|
| 1 | +"""MkDocs plugin that generates a Markdown file at the end of the build.""" |
| 2 | + |
| 3 | +from __future__ import annotations |
| 4 | + |
| 5 | +import fnmatch |
| 6 | +from collections import defaultdict |
| 7 | +from itertools import chain |
| 8 | +from pathlib import Path |
| 9 | +from typing import TYPE_CHECKING |
| 10 | + |
| 11 | +import mdformat |
| 12 | +from bs4 import BeautifulSoup as Soup |
| 13 | +from bs4 import Tag |
| 14 | +from markdownify import ATX, MarkdownConverter |
| 15 | +from mkdocs.config.defaults import MkDocsConfig |
| 16 | +from mkdocs.exceptions import PluginError |
| 17 | +from mkdocs.plugins import BasePlugin |
| 18 | + |
| 19 | +from mkdocs_llmstxt.config import PluginConfig |
| 20 | +from mkdocs_llmstxt.logger import get_logger |
| 21 | +from mkdocs_llmstxt.preprocess import autoclean, preprocess |
| 22 | + |
| 23 | +if TYPE_CHECKING: |
| 24 | + from typing import Any |
| 25 | + |
| 26 | + from mkdocs.config.defaults import MkDocsConfig |
| 27 | + from mkdocs.structure.files import Files |
| 28 | + from mkdocs.structure.pages import Page |
| 29 | + |
| 30 | + |
| 31 | +logger = get_logger(__name__) |
| 32 | + |
| 33 | + |
| 34 | +class MkdocsLLMsTxtPlugin(BasePlugin[PluginConfig]): |
| 35 | + """The MkDocs plugin to generate an `llms.txt` file. |
| 36 | +
|
| 37 | + This plugin defines the following event hooks: |
| 38 | +
|
| 39 | + - `on_page_content` |
| 40 | + - `on_post_build` |
| 41 | +
|
| 42 | + Check the [Developing Plugins](https://www.mkdocs.org/user-guide/plugins/#developing-plugins) page of `mkdocs` |
| 43 | + for more information about its plugin system. |
| 44 | + """ |
| 45 | + |
| 46 | + mkdocs_config: MkDocsConfig |
| 47 | + |
| 48 | + def __init__(self) -> None: # noqa: D107 |
| 49 | + self.html_pages: dict[str, dict[str, str]] = defaultdict(dict) |
| 50 | + |
| 51 | + def _expand_inputs(self, inputs: list[str], page_uris: list[str]) -> list[str]: |
| 52 | + expanded: list[str] = [] |
| 53 | + for input_file in inputs: |
| 54 | + if "*" in input_file: |
| 55 | + expanded.extend(fnmatch.filter(page_uris, input_file)) |
| 56 | + else: |
| 57 | + expanded.append(input_file) |
| 58 | + return expanded |
| 59 | + |
| 60 | + def on_config(self, config: MkDocsConfig) -> MkDocsConfig | None: |
| 61 | + """Save the global MkDocs configuration. |
| 62 | +
|
| 63 | + Hook for the [`on_config` event](https://www.mkdocs.org/user-guide/plugins/#on_config). |
| 64 | + In this hook, we save the global MkDocs configuration into an instance variable, |
| 65 | + to re-use it later. |
| 66 | +
|
| 67 | + Arguments: |
| 68 | + config: The MkDocs config object. |
| 69 | +
|
| 70 | + Returns: |
| 71 | + The same, untouched config. |
| 72 | + """ |
| 73 | + self.mkdocs_config = config |
| 74 | + return config |
| 75 | + |
| 76 | + def on_files(self, files: Files, *, config: MkDocsConfig) -> Files | None: # noqa: ARG002 |
| 77 | + """Expand inputs for generated files. |
| 78 | +
|
| 79 | + Hook for the [`on_files` event](https://www.mkdocs.org/user-guide/plugins/#on_files). |
| 80 | + In this hook we expand inputs for generated file (glob patterns using `*`). |
| 81 | +
|
| 82 | + Parameters: |
| 83 | + files: The collection of MkDocs files. |
| 84 | + config: The MkDocs configuration. |
| 85 | +
|
| 86 | + Returns: |
| 87 | + Modified collection or none. |
| 88 | + """ |
| 89 | + for file in self.config.files: |
| 90 | + file["inputs"] = self._expand_inputs(file["inputs"], page_uris=list(files.src_uris.keys())) |
| 91 | + return files |
| 92 | + |
| 93 | + def on_page_content(self, html: str, *, page: Page, **kwargs: Any) -> str | None: # noqa: ARG002 |
| 94 | + """Record pages contents. |
| 95 | +
|
| 96 | + Hook for the [`on_page_content` event](https://www.mkdocs.org/user-guide/plugins/#on_page_content). |
| 97 | + In this hook we simply record the HTML of the pages into a dictionary whose keys are the pages' URIs. |
| 98 | +
|
| 99 | + Parameters: |
| 100 | + html: The rendered HTML. |
| 101 | + page: The page object. |
| 102 | + """ |
| 103 | + for file in self.config.files: |
| 104 | + if page.file.src_uri in file["inputs"]: |
| 105 | + logger.debug(f"Adding page {page.file.src_uri} to page {file['output']}") |
| 106 | + self.html_pages[file["output"]][page.file.src_uri] = html |
| 107 | + return html |
| 108 | + |
| 109 | + def on_post_build(self, config: MkDocsConfig, **kwargs: Any) -> None: # noqa: ARG002 |
| 110 | + """Combine all recorded pages contents and convert it to a Markdown file with BeautifulSoup and Markdownify. |
| 111 | +
|
| 112 | + Hook for the [`on_post_build` event](https://www.mkdocs.org/user-guide/plugins/#on_post_build). |
| 113 | + In this hook we concatenate all previously recorded HTML, and convert it to Markdown using Markdownify. |
| 114 | +
|
| 115 | + Parameters: |
| 116 | + config: MkDocs configuration. |
| 117 | + """ |
| 118 | + |
| 119 | + def language_callback(tag: Tag) -> str: |
| 120 | + for css_class in chain(tag.get("class", ()), tag.parent.get("class", ())): |
| 121 | + if css_class.startswith("language-"): |
| 122 | + return css_class[9:] |
| 123 | + return "" |
| 124 | + |
| 125 | + converter = MarkdownConverter( |
| 126 | + bullets="-", |
| 127 | + code_language_callback=language_callback, |
| 128 | + escape_underscores=False, |
| 129 | + heading_style=ATX, |
| 130 | + ) |
| 131 | + |
| 132 | + for file in self.config.files: |
| 133 | + try: |
| 134 | + html = "\n\n".join(self.html_pages[file["output"]][input_page] for input_page in file["inputs"]) |
| 135 | + except KeyError as error: |
| 136 | + raise PluginError(str(error)) from error |
| 137 | + |
| 138 | + soup = Soup(html, "html.parser") |
| 139 | + if self.config.autoclean: |
| 140 | + autoclean(soup) |
| 141 | + if self.config.preprocess: |
| 142 | + preprocess(soup, self.config.preprocess, file["output"]) |
| 143 | + |
| 144 | + output_file = Path(config.site_dir).joinpath(file["output"]) |
| 145 | + output_file.parent.mkdir(parents=True, exist_ok=True) |
| 146 | + markdown = mdformat.text(converter.convert_soup(soup), options={"wrap": "no"}) |
| 147 | + output_file.write_text(markdown, encoding="utf8") |
| 148 | + |
| 149 | + logger.info(f"Generated file /{file['output']}") |
0 commit comments