langchain/docs/scripts/last_updated.py

import glob
import re
import subprocess
from pathlib import Path

import nbformat

_DOCS_DIR = str(Path(__file__).parent.absolute() / ".." / "docs")


def update_contents(md, last_updated):
    last_updated_str = f"*Last updated: {last_updated}*"
    if re.search("\*Last updated: \d\d\d\d-\d\d-\d\d\*", md) is not None:
        return re.sub("\*Last updated: \d\d\d\d-\d\d-\d\d\*", last_updated_str, md)
    first_header = md.find("# ")
    if first_header >= 0:
        if "\n" in md[first_header:]:
            body = md[first_header:].replace("\n", f"\n{last_updated_str}\n", 1)
        else:
            body = md[first_header:] + f"\n{last_updated_str}"
        return md[:first_header] + body
    else:
        return f"{last_updated_str}\n\n{md}"


def last_updated(path):
    process = subprocess.Popen(
        # ["git", "log", "-n", "1", "--pretty=format:%cs", path], stdout=subprocess.PIPE
        ["stat", path], stdout=subprocess.PIPE
    )
    output = process.communicate()
    return output[0].decode()


def update_nb(path):
    nb = nbformat.read(path, as_version=4)
    first_md = None
    for cell in nb.cells:
        if cell["cell_type"] == "markdown":
            first_md = cell
            break
    if first_md is None:
        first_md = nbformat.from_dict(
            {"cell_type": "markdown", "source": "", "metadata": {}}
        )
        nb.cells = [first_md] + nb.cells

    _last_updated = last_updated(path)
    first_md.source = update_contents(first_md.source, _last_updated)
    nbformat.validate(nb)
    nbformat.write(nb, path)


def update_md(path):
    with open(path, "r") as f:
        md = f.read()
    _last_updated = last_updated(path)
    md = update_contents(md, _last_updated)
    with open(path, "w") as f:
        f.write(md)


for path in glob.glob(_DOCS_DIR + "/**/*.*", recursive=True):
    if Path(path).suffix == ".ipynb":
        update_nb(path)
    elif Path(path).suffix.startswith(".md"):
        update_md(path)