Spaces:
Runtime error
Runtime error
File size: 3,769 Bytes
dbdb640 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 |
"""
This module is taken with slight modifications from:
https://github.com/PrithivirajDamodaran/Gramformer/blob/main/gramformer/gramformer.py
"""
import re
import pandas as pd
from annotated_text import annotated_text
from bs4 import BeautifulSoup
def show_highlights(annotator, input_text, corrected_sentence):
strikeout = lambda x: "\u0336".join(x) + "\u0336"
highlight_text = highlight(annotator, input_text, corrected_sentence)
color_map = {"d": "#faa", "a": "#afa", "c": "#fea"}
tokens = re.split(r"(<[dac]\s.*?<\/[dac]>)", highlight_text)
annotations = []
for token in tokens:
soup = BeautifulSoup(token, "html.parser")
tags = soup.findAll()
if tags:
_tag = tags[0].name
_type = tags[0]["type"]
_text = tags[0]["edit"]
_color = color_map[_tag]
if _tag == "d":
_text = strikeout(tags[0].text)
annotations.append((_text, _type, _color))
else:
annotations.append(token)
annotated_text(*annotations)
def show_edits(annotator, input_text, corrected_sentence):
edits = get_edits(annotator, input_text, corrected_sentence)
df = pd.DataFrame(
edits,
columns=[
"type",
"original word",
"original start",
"original end",
"correct word",
"correct start",
"correct end",
],
)
return df.set_index("type")
def highlight(annotator, orig, cor):
edits = get_edits(annotator, orig, cor)
orig_tokens = orig.split()
ignore_indexes = []
for edit in edits:
edit_type = edit[0]
edit_str_start = edit[1]
edit_spos = edit[2]
edit_epos = edit[3]
edit_str_end = edit[4]
for i in range(edit_spos + 1, edit_epos):
ignore_indexes.append(i)
if edit_str_start == "":
if edit_spos >= 1:
new_edit_str = orig_tokens[edit_spos - 1]
# print("edit_spos >= 1", new_edit_str)
edit_spos -= 1
else:
new_edit_str = orig_tokens[edit_spos + 1]
# print("new", new_edit_str)
edit_spos += 1
if edit_type == "PUNCT":
st = (
"<a type='"
+ edit_type
+ "' edit='"
+ edit_str_end
+ "'>"
+ new_edit_str
+ "</a>"
)
else:
st = (
"<a type='"
+ edit_type
+ "' edit='"
+ new_edit_str
+ " "
+ edit_str_end
+ "'>"
+ new_edit_str
+ "</a>"
)
elif edit_str_end == "":
st = "<d type='" + edit_type + "' edit=''>" + edit_str_start + "</d>"
else:
st = (
"<c type='" + edit_type + "' edit='" + edit_str_end + "'>" + edit_str_start + "</c>"
)
orig_tokens[edit_spos] = st
for i in sorted(ignore_indexes, reverse=True):
del orig_tokens[i]
return " ".join(orig_tokens)
def get_edits(annotator, orig, cor):
orig = annotator.parse(orig)
cor = annotator.parse(cor)
alignment = annotator.align(orig, cor)
edits = annotator.merge(alignment)
if len(edits) == 0:
return []
edit_annotations = []
for e in edits:
e = annotator.classify(e)
edit_annotations.append(
(e.type[2:], e.o_str, e.o_start, e.o_end, e.c_str, e.c_start, e.c_end)
)
return edit_annotations or []
|