vivtsai commited on
Commit
bdfae41
1 Parent(s): dd7b8fe

Remove obsolete file

Browse files

This script has been replaced by `json_to_md.py`, which converts JSON output from `reformat_json.py` directly to the Markdown format specified by Data Cards Labs.

Files changed (1) hide show
  1. formatting/format_as_md.py +0 -78
formatting/format_as_md.py DELETED
@@ -1,78 +0,0 @@
1
- import argparse
2
- import json
3
- import pathlib
4
- import os
5
-
6
- parser = argparse.ArgumentParser(
7
- description="Format the output of the data card tool as .md for the hub."
8
- )
9
- parser.add_argument("--input_path", "-i", type=pathlib.Path, required=True)
10
- parser.add_argument("--output_path", "-o", type=pathlib.Path, required=True)
11
- args = parser.parse_args()
12
-
13
- def read_json_file(json_path: pathlib.Path):
14
- """Load a json file and return it as object."""
15
- with open(json_path, "r") as f:
16
- data = json.load(f)
17
- return data
18
-
19
- def save_md_file(md_path: pathlib.Path, md_string: str):
20
- """Takes a string and saves it as .md file."""
21
- with open(md_path, 'w') as f:
22
- f.write(md_string)
23
-
24
- def construct_md(data_card_data: dict, text_by_key: dict):
25
- """Constructs the markdown file
26
-
27
- This function iterates through text_by_key and extracts all answers from
28
- the data_card_data object. It uses the levels of hierarchy as indicator for
29
- the heading indentation and does not change the order in which anything
30
- appears.
31
-
32
- Args:
33
- data_card_data: Output from the data card tool
34
- text_by_key: configuration defined in key_to_question.json
35
-
36
- Returns:
37
- data_card_md_string: Markdown-formatted content
38
- """
39
-
40
- data_card_md_string = ""
41
-
42
- for main_key, main_content in text_by_key.items():
43
- section_header = main_content['section-title']
44
- # Remove it so that we don't iterate over it.
45
- del main_content['section-title']
46
- # Add it to string.
47
- data_card_md_string += f"## {section_header} \n\n"
48
- # Iterate over the subsections.
49
- for second_key, second_content in main_content.items():
50
- subsection_header = second_content['section-title']
51
- # Remove it so that we don't iterate over it.
52
- del second_content['section-title']
53
- # Add it to string.
54
- data_card_md_string += f"### {subsection_header} \n\n"
55
- # Finally, iterate over actual questions.
56
- for question_key, question_text in second_content.items():
57
- # Now grab the answer text.
58
- reply = data_card_data[main_key][second_key].get(question_key, "N/A")
59
- if reply and reply != "N/A":
60
- data_card_md_string += f"#### {question_text} \n\n"
61
- # Special parsing options here.
62
- if question_key == question_key == "paper-bibtext":
63
- data_card_md_string += f"```\n{reply}\n``` \n\n"
64
- elif question_key == "structure-example":
65
- data_card_md_string += f"```json\n{reply}\n``` \n\n"
66
- elif isinstance(reply, list):
67
- data_card_md_string += f"{' '.join(reply)} \n\n"
68
- else:
69
- data_card_md_string += f"{reply} \n\n"
70
-
71
- return data_card_md_string
72
-
73
-
74
- if __name__ == "__main__":
75
- data_card_data = read_json_file(args.input_path)
76
- text_by_key = read_json_file(os.path.join(os.path.dirname(__file__), "key_to_question.json"))
77
- data_card_md_string = construct_md(data_card_data, text_by_key)
78
- save_md_file(args.output_path, data_card_md_string)