|
| 1 | +import pandas |
| 2 | +import glob |
| 3 | +from collections import defaultdict |
| 4 | + |
| 5 | +""" |
| 6 | +Script used to generate readme programmatically works by taking the links of all the notebooks |
| 7 | +then dividing them to different tables based on directory name. Pandas is used to make the tables. Using inline HTML to support our doc page. |
| 8 | +""" |
| 9 | + |
| 10 | +SDK_EXAMPLE_HEADER = """ |
| 11 | +# Labelbox SDK Examples\n |
| 12 | +- Learn how to use the SDK by following along\n |
| 13 | +- Run in google colab, view the notebooks on github, or clone the repo and run locally\n |
| 14 | +""" |
| 15 | + |
| 16 | +COLAB_TEMPLATE = "https://colab.research.google.com/github/Labelbox/labelbox-python/blob/develop/examples/{filename}" |
| 17 | +GITHUB_TEMPLATE = ( |
| 18 | + "https://github.com/Labelbox/labelbox-python/tree/develop/examples/{filename}" |
| 19 | +) |
| 20 | + |
| 21 | + |
| 22 | +def create_header(link: str) -> str: |
| 23 | + """Creates headers of tables with h2 tags to support readme docs |
| 24 | +
|
| 25 | + Args: |
| 26 | + link (str): file path |
| 27 | +
|
| 28 | + Returns: |
| 29 | + str: formatted file path for header |
| 30 | + """ |
| 31 | + # Splits up link uses directory name |
| 32 | + split_link = link.split("/")[1].replace("_", " ").split(" ") |
| 33 | + header = [] |
| 34 | + |
| 35 | + # Capitalize first letter of each word |
| 36 | + for word in split_link: |
| 37 | + header.append(word.capitalize()) |
| 38 | + return f"<h2>{' '.join(header)}</h2>" |
| 39 | + |
| 40 | + |
| 41 | +def create_title(link: str) -> str: |
| 42 | + """Create notebook titles will be name of notebooks with _ replaced with spaces and file extension removed |
| 43 | +
|
| 44 | + Args: |
| 45 | + link (str): file path |
| 46 | +
|
| 47 | + Returns: |
| 48 | + str: formatted file path for notebook title |
| 49 | + """ |
| 50 | + split_link = link.split(".")[-2].split("/")[-1].replace("_", " ").split(" ") |
| 51 | + title = [] |
| 52 | + |
| 53 | + # List to lower case certain words and list to keep certain acronyms capitalized |
| 54 | + lower_case_words = ["to"] |
| 55 | + acronyms = ["html", "pdf", "llm", "dicom", "sam"] |
| 56 | + |
| 57 | + for word in split_link: |
| 58 | + if word.lower() in acronyms: |
| 59 | + title.append(word.upper()) |
| 60 | + elif word.lower() in lower_case_words: |
| 61 | + title.append(word.lower()) |
| 62 | + else: |
| 63 | + title.append(word.capitalize()) |
| 64 | + return " ".join(title).split(".")[0] |
| 65 | + |
| 66 | + |
| 67 | +def make_link(link: str, photo: str, link_type: str) -> str: |
| 68 | + """Creates the links for the notebooks as an anchor tag |
| 69 | +
|
| 70 | + Args: |
| 71 | + link (str): file path |
| 72 | + photo (str): photo link |
| 73 | + link_type (str): type of link (github, google colab) |
| 74 | +
|
| 75 | + Returns: |
| 76 | + str: anchor tag with image |
| 77 | + """ |
| 78 | + return f'<a href="{link}" target="_blank"><img src="{photo}" alt="Open In {link_type}" onclick="(function prevent(e){{e.preventDefault()}}()"></a>' |
| 79 | + |
| 80 | + |
| 81 | +def make_links_dict(links: str) -> defaultdict[list]: |
| 82 | + """Creates dictionary needed for pandas to generate the table takes all the links and makes each directory its own table |
| 83 | +
|
| 84 | + Args: |
| 85 | + links (list[str]): list of links to notebooks from glob |
| 86 | +
|
| 87 | + Returns: |
| 88 | + defaultdict[list]: returns dict that is in pandas dataFrame format |
| 89 | + """ |
| 90 | + link_dict = defaultdict(list) |
| 91 | + for link in links: |
| 92 | + split_link = link.split("/")[1] |
| 93 | + link_dict[split_link].append(link) |
| 94 | + return link_dict |
| 95 | + |
| 96 | + |
| 97 | +def make_table(base: str = "") -> str: |
| 98 | + """main function to make table |
| 99 | +
|
| 100 | + Args: |
| 101 | + base (str): Header of file generated. Defaults to an empty string. |
| 102 | +
|
| 103 | + Returns: |
| 104 | + str: markdown string file |
| 105 | + """ |
| 106 | + link_dict = make_links_dict(glob.glob("**/examples/**/*.ipynb", recursive=True)) |
| 107 | + generated_markdown = base |
| 108 | + for link_list in link_dict.values(): |
| 109 | + pandas_dict = {"Notebook": [], "Github": [], "Google Colab": []} |
| 110 | + generated_markdown += f"{create_header(link_list[0])}\n\n" |
| 111 | + for link in link_list: |
| 112 | + pandas_dict["Notebook"].append(create_title(link)) |
| 113 | + pandas_dict["Github"].append( |
| 114 | + make_link( |
| 115 | + GITHUB_TEMPLATE.format(filename="/".join(link.split("/")[1:])), |
| 116 | + "https://img.shields.io/badge/GitHub-100000?logo=github&logoColor=white", |
| 117 | + "Github", |
| 118 | + ) |
| 119 | + ) |
| 120 | + pandas_dict["Google Colab"].append( |
| 121 | + make_link( |
| 122 | + COLAB_TEMPLATE.format(filename="/".join(link.split("/")[1:])), |
| 123 | + "https://colab.research.google.com/assets/colab-badge.svg", |
| 124 | + "Colab", |
| 125 | + ) |
| 126 | + ) |
| 127 | + df = pandas.DataFrame(pandas_dict) |
| 128 | + generated_markdown += f"{df.to_html(col_space={'Notebook':400}, index=False, escape=False, justify='left')}\n\n" |
| 129 | + return generated_markdown |
| 130 | + |
| 131 | + |
| 132 | +if __name__ == "__main__": |
| 133 | + with open("./examples/README.md", "w") as readme: |
| 134 | + readme.write(f"{make_table(SDK_EXAMPLE_HEADER).rstrip()}\n") |
0 commit comments