* Tmp commit of NER refactoring

spacy/structs.pxd

@@ -1,4 +1,4 @@
-from libc.stdint cimport int8_t, uint8_t, uint16_t, uint32_t
+from libc.stdint cimport uint8_t, uint32_t, int32_t
 
 from .typedefs cimport flags_t, attr_t, id_t, hash_t
 from .parts_of_speech cimport univ_pos_t
@@ -42,28 +42,9 @@ cdef struct PosTag:
     univ_pos_t pos
 
 
-# Start and end will be offsets: i + ent.start will always take you to the
-# "next" entity start. If inside an entity, ent.start will be negative ---
-# the next entity is the start of the one the token is inside.  If i _is_
-# the start of an entity, then ent.start will be the beginning of the next one.
-#
-# The same/inverse is true for end. If ent.end has a negative value, we are either
-# at the end of an entity, or outside one.  If we're inside an entity, ent.end
-# will have a positive value.
-#
-# This allows us to easily find the span of an entity we might be inside, while
-# naturally sharing an API with iterating through all entities in the sentence
-cdef struct Entity:
-    int32_t tag
-    uint16_t flags
-    int8_t start
-    int8_t end
-
-
 cdef struct TokenC:
     const LexemeC* lex
     Morphology morph
-    Entity ent
     univ_pos_t pos
     int tag
     int idx

spacy/syntax/arc_eager.pyx

@@ -123,7 +123,7 @@ cdef int _break_cost(const State* s, const int* gold) except -1:
     return cost
 
 
-cdef class TransitionSystem:
+cdef class ArcEager(TransitionSystem):
     def __init__(self, list left_labels, list right_labels):
         self.mem = Pool()
         left_labels.sort()
@@ -163,7 +163,7 @@ cdef class TransitionSystem:
         moves[i].label = 0
         moves[i].clas = i
         i += 1
-        self._moves = moves
+        self.c = moves
 
     cdef int transition(self, State *s, const Transition* t) except -1:
         if t.move == SHIFT:

spacy/syntax/parser.pyx

@@ -83,7 +83,7 @@ cdef class GreedyParser:
             fill_context(context, state)
             scores = self.model.score(context)
             guess = self.moves.best_valid(scores, state)
-            self.moves.transition(state, &guess)
+            guess.do(&guess, state)
         # Messily tell Tokens object the string names of the dependency labels
         dep_strings = [None] * len(self.moves.label_ids)
         for label, id_ in self.moves.label_ids.items():
@@ -129,9 +129,9 @@ cdef class GreedyParser:
             history.append((py_moves[best.move], print_state(state, py_words)))
             self.model.update(context, guess.clas, best.clas, guess.cost)
             if force_gold:
-                self.moves.transition(state, &best)
+                best.do(&best, state)
             else:
-                self.moves.transition(state, &guess)
+                guess.do(&guess, state)
         cdef int n_corr = 0
         for i in range(tokens.length):
             if gold_heads[i] != -1: