Transformations

Column transforms are JavaScript expressions evaluated once per visible cell in that column. Open the column index menu → Transform, check the preview, then apply.

Hidden (filtered-out) rows are not transformed.

Expression context

Each expression runs with these bindings:

Name	Description
`self`	Current cell value (before transform)
`cols`	Array of cell values by column index (`cols[0]` = first data column)
`flds`	Object of cell values keyed by field name
`$0`, `$1`, …	Legacy aliases for `cols[N]`
`_ColumnName`	Legacy aliases for `flds.ColumnName`
`data`	Row sidecar — `data.csv` (original import), `data.zbx` / custom Pull labels
`ws`	All worksheets — `ws.sheet1`, `ws.MySheet`, …
`lastResponse`	Last Zabbix API response for this row (Pull/Push)
`lastError`	Last error for this row
`json()`	`JSON.stringify` shorthand
`obj()`	`JSON.parse` shorthand
`formatUnix()`	Unix seconds or ms → locale date/time string
`parseUnixMs()`	Parse Unix seconds/ms to a millisecond timestamp

Standard JavaScript operators and built-ins are available.

Objects returned from expressions are JSON-stringified for display in the cell when needed.

Helper functions

json(value)        // JSON.stringify(value)
obj(string)        // JSON.parse(string)
formatUnix(value)  // Unix seconds or ms → locale date/time string
parseUnixMs(value) // Unix seconds or ms → number (ms)

Typical pattern for nested API fields:

json(obj(data.csv.tags).concat({"tag":"env","value":"prod"}))

Convert a Zabbix clock column:

formatUnix(self)

Or use the column menu Format as date/time for a one-click convert of visible cells (this mutates stored and exported values).

Legacy vs modern names

These are equivalent:

flds.hostname
_hostname

cols[2]
$2

New examples in this documentation use flds / cols.

Saved expressions

The transform dialog can:

Load saved expressions (transfo_* in localStorage)
Save the current expression
Delete a saved expression

DataTable API (`ws.sheetName`)

Access another sheet (or the current one by name):

ws.sheet1.rows          // array of Row objects
ws.sheet1.col(2)        // all cells in column 2, or .col("hostid")
ws.sheet1.get_row(0)    // row by index
ws.sheet1.lookup(subject, searchCol, valueCol, regexp?)
ws.sheet1.lookup2(terms, valueCol, regexp?, defaultOnEmpty?)

`lookup(subject, searchCol, valueCol, regexp?)`

Find the first row where searchCol matches subject, return valueCol.

Parameter	Type	Description
`subject`	string	Value or regex pattern to match
`searchCol`	number or string	Column index or field name to search
`valueCol`	number or string	Column to return
`regexp`	boolean	If true, treat `subject` as regex (default: exact match, case-insensitive)

Example — map host to asset ID from a lookup sheet:

ws.assets.lookup(flds.hostname, "hostname", "asset_id", false)

`lookup2(terms, valueCol, regexp?, defaultOnEmpty?)`

Multi-column exact match. terms is an object of field → value pairs; every pair must match the same row.

ws.assets.lookup2({ hostid: flds.hostid, env: "prod" }, "owner", false, "")

Parameter	Type	Description
`terms`	object	e.g. `{ hostid: "1001", env: "prod" }`
`valueCol`	number or string	Column to return from the matched row
`regexp`	boolean	Treat term values as regex (default: exact, case-insensitive)
`defaultOnEmpty`	any	Returned when no row matches

Row object (in scripts / debug)

When iterating ws.sheet1.rows or inspecting row debug info:

Property / method	Description
`cells`	Array of Cell objects
`vals`	Array of cell values by column index
`fld_vals`	Object of field name → value
`isHidden`	`true` if filtered out
`isSelected`	`true` if checkbox checked
`data`	Attached data objects
`cell(idx)`	Cell by index or field name

Common patterns

Copy value from another column

cols[0]

Uppercase a field

flds.hostname.toUpperCase()

Compose a name from several columns (legacy style still fine)

flds.CountryCode + " - " + flds.DevType + " - " + flds.IP

Build JSON for a push column

json([{"tag":"owner","value":cols[2]}])

Conditional value

cols[3] === "0" ? "disabled" : "enabled"

Count pulled items

data.items ? data.items.length : 0

Cross-sheet lookup

ws.assets.lookup(flds.hostid, "hostid", "owner", false) || "unknown"

Preview vs Apply

Action	Behaviour
Editing the expression	Updates preview using the first visible row
Apply	Runs on all visible rows in the column

Errors in expressions show as the error message string in the cell.

Tips

Use filters to limit which rows get transformed before Apply
Keep complex enrichment in a Pull, then transform the flattened result
Double-click a field name to rename columns for clearer flds references
Prefer small, readable expressions; use the script editor for multi-step automation

See Examples for full walkthroughs.

Edit this page on GitHub

← All documentation