import path from "path";
import type { DocChunk, Frontmatter, ParsedDoc, RawDoc } from "./types.js";
export function parseDoc(raw: RawDoc): ParsedDoc {
const { frontmatter, body } = extractFrontmatter(raw.content);
const cleaned = stripMdx(body);
const name = pageNameFromPath(raw.path);
const title = frontmatter.title?.trim() || titleFromName(name);
const chunks = chunkByH2(cleaned, name, raw.url, title);
return {
name,
title,
description: frontmatter.description?.trim(),
path: raw.path,
url: raw.url,
content: cleaned.trim(),
chunks,
};
}
function extractFrontmatter(content: string): {
frontmatter: Frontmatter;
body: string;
} {
const frontmatter: Frontmatter = {};
if (!content.startsWith("---")) {
return { frontmatter, body: content };
}
const endIndex = content.indexOf("\n---", 3);
if (endIndex === -1) {
return { frontmatter, body: content };
}
const raw = content.slice(3, endIndex).trim();
let body = content.slice(endIndex + 4);
if (body.startsWith("\n")) {
body = body.slice(1);
}
for (const line of raw.split(/\r?\n/)) {
const match = line.match(/^(\w+):\s*(.+)$/);
if (!match) {
continue;
}
const key = match[1];
const value = stripQuotes(match[2].trim());
if (key === "sidebar_position") {
const parsed = Number.parseInt(value, 10);
if (!Number.isNaN(parsed)) {
frontmatter.sidebar_position = parsed;
}
} else if (key === "title") {
frontmatter.title = value;
} else if (key === "description") {
frontmatter.description = value;
}
}
return { frontmatter, body };
}
function stripMdx(content: string): string {
const codeBlocks: string[] = [];
const placeholder = "__NIIVUE_CODE_BLOCK_";
let output = content.replace(/```[\s\S]*?```/g, (match) => {
const index = codeBlocks.length;
codeBlocks.push(match);
return `${placeholder}${index}__`;
});
output = output.replace(/^\s*import .*$/gm, "");
output = output.replace(/<([A-Z][\w]*)\b[^>]*>[\s\S]*?<\/\1>/g, "");
output = output.replace(/<([A-Z][\w]*)\b[^>]*\/>/g, "");
output = output.replace(new RegExp(`${placeholder}(\\d+)__`, "g"), (_match, index) => {
const idx = Number(index);
return codeBlocks[idx] ?? "";
});
output = output.replace(/\n{3,}/g, "\n\n");
return output;
}
function chunkByH2(
content: string,
page: string,
url: string,
fallbackTitle: string
): DocChunk[] {
const lines = content.split(/\r?\n/);
const chunks: DocChunk[] = [];
let currentTitle = fallbackTitle;
let currentLines: string[] = [];
let index = 0;
const pushChunk = () => {
const chunkContent = currentLines.join("\n").trim();
if (!chunkContent) {
return;
}
chunks.push({
id: `${page}:${index}`,
title: currentTitle,
content: chunkContent,
page,
url,
});
index += 1;
};
for (const line of lines) {
if (line.startsWith("## ")) {
pushChunk();
currentTitle = line.replace(/^##\s+/, "").trim() || fallbackTitle;
currentLines = [];
continue;
}
currentLines.push(line);
}
pushChunk();
return chunks;
}
function pageNameFromPath(docPath: string): string {
const base = path.basename(docPath);
return base.replace(/\.(md|mdx)$/i, "");
}
function titleFromName(name: string): string {
return name.replace(/[-_]/g, " ").replace(/\b\w/g, (char) => char.toUpperCase());
}
function stripQuotes(value: string): string {
if (
(value.startsWith("\"") && value.endsWith("\"")) ||
(value.startsWith("'") && value.endsWith("'"))
) {
return value.slice(1, -1);
}
return value;
}
