Dissect the tracking URLs

A marketing analyst in Toronto receives hundreds of campaign URLs such as HTTPS://Shop.Example.COM:8080/Men/Shoes?color=Red&size=42&color=blue#reviews and needs them split into readable parts.

Everything after the first # is a fragment and is ignored (even a ? inside it).
The host is the text between :// and the first / or ?, in lower case, with any :port removed.
The path starts at the first / after the host and ends before the ?; keep its case; if there is none the path is /.
The query is the text after the first ?. It is split on &; empty parts are skipped; each part is split at its first = into key and value (no = means an empty value). Keys and values are percent-decoded and + becomes a space (urllib.parse.unquote_plus does exactly this).

Task: line 1 is N; then N URLs, each containing ://. For each URL print host=..., then path=..., then one line per distinct key, sorted by key, as key=v1|v2 where the values of repeated keys are joined with | in order of appearance. After each URL print a line with ---.

Sample tests

Input

2
HTTPS://Shop.Example.COM:8080/Men/Shoes?color=Red&size=42&color=blue#reviews
http://example.org

Expected output

host=shop.example.com
path=/Men/Shoes
color=Red|blue
size=42
---
host=example.org
path=/
---

Input

1
https://news.site.pk/search?q=data+analyst%20jobs&city=Lahore&q=sql

Expected output

host=news.site.pk
path=/search
city=Lahore
q=data analyst jobs|sql
---

Your answer is also checked against 2 hidden tests.