explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-08-04 12:20:20 +03:00
Code Issues Packages Projects Releases Wiki Activity

Work in progress

Browse Source
This commit is contained in:
Richard Hudson 2021-11-26 22:01:35 +01:00
parent c446d56ce8
commit 9d97a0ff0c
1 changed files with 29 additions and 46 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

69
spacy/visualization.py
View File

@ -18,34 +18,35 @@ class Visualizer:
for token in doc for token in doc
] ]
children_lists = [[] for _ in range(len(doc))] children_lists = [[] for _ in range(len(doc))]
for child, head in enumerate(h for h in heads): for child, head in enumerate(heads):
if head != -1: if head != -1:
children_lists[head].append(child) children_lists[head].append(child)
all_indices_ordered_by_column = [] all_indices_ordered_by_column = []
indices_in_current_column = [i for i, h in enumerate(heads) if h == -1] indices_in_current_column = [i for i, h in enumerate(heads) if h == -1]
while len(indices_in_current_column) > 0: while len(indices_in_current_column) > 0:
all_indices_ordered_by_column.extend(indices_in_current_column) all_indices_ordered_by_column = (
indices_in_current_column + all_indices_ordered_by_column
)
indices_in_current_column = [ indices_in_current_column = [
i i
for index_in_current_column in indices_in_current_column for index_in_current_column in indices_in_current_column
for i in children_lists[index_in_current_column] for i in children_lists[index_in_current_column]
if i not in all_indices_ordered_by_column if i not in all_indices_ordered_by_column
] ]
all_indices_ordered_by_column = reversed(all_indices_ordered_by_column) # -1: root token with no arrow; 0: length not yet set
# -1: no arrow; None: not yet set horizontal_line_lengths = [
horizontal_arrow_positions = [
-1 -1
if heads[i] == -1 if heads[i] == -1
else 0 else 1
if len(children_lists[i]) == 0 and abs(heads[i] - i) == 1 if len(children_lists[i]) == 0 and abs(heads[i] - i) == 1
else None else 0
for i in range(len(doc)) for i in range(len(doc))
] ]
while None in horizontal_arrow_positions: while 0 in horizontal_line_lengths:
for working_token_index in ( for working_token_index in (
i i
for i in all_indices_ordered_by_column for i in all_indices_ordered_by_column
if horizontal_arrow_positions[i] is None if horizontal_line_lengths[i] == 0
): ):
first_index_in_relation = min( first_index_in_relation = min(
working_token_index, working_token_index,
@ -55,61 +56,43 @@ class Visualizer:
working_token_index, working_token_index,
heads[working_token_index], heads[working_token_index],
) )
inbetween_indexes = range(
first_index_in_relation + 1, second_index_in_relation
)
inbetween_horizontal_arrow_positions = horizontal_arrow_positions[
first_index_in_relation + 1 : second_index_in_relation
]
if (
-1 in inbetween_horizontal_arrow_positions
and None in inbetween_horizontal_arrow_positions
):
continue
if None in (
horizontal_arrow_positions[i]
for i in children_lists[working_token_index]
):
continue
if len(children_lists[working_token_index]) > 0: if len(children_lists[working_token_index]) > 0:
working_horizontal_arrow_position = ( working_horizontal_arrow_position = max(
max(
[ [
horizontal_arrow_positions[i] horizontal_line_lengths[i]
for i in children_lists[working_token_index] for i in children_lists[working_token_index]
] ]
) )
+ 1
)
else: else:
working_horizontal_arrow_position = -1 working_horizontal_arrow_position = 0
for inbetween_index in ( for inbetween_index in (
i i
for i in inbetween_indexes for i in range(
first_index_in_relation + 1, second_index_in_relation
)
if i not in children_lists[working_token_index] if i not in children_lists[working_token_index]
and horizontal_arrow_positions[i] is not None and horizontal_line_lengths[i] != 0
): ):
working_horizontal_arrow_position = max( working_horizontal_arrow_position = max(
working_horizontal_arrow_position, working_horizontal_arrow_position,
horizontal_arrow_positions[inbetween_index] horizontal_line_lengths[inbetween_index] - 1
if inbetween_index in children_lists[heads[working_token_index]] if inbetween_index in children_lists[heads[working_token_index]]
else horizontal_arrow_positions[inbetween_index] + 1, else horizontal_line_lengths[inbetween_index],
) )
for child_horizontal_arrow_position in ( for child_horizontal_arrow_position in (
horizontal_arrow_positions[i] horizontal_line_lengths[i]
for i in children_lists[working_token_index] for i in children_lists[working_token_index]
if (i < first_index_in_relation or i > second_index_in_relation) if (i < first_index_in_relation or i > second_index_in_relation)
and horizontal_arrow_positions[i] is not None and horizontal_line_lengths[i] != 0
): ):
working_horizontal_arrow_position = max( working_horizontal_arrow_position = max(
working_horizontal_arrow_position, working_horizontal_arrow_position,
child_horizontal_arrow_position + 1, child_horizontal_arrow_position,
) )
if working_horizontal_arrow_position > -1: horizontal_line_lengths[working_token_index] = (
horizontal_arrow_positions[ working_horizontal_arrow_position + 1
working_token_index )
] = working_horizontal_arrow_position print(horizontal_line_lengths)
print(horizontal_arrow_positions)
import spacy import spacy