pup

Extract element by selector

$ cat [file.html] | pup '[selector]'

$ cat [file.html] | pup '[selector] text{}'

$ cat [file.html] | pup '[selector] attr{href}'

$ curl -s [url] | pup '[selector]'

$ cat [file.html] | pup '[selector] json{}'

pup [options] [selectors]

pup is a command-line tool for parsing and extracting data from HTML documents using CSS selectors. It reads HTML from standard input and applies selectors to filter elements, making it straightforward to extract structured data from web pages when combined with tools like curl or wget.
Selectors follow standard CSS syntax including element names, classes, IDs, attributes, and pseudo-classes. Special display functions like text{} extract text content, attr{name} retrieves attribute values, and json{} outputs matching elements as structured JSON. Multiple selectors can be chained to navigate nested document structures.
pup fills a similar role to jq but for HTML rather than JSON, providing a lightweight alternative to full-featured scraping frameworks for quick data extraction tasks in shell scripts and pipelines.

SELECTORS

CSS selectors.

Extract text content.

Extract attribute.

JSON output.

Pretty print.

Number results.

CSS selector syntax. Written in Go.

pup was created as a command-line HTML parser using CSS selectors.

jq(1), htmlq(1), xidel(1)