:00
def is_gold_parse(self, StateClass state, gold):
raise NotImplementedError
2017-05-26 19:31:23 +03:00
2020-06-14 18:22:14 +03:00
def has_gold(self, gold, start=0, end=None):
raise NotImplementedError
def preprocess_gold(self, example):
2020-06-14 18:22:14 +03:00
raise NotImplementedError
2020-06-19 01:11:59 +03:00
def init_gold_batch(self, examples):
states = self.init_batch([eg.predicted for eg in examples])
keeps = [i for i, s in enumerate(states) if not s.is_final()]
states = [states[i] for i in keeps]
2020-06-19 03:29:16 +03:00
golds = [ArcEagerGold(self, states[i], examples[i]) for i in keeps]
2020-06-20 03:36:12 +03:00
n_steps = sum([len(s.queue) * 4 for s in states])
2020-06-19 03:29:16 +03:00
return states, golds, n_steps
2020-06-19 01:11:59 +03:00
cdef Transition lookup_transition(self, object name_or_id) except *:
if isinstance(name_or_id, int):
return self.c[name_or_id]
name = name_or_id
if '-' in name:
move_str, label_str = name.split('-', 1)
2015-08-09 00:31:54 +03:00
label = self.strings[label_str]
else:
2015-08-09 00:31:54 +03:00
move_str = name
label = 0
move = MOVE_NAMES.index(move_str)
for i in range(self.n_moves):
if self.c[i].move == move and self.c[i].label == label:
return self.c[i]
return Transition(clas=0, move=MISSING, label=0)
2017-05-28 15:06:40 +03:00
def move_name(self, int move, attr_t label):
label_str = self.strings[label]
if label_str:
return MOVE_NAMES[move] + '-' + label_str
else:
return MOVE_NAMES[move]
def class_name(self, int i):
return self.move_name(self.c[i].move, self.c[i].label)
2017-05-28 15:06:40 +03:00
cdef Transition init_transition(self, int clas, int move, attr_t label) except *:
# TODO: Apparent Cython bug here when we try to use the Transition()
# constructor with the function pointers
cdef Transition t
t.score = 0
t.clas = clas
t.move = move
t.label = label
if move == SHIFT:
t.is_valid = Shift.is_valid
t.do = Shift.transition
t.get_cost = Shift.cost
elif move == REDUCE:
t.is_valid = Reduce.is_valid
t.do = Reduce.transition
t.get_cost = Reduce.cost
elif move == LEFT:
t.is_valid = LeftArc.is_valid
t.do = LeftArc.transition
t.get_cost = LeftArc.cost
elif move == RIGHT:
t.is_valid = RightArc.is_valid
t.do = RightArc.transition
t.get_cost = RightArc.cost
elif move == BREAK:
t.is_valid = Break.is_valid
t.do = Break.transition
t.get_cost = Break.cost
else:
raise ValueError(Errors.E019.format(action=move, src='arc_eager'))
return t
cdef int initialize_state(self, StateC* st) nogil:
for i in range(st.length):
if st._sent[i].dep == 0:
st._sent[i].l_edge = i
st._sent[i].r_edge = i
st._sent[i].head = 0
st._sent[i].dep = 0
st._sent[i].l_kids = 0
st._sent[i].r_kids = 0
st.fast_forward()
cdef int finalize_state(self, StateC* st) nogil:
cdef int i
for i in range(st.length):
if st._sent[i].head == 0:
st._sent[i].dep = self.root_label
2018-05-02 17:19:22 +03:00
def finalize_doc(self, Doc doc):
doc.is_parsed = True
2018-05-02 17:19:22 +03:00
set_children_from_heads(doc.c, doc.length)
2016-02-01 05:07:37 +03:00
cdef int set_valid(self, int* output, const StateC* st) nogil:
cdef bint[N_MOVES] is_valid
2017-05-30 21:37:24 +03:00
is_valid[SHIFT] = Shift.is_valid(st, 0)
is_valid[REDUCE] = Reduce.is_valid(st, 0)
is_valid[LEFT] = LeftArc.is_valid(st, 0)
is_valid[RIGHT] = RightArc.is_valid(st, 0)
is_valid[BREAK] = Break.is_valid(st, 0)
cdef int i
for i in range(self.n_moves):
if self.c[i].label == SUBTOK_LABEL:
output[i] = self.c[i].is_valid(st, self.c[i].label)
else:
output[i] = is_valid[self.c[i].move]
cdef int set_costs(self, int* is_valid, weight_t* costs,
2020-06-20 22:49:46 +03:00
StateClass stcls, gold) except -1:
2020-06-21 02:01:09 +03:00
if not isinstance(gold, ArcEagerGold):
raise TypeError("Expected ArcEagerGold")
cdef ArcEagerGold gold_ = gold
gold_.update(stcls)
gold_state = gold_.c
n_gold = 0
for i in range(self.n_moves):
if self.c[i].is_valid(stcls.c, self.c[i].label):
is_valid[i] = True
2020-06-21 02:01:09 +03:00
costs[i] = self.c[i].get_cost(stcls, &gold_state, self.c[i].label)
n_gold += 1
2015-06-10 07:56:35 +03:00
else:
is_valid[i] = False
costs[i] = 9000
if n_gold < 1:
2020-06-20 22:49:46 +03:00
raise ValueError
#failure_state = stcls.print_state([t.text for t in example])
#raise ValueError(
# Errors.E021.format(n_actions=self.n_moves, state=failure_state))