URL Analyzer

URL Analyzer reads each line with the browser's own URL parser and reports what it found: the scheme, the host and any non-default port, the path, how many segments deep it goes, the final slug, the query, the fragment and the total character count. Work is done a batch at a time, one address per line; blank lines and lines beginning with # are skipped so you can label groups, and each URL gets its own breakdown and its own list of findings.

Those findings come in three tiers. Errors are the things that are wrong rather than untidy: http:// instead of https, a scheme that is neither, a session parameter such as sessionid or phpsessid that mints a fresh address on every crawl, and any URL past 2,048 characters. Warnings cover the conventions — an address longer than about a hundred characters, uppercase letters in the path or in the host as you typed it, underscores where hyphens belong, an encoded space or a plus sign in the path, a double slash, a #fragment, a bare IP address for a host, an index.php ending, a path more than four levels deep, more than four query parameters, a tracking tag from the utm_, gclid and fbclid family, a word repeated inside the slug, and a last segment that is an opaque ID. Notes are neutral observations rather than complaints: a trailing slash present or absent, a date in the path, a .html or .php extension, a non-default port, non-ASCII characters in the path, an internationalised domain, two or more English stop words in the slug, a slug running past six words, and parameters carrying empty values. Only errors and warnings move the verdict; the notes are printed and never counted.

Nothing is fetched. The analysis is of the address as a string, so this cannot tell you whether the page answers at all, what status it returns, where it redirects, which canonical it declares or whether it is indexed — Redirect Checker, Canonical Checker and HTTP Header Inspector each make the request that this tool deliberately does not. Four behaviours matter before you read a result as a verdict. A line typed without a scheme is silently upgraded to https:// before parsing, so example.com/page can never be flagged for http. The tracking list is a fixed set of about thirty known parameter names, so an in-house tracking tag is counted as ordinary content. The two case checks are blunt string tests rather than real case analysis: the path check looks for any A–Z in the parsed path, and because an accented character is stored percent-encoded with capital hex digits — /café is really /caf%C3%A9 — an all-lowercase accented address collects an uppercase warning it does not deserve. The host check reads whatever you typed before the first slash, so a capital inside a username, a password or the query of a line that has no path is reported as a capital in the host name. The verdict is arithmetic — any error makes it Needs fixing, more than two warnings Fair, one or two Good, none Clean — and not a prediction about ranking.

All of it runs in this browser tab. No URL is transmitted and none is requested, which is worth more here than it first looks: the addresses pasted into a URL checker are usually copied out of analytics exports, spreadsheets and server logs, and they routinely carry session identifiers, reset tokens, internal hostnames and staging addresses that were never meant to travel. Nothing is logged or stored, and the tool goes on working with the network disconnected.

Runs on your device. Your file is processed directly in this browser.