胡飞
481a856465
- README/README_CN: rewrite feature lists covering TDengine, MQTT, batch cell editing, session-log batch ops, known hosts, toolbox, ad-hoc SSH, settings redesign; add TDengine/MQTT badges and Scoop install - docs-site: add Known Hosts guide section (zh/en/zh-TW), Scoop install chapter, TDengine support notes, settings shortcut clear/disable, batch cell editing, home grid/tree + ad-hoc SSH + batch management
2.5 KiB
Browse and edit table data
Table data views provide paging, filtering, copying, and row editing. Committing additions, edits, or deletions changes the real database. Confirm keys, filters, and environment before submitting.
Browse and filter
Open Data from a table or view and use paging to control load. Search on the current page filters only rows already loaded; use SQL for server-wide filtering or aggregation. Refresh to read the latest server state.
Keep page size and maximum rows reasonable for large tables. Whether a view is editable depends on the database and view definition; use the base table or explicit SQL when it is read-only.
Edit values correctly
Edit cells directly and use the large-text editor for long strings, JSON, or multiline content. NULL and an empty string are different values, so choose the explicit NULL action rather than inferring it from a blank field.
New rows can be entered from scratch or cloned. Change primary and unique fields on a clone; generate a UUID only when the column and business rules expect one. Binary, generated, and driver-specific types may have editing limits. Refresh after commit to verify the stored representation.
Add, clone, and delete rows
Local row changes remain pending until submitted. Before deleting, confirm the count and primary keys, especially while sorted, paged, or filtered. Tables without a stable key may not be safe for single-row editing.
If the selection is wrong, discard pending changes before submission. Selecting multiple cells enables batch editing, applying one value to every selected cell at once — useful for normalizing status or category fields. Production deletions should use a controlled query, transaction, and backup rather than relying on manual recovery.
Preview SQL and submit
Open SQL Preview to inspect generated INSERT, UPDATE, and DELETE statements, including table, columns, values, and row predicates. The preview confirms the generated operation but cannot validate business intent. Submit only after review.
On partial failure, inspect the exact statement and transaction behavior before retrying. Refresh and query key rows after a successful commit.
Copy and export
Copy cells, rows, or results in supported formats. Export to .xlsx, .csv, or .sql, and verify column range, encoding, date formatting, and NULL representation. Spreadsheet programs may silently transform long numbers or dates.
Exported files leave database access controls and may contain sensitive data. Store them in controlled locations and delete them according to policy.