def volt_throttler(c):
(part, attr) = c
if same_table((part, attr)):
return is_horz_aligned((part, attr)) or is_vert_aligned((part, attr))
if LF_part_miss_match((part, attr)) < 0:
return False
return True
def stg_temp_filter(c):
(part, attr) = c
if same_table((part, attr)):
return is_horz_aligned((part, attr)) or is_vert_aligned((part, attr))
# Filter if not inside of a table
return part.context.sentence.is_tabular()
def LF_part_ce_keywords_in_row_prefix_same_table(c):
ngrams_part = _filter_non_parts(get_row_ngrams(c[1], n_max=3))
return (TRUE if same_table(c) and is_horz_aligned(c) and overlap(
_CE_KEYWORDS.union(_CE_ABBREVS), get_row_ngrams(c[1], n_max=3))
and overlap(_CE_KEYWORDS.union(_CE_ABBREVS),
get_row_ngrams(c.part, n_max=3)) and any([
c.part.context.get_span().lower().startswith(_)
for _ in ngrams_part
])
and not overlap(_NON_CEV_KEYWORDS, get_row_ngrams(c.part, n_max=3))
and not overlap(_NON_CEV_KEYWORDS, get_row_ngrams(c[1], n_max=3))
and not LF_current_in_row(c) else ABSTAIN)
def polarity_filter(c):
(part, attr) = c
# Check if the polarities are not matched with the part
ngrams_part = set(x for x in get_row_ngrams(part, n_max=1, lower=False)
if (x and polarity_pattern.match(x)))
if len(ngrams_part) != 0 and all(
not attr.context.get_span().lower().startswith(_.lower())
for _ in ngrams_part):
logger.debug(
f"ngrams_part: {ngrams_part}\nattr: {attr.context.get_span().lower()}"
)
return False
if same_table(c):
return is_horz_aligned(c) or is_vert_aligned(c)
return True
def _vizlib_multinary_features(
spans: Tuple[SpanMention, ...]) -> Iterator[Tuple[str, int]]:
"""Visual-related features for multiple spans."""
if same_page(spans):
yield "SAME_PAGE", DEF_VALUE
if is_horz_aligned(spans):
yield "HORZ_ALIGNED", DEF_VALUE
if is_vert_aligned(spans):
yield "VERT_ALIGNED", DEF_VALUE
if is_vert_aligned_left(spans):
yield "VERT_ALIGNED_LEFT", DEF_VALUE
if is_vert_aligned_right(spans):
yield "VERT_ALIGNED_RIGHT", DEF_VALUE
if is_vert_aligned_center(spans):
yield "VERT_ALIGNED_CENTER", DEF_VALUE
def _vizlib_binary_features(span1: SpanMention,
span2: SpanMention) -> Iterator[Tuple[str, int]]:
"""Visual-related features for a pair of spans."""
if same_page((span1, span2)):
yield "SAME_PAGE", DEF_VALUE
if is_horz_aligned((span1, span2)):
yield "HORZ_ALIGNED", DEF_VALUE
if is_vert_aligned((span1, span2)):
yield "VERT_ALIGNED", DEF_VALUE
if is_vert_aligned_left((span1, span2)):
yield "VERT_ALIGNED_LEFT", DEF_VALUE
if is_vert_aligned_right((span1, span2)):
yield "VERT_ALIGNED_RIGHT", DEF_VALUE
if is_vert_aligned_center((span1, span2)):
yield "VERT_ALIGNED_CENTER", DEF_VALUE
def ce_v_max_filter(c):
(part, attr) = c
if same_table(c):
return is_horz_aligned(c) or is_vert_aligned(c)
# Check if the ce_v_max's are not matched with the part
ngrams_part = _filter_non_parts(
set(x for x in get_vert_ngrams(attr, n_max=1)))
ngrams_part = _filter_non_parts(
ngrams_part.union(set(x for x in get_horz_ngrams(attr, n_max=1))))
if len(ngrams_part) != 0 and all(
not part.context.get_span().lower().startswith(_.lower())
for _ in ngrams_part):
logger.debug(
f"ngrams_part: {ngrams_part}\npart: {part.context.get_span().lower()}"
)
return False
return True
def _vizlib_binary_features(span1, span2):
"""
Visual-related features for a pair of spans
"""
if same_page((span1, span2)):
yield "SAME_PAGE", DEF_VALUE
if is_horz_aligned((span1, span2)):
yield "HORZ_ALIGNED", DEF_VALUE
if is_vert_aligned((span1, span2)):
yield "VERT_ALIGNED", DEF_VALUE
if is_vert_aligned_left((span1, span2)):
yield "VERT_ALIGNED_LEFT", DEF_VALUE
if is_vert_aligned_right((span1, span2)):
yield "VERT_ALIGNED_RIGHT", DEF_VALUE
if is_vert_aligned_center((span1, span2)):
yield "VERT_ALIGNED_CENTER", DEF_VALUE
def LF_same_table_must_align(c):
return (FALSE if
(same_table(c)
and not (is_horz_aligned(c) or is_vert_aligned(c))) else ABSTAIN)
def LF_aligned_or_global(c):
return (TRUE if (same_row(c) or is_horz_aligned(c) or same_col(c)
or is_vert_aligned(c)
or not c.part.context.sentence.is_tabular()) else ABSTAIN)
def LF_polarity_part_viz_aligned(c):
return TRUE if is_horz_aligned(c) or is_vert_aligned(c) else ABSTAIN
def temp_throttler(c):
"""Temperature throttler."""
(part, attr) = c
if same_table((part, attr)):
return is_horz_aligned((part, attr)) or is_vert_aligned((part, attr))
return True
