Spaces:
Paused
Paused
Commit ·
1f7ca14
1
Parent(s): fe9092a
Add card_processing.py for markdown parsing and text loading
Browse files- card_processing.py +72 -0
card_processing.py
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import re
|
| 2 |
+
|
| 3 |
+
from huggingface_hub import DatasetCard
|
| 4 |
+
|
| 5 |
+
|
| 6 |
+
def parse_markdown(markdown_text):
|
| 7 |
+
lines = markdown_text.split("\n")
|
| 8 |
+
parsed_lines = []
|
| 9 |
+
skip_section = False
|
| 10 |
+
empty_section = True
|
| 11 |
+
table_of_contents = False
|
| 12 |
+
more_info_pattern = re.compile(r"\[More Information Needed\]\(https?://\S+\)")
|
| 13 |
+
html_comment_pattern = re.compile(r"<!--.*?-->")
|
| 14 |
+
|
| 15 |
+
for line in lines:
|
| 16 |
+
if "Table of Contents" in line:
|
| 17 |
+
table_of_contents = True
|
| 18 |
+
continue
|
| 19 |
+
if table_of_contents:
|
| 20 |
+
if line.startswith("#"):
|
| 21 |
+
table_of_contents = False
|
| 22 |
+
else:
|
| 23 |
+
continue
|
| 24 |
+
|
| 25 |
+
if line.startswith("#"):
|
| 26 |
+
if skip_section or empty_section:
|
| 27 |
+
continue
|
| 28 |
+
empty_section = True
|
| 29 |
+
|
| 30 |
+
if skip_section:
|
| 31 |
+
if line.startswith("#"):
|
| 32 |
+
skip_section = False
|
| 33 |
+
|
| 34 |
+
else:
|
| 35 |
+
continue
|
| 36 |
+
if more_info_pattern.match(line.strip()):
|
| 37 |
+
skip_section = True
|
| 38 |
+
empty_section = True
|
| 39 |
+
continue
|
| 40 |
+
|
| 41 |
+
if html_comment_pattern.match(line.strip()):
|
| 42 |
+
continue
|
| 43 |
+
|
| 44 |
+
if line.strip():
|
| 45 |
+
empty_section = False
|
| 46 |
+
parsed_lines.append(line)
|
| 47 |
+
|
| 48 |
+
if skip_section or empty_section:
|
| 49 |
+
while parsed_lines and parsed_lines[-1].startswith("#"):
|
| 50 |
+
parsed_lines.pop()
|
| 51 |
+
|
| 52 |
+
return "\n".join(parsed_lines)
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
def is_empty_template(text):
|
| 56 |
+
# Define the placeholder phrases
|
| 57 |
+
placeholders = [r"\[More Information Needed\]", r"\[optional\]"]
|
| 58 |
+
# Remove the placeholder phrases from the text
|
| 59 |
+
for placeholder in placeholders:
|
| 60 |
+
text = re.sub(placeholder, "", text)
|
| 61 |
+
# Remove whitespace and newline characters
|
| 62 |
+
text = text.strip()
|
| 63 |
+
# Check if the remaining text is empty
|
| 64 |
+
return not text
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
def try_load_text(row):
|
| 68 |
+
try:
|
| 69 |
+
return DatasetCard(row["card"]).text
|
| 70 |
+
except Exception as e:
|
| 71 |
+
print(e)
|
| 72 |
+
return None
|