Query String Parser & Builder

Inspect and rebuild parameters

1. Choose Query string or HTTP(S) URL, paste the source and select Parse. A blank Query string starts an empty builder. Changing the source or mode discards row edits and requires another Parse.
2. Edit decoded names and values, add a row, or use Move up, Move down and Remove. An empty name or value is data; a completely empty row still represents one parameter.
3. Review the normalized output and copy the displayed query or URL. If clipboard access fails, use the readonly fallback. Invalid fields have no current copyable output.

Understand the parameter list

A query is an ordered list, so label=morning&view=list&label=evening has three entries. The two label rows remain separate, including the view row between them. Brackets and dots in a name do not create nested objects here.

Under form encoding, a literal plus in the source means a space. A percent-encoded %2B means a plus character. Values are decoded once: %252B becomes the text %2B, not a plus or a space.

The normalized query uses a plus for a space, uppercase percent escapes and an equals sign for every row. Empty separator segments disappear. Those spelling changes preserve admitted ordered text pairs, not the original bytes or a signed link’s validity.

The source field stays as the source for parsing; it is not rewritten while you edit rows. The rows become the current editing session and drive the result. Start a new parse to replace that session.

Examples of meaning and spelling

The example query contains label twice, view once, q with the text tea + café, an empty value, and a blank name with value note. It produces six independent rows; both the empty value and the value attached to the blank name remain.

For q=tea%20%2B%20caf%C3%A9, the decoded value is tea + café and the normalized output is q=tea+%2B+caf%C3%A9. The space spelling changes while the decoded value stays the same.

In Query string mode, x=1#frag contains a literal hash in the value and becomes x=1%23frag. In HTTP(S) URL mode, a fragment after # stays outside the parameter list, including question marks inside it.

Strict text and normalization limits

This editor rejects malformed percent escapes, invalid UTF-8, lone UTF-16 surrogates, and decoded U+0000–U+001F or U+007F controls, including NUL, tab and line breaks. They are not silently repaired or removed. Other scalar Unicode text, including a genuine replacement character or byte-order-mark character, remains data.

The source limit is 65536 UTF-8 bytes and the row limit is 200. Each decoded name or value is at most 4096 UTF-8 bytes, the combined decoded text at most 65536, and each output at most 262144. An excessive result fails as a whole rather than copying a shortened prefix.

URL mode accepts literal absolute HTTP(S) syntax without credentials, raw ASCII whitespace, controls or backslashes. Native URL serialization can normalize the host, default port and path. Strict text checks cover the query fields, not the whole address, its safety, server interpretation or reachability.

Output is normalized, not byte-for-byte reconstruction. Empty query markers may disappear, bare names gain an equals sign and escaping can change. Rewriting even an otherwise unchanged signed URL can invalidate its signature. No URL is requested or opened by this tool.

Edit parameter text locally

The tool adds no upload, request to the entered address, saved draft history, parameter-bearing link or telemetry carrying your data. Copy is explicit. Existing site services, extensions and copies you keep are outside a universal privacy or erasure promise.