@analogjs/vite-plugin-nitro/src/lib/utils/get-content-files.js

A Vite plugin for adding a nitro API server

import { readFileSync } from 'node:fs';
import { join, relative, resolve } from 'node:path';
import { normalizePath } from 'vite';
import { createRequire } from 'node:module';
import { globSync } from 'tinyglobby';
const require = createRequire(import.meta.url);
/**
 * Discovers content files with front matter and extracts metadata for prerendering.
 *
 * This function:
 * 1. Discovers all content files matching the specified glob pattern
 * 2. Reads each file and parses front matter metadata
 * 3. Extracts file name, extension, and path information
 * 4. Returns structured data for prerendering content pages
 *
 * @param workspaceRoot The workspace root directory path
 * @param rootDir The project root directory relative to workspace
 * @param glob The glob pattern to match content files (e.g., 'content/blog')
 * @returns Array of PrerenderContentFile objects with metadata and front matter
 *
 * Example usage:
 * const contentFiles = getMatchingContentFilesWithFrontMatter(
 *   '/workspace',
 *   'apps/my-app',
 *   'content/blog'
 * );
 *
 * Sample discovered file paths:
 * - /workspace/apps/my-app/content/blog/first-post.md
 * - /workspace/apps/my-app/content/blog/2024/01/hello-world.md
 * - /workspace/apps/my-app/content/blog/tech/angular-v17.mdx
 * - /workspace/apps/my-app/content/blog/about/index.md
 *
 * Sample output structure:
 * {
 *   name: 'first-post',
 *   extension: 'md',
 *   path: 'content/blog',
 *   attributes: { title: 'My First Post', date: '2024-01-01', tags: ['intro'] }
 * }
 *
 * tinyglobby vs fast-glob comparison:
 * - Both support the same glob patterns for file discovery
 * - Both are efficient for finding content files
 * - tinyglobby is now used instead of fast-glob
 * - tinyglobby provides similar functionality with smaller bundle size
 * - tinyglobby's globSync returns absolute paths when absolute: true is set
 *
 * Front matter parsing:
 * - Uses front-matter library to parse YAML/TOML front matter
 * - Extracts metadata like title, date, tags, author, etc.
 * - Supports both YAML (---) and TOML (+++) delimiters
 * - Returns structured attributes for prerendering
 *
 * File path processing:
 * - Normalizes paths for cross-platform compatibility
 * - Extracts file name without extension
 * - Determines file extension for content type handling
 * - Maintains relative path structure for routing
 */
export function getMatchingContentFilesWithFrontMatter(workspaceRoot, rootDir, glob, recursive = false) {
    // Dynamically require front-matter library
    // eslint-disable-next-line @typescript-eslint/no-var-requires
    const fm = require('front-matter');
    // Normalize the project root path for consistent path handling
    const root = normalizePath(resolve(workspaceRoot, rootDir));
    // Resolve the content directory path relative to the project root
    const resolvedDir = normalizePath(relative(root, join(root, glob)));
    // Discover all content files in the specified directory.
    // Default pattern matches only top-level files; recursive opt-in walks subdirectories.
    const pattern = recursive
        ? `${root}/${resolvedDir}/**/*`
        : `${root}/${resolvedDir}/*`;
    const contentFiles = globSync([pattern], {
        dot: true,
        absolute: true,
        onlyFiles: true,
    });
    const dirPrefix = `${root}/${resolvedDir}`;
    // Process each discovered content file to extract metadata and front matter
    const mappedFilesWithFm = contentFiles.map((f) => {
        // Read the file contents as UTF-8 text
        const fileContents = readFileSync(f, 'utf8');
        // Parse front matter from the file content
        const raw = fm(fileContents);
        // Get the relative file path by removing the root directory
        const filepath = f.replace(root, '');
        // Extract file name and extension using regex
        // Matches: /filename.ext or /filename (with optional extension)
        const match = filepath.match(/\/([^/.]+)(\.([^/.]+))?$/);
        let name = '';
        let extension = '';
        if (match) {
            name = match[1]; // File name without extension
            extension = match[3] || ''; // File extension or empty string if no extension
        }
        // Path of the file's directory relative to the configured contentDir.
        // For top-level files this is an empty string; for nested files it
        // gives transforms enough context to disambiguate identically-named
        // files (e.g. docs/a/post.md vs docs/b/post.md).
        const relativeDir = normalizePath(relative(dirPrefix, f));
        const lastSlash = relativeDir.lastIndexOf('/');
        const relativePath = lastSlash === -1 ? '' : relativeDir.slice(0, lastSlash);
        // Return structured content file data for prerendering
        return {
            name,
            extension,
            path: resolvedDir,
            attributes: raw.attributes,
            content: fileContents,
            relativePath,
        };
    });
    return mappedFilesWithFm;
}
//# sourceMappingURL=get-content-files.js.map