Skip to content
FrostyBee Docs LogoFrostyBee Docs

Known limitations

Each entry names something JSON Digger does not do and the way around it when there is one. For the numeric caps, such as the 4,096 children a tree node shows, see Limits.

Opening files

Section titled “Opening files”
  • UTF-16 files are not decoded. Convert the file to UTF-8 first; see Encodings and byte order marks.
  • Changes on some network drives go unnoticed. Open the file again to reload it; see Network drives and locked files.
  • A .gz or .zst file needs free disk space for its decompressed copy. When the disk is full, the open fails with a message that says so. Free space, or decompress the file elsewhere and open the result.
  • A malformed .json file opens as far as its first error. The status bar shows the error; the Repair offer appears only when the open itself fails. See Open malformed JSON.

Views and tabs

Section titled “Views and tabs”
  • Tabs cannot be reordered by dragging and have no context menu. Switch tabs with Window → Next Tab and close a tab with its close button.
  • The tree and table state is not restored. After a restart the tree starts collapsed and the table’s sort, filters, hidden columns and pinned columns start over; see Session restore.
  • NDJSON records past the first 4,096 have no tree row. A search or query result in such a record opens in the NDJSON view at its line; see Very large objects and arrays.
  • The NDJSON view does not wrap long lines. Open the record in the tree, or use the Text view for the raw line.
  • Search results are not kept with a workspace. Run the search again after a restore; saved searches keep their text; see Search history and saved searches.

Search and queries

Section titled “Search and queries”
  • A live stream tab cannot be searched or queried. Freeze the stream into a document first; see What works on a live stream.
  • JSONPath and filter queries on an NDJSON file run on one record. To query every record, use jq or SQL.
  • A jq query that runs while a large NDJSON file is still indexing covers the records indexed so far. Wait for the status bar’s indexing indicator to finish, then run the query again.
  • Key frequency reads the whole file. On a multi-gigabyte file it takes a while; the Statistics panel reads a sample instead.
  • Case folding differs by search mode. Literal and Exact mode ignore case for ASCII letters only; Regex and Fuzzy mode fold Unicode letters too, and Fuzzy mode always ignores case.
  • Fuzzy matches on a minified file highlight a wide range. The results panel shows the region that holds the match rather than the exact bytes. Use Literal or Regex mode for an exact span.
  • The Analyze view has no scope. A pipeline runs on the whole document; for one array, run a jq query on the selected node instead; see jq.

Editing

Section titled “Editing”
  • Undo keeps the last edit only. Each edit replaces the one before it in the undo history; see How editing works.
  • Members cannot be reordered, moved, or edited in batches. Rename, insert and delete act on one member at a time; a multi-selection deletes nothing. To reorder, use Sort an array or apply a JSON Patch.
  • NDJSON and plain text files are read-only. Edit them in a text editor; see What you can edit.
  • Files over 2 GiB cannot be edited. Use a text editor or a command-line tool for them.
  • Free space is checked on the file’s own folder. The edit journal and the index cache live in the data folders; when those are on a fuller volume, a save can still fail with a disk-space message.

jdig

Section titled “jdig”