Docs
Searching inside large JSON documents
Search runs against the raw source once and maps each hit back onto a node, which keeps its cost independent of how many nodes the document has.
Scope
A query can be matched against keys, values, or both. Scoping to keys is how you answer "does this payload contain a field called token anywhere"; scoping to values is how you find the record you are actually looking for.
Matching options
| Option | Behaviour |
|---|---|
| Case sensitive | Off by default. Turn on when case is the distinguishing feature. |
| Whole word | Wraps the pattern in word boundaries, so id stops matching uuid and video. |
| Regex | Treats the query as a JavaScript regular expression rather than literal text. |
Regex and whole word combine: the pattern is wrapped in \b(?:...)\b. An invalid expression is reported inline instead of silently returning nothing.
Useful patterns
| Pattern | Finds |
|---|---|
"(access_)?token" | Credential-shaped keys before you attach a payload to a ticket. |
\\d{4}-\\d{2}-\\d{2}T | ISO timestamps, to see which fields carry dates. |
null | Slots that are empty in this sample — the fields most likely to be mistyped by a generator. |
^\\s*$ | Empty strings hiding as populated fields. |
Limits
Search stops collecting after 200,000 matches and the status bar reports that results were truncated. That ceiling exists so a one-character query on a large file returns in bounded time rather than trying to materialise a match per node. Narrow the query or scope it to keys.
Frequently asked questions
Does search cover collapsed parts of the tree?
Yes. Search runs over the whole document, not the visible rows, and the tree expands to reveal each match as you step through results.
Why was my search truncated?
It produced more than 200,000 matches. Add more of the string, enable whole word, or restrict the scope to keys or values.
Last updated .