# html/select Query a parsed document with a CSS selector list and get the matching elements as Deft trees. `.class`, `#id`, attribute selectors, combinators (`div > p`, `a b`), pseudo-classes (`:first-child`, `:has()`, …) — the full selector language. ## Syntax 
[html/select $src|$doc $selector]
## Parameters - `$src` — HTML source text (string) or a `%HtmlDocument` handle from `html/open` - `$selector` — CSS selector list; comma separates alternatives ## Examples 
set hits [html/select "<div><p class=x>1</p><p id=y>2</p><p>3</p></div>" "p.x, p#y"]
[len $hits]                          # => 2
[nth [nth $hits 0]~children 0]       # => "1"

# Combinators
set kids [html/select "<div><section><p>a</p></section><p>b</p></div>" "div > p"]
[len $kids]                          # => 1

# Attribute selectors
[html/select "<a href='x'>l</a><a>plain</a>" "a[href]"]   # 1 hit

# Repeat queries against one parse
set d [html/open $page]
each [html/select $d "a[href]"] {|a| echo $a~attrs~href }
[html/close $d]
## Returns A vector of element maps in the `html/parse` tree shape (`%{ :tag :attrs :children }`), in document order. A node matching several comma-list selectors appears once (matches are deduped). No match is not an error — it returns an empty vector `@{}`. An unparseable selector returns `%Err{ :message "html/select: invalid` CSS selector" }`; so do parse failures when a string input is parsed.` ## See Also `html/open`, `html/parse`, `html/text`, `re/match`
← html/parse
html/serialize →