Relparse

Multipurpose parsing CLI (optimized for Bun users only).

Features:

• HTML: extract title, meta, OpenGraph, links, canonical, JSON-LD (streamed via HTMLRewriter)
• RSS/Atom: parse feed items (title/link/id/date)
• Sitemap: parse sitemap.xml and sitemap index files
• File: parse local files by extension (json|yaml|md|xml|html)
• Output: JSON, YAML, CSV
• Caching: optional response caching for HTML requests
• Crawling: site-agnostic via CLI flags (multi-page category → target emails)

Installation

bun add -D @reliverse/relparse
# or globally: bun add -g @reliverse/relparse

Usage

bun relparse <command> [options]
# or globally: relparse <command> [options]

Commands

• help: show help
• version: show CLI version
• crawl : crawl category pages and extract target emails
  • flags: --pages 1-15|1,3,10 (default 1-15) --per-page (default 25) --ua --timeout --retries --format json|yaml|csv --out --stdout --selector --rel-prefix --abs-base --host --path-contains --jsonld-types <t1,t2,...> (empty or omit = all types)
• html : parse an HTML page
  • flags: --ua --timeout --retries --cache --format json|yaml|csv --out --stdout
• rss : parse an RSS/Atom feed
  • flags: --ua --timeout --retries --format json|yaml|csv --out --stdout
• sitemap : parse sitemap.xml or a sitemap index
  • flags: --ua --timeout --retries --format json|yaml|csv --out --stdout
• file : parse local files by extension
  • flags: --format json|yaml|csv --out --stdout
• config : run tasks from a YAML/JSON config
• cache clear: clear the response/result cache

Examples:

# Category pages → emails.csv (pass selectors & URL rules via flags)
bun relparse crawl https://example.com/category/widgets \
  --selector a.item-link --rel-prefix /widgets/ --abs-base https://example.com \
  --host example.com --path-contains /widgets/ \
  --pages 1-15 --per-page 25 --format csv --out emails.csv

# HTML → YAML to stdout
bun relparse html https://example.com --format yaml --stdout

# HTML → JSON to file with cache enabled
bun relparse html https://example.com --cache --out results/example.json

# RSS → JSON to stdout
bun relparse rss https://example.com/feed.xml --stdout

# Sitemap → JSON
bun relparse sitemap https://example.com/sitemap.xml --format json

# Local files (JSON/YAML/MD/XML/HTML) → JSON
bun relparse file "content/**/*.{json,yaml,yml,md,html,xml}" --format json

# Clear cache
bun relparse cache clear

Crawl flags

The crawl command is site-agnostic and configured entirely via flags:

• Required for category crawl:

  • --selector <css>: CSS selector for links on category pages
  • --rel-prefix <prefix>: accept only relative hrefs with this prefix
  • --abs-base <url>: absolute base to resolve relative target URLs

• Optional, but recommended for direct target detection (when the input URL is a target page):

  • --host <hostname>: hostname to match
  • --path-contains <substr>: substring to match in the pathname

• Email/entity extraction from JSON-LD:

  • --jsonld-types <t1,t2,...>: comma-separated @type allowlist (case-insensitive). If omitted or empty, all detected types are allowed.

Examples:

# Direct target URL (no category crawl):
bun relparse crawl https://example.com/widgets/some-item \
  --host example.com --path-contains /widgets/

# Limit fields in output (CSV):
bun relparse crawl https://example.com/category/widgets \
  --selector a.item-link --rel-prefix /widgets/ --abs-base https://example.com \
  --get name,email,url --format csv --out emails.csv