The table used in every example

All results on this page were produced by Tabulizer's own range engine on this 10-row table. Rows and columns are numbered from 1, and row 1 is the header.

1. Element: what kind of thing the rule targets

The element decides how the range below is read: for rows and columns it selects row or column numbers, for cells it selects cells, and for the table it is a condition the whole table must meet.

2. Range: which rows, columns or cells

Counting from the end with L

Wherever a row or column number is expected you can count from the end: L1 is the last one, L2 the one before it, and so on. In a 10-row table L3 is row 8. This keeps a rule correct when the number of rows changes, for example 2-L1 for "every row except the header".

Row and column ranges

The part in square brackets after MOD, ODD and EVEN is a limit subrange. It accepts from-to, >, >=, <, <=, FIRST, LAST and ALL. Numbering stays the table's own: ODD [2-6] gives rows 3 and 5, not "the first and third rows after 2".

Column ranges work the same way on column numbers. On the sample table, LAST 2 selects columns 4 and 5.

Combining ranges: ; # ^

Three symbols let you build a precise selection out of simple pieces:

• ; union: separates subranges; the result is everything any of them selects.
• # filter: keeps only what the next part also selects.
• ^ exclude: removes what the next part selects.

Filters and exclusions work inside their own subrange, before the union. Take <3;>=5#2-4;4-6^3-5:

In practice the most useful combinations are short. EVEN;1 selects the even rows and the header (1, 2, 4, 6, 8, 10), and 2-L1#ODD selects the odd data rows (3, 5, 7, 9).

Selecting rows and columns by their content

A row or column can also be chosen by what its cells contain. Text and comparison values are stored base64-encoded in the ruleset file, so the rule editor includes a Base64 encoder/decoder, and the online wizards encode for you. The tables below show the plain value and the stored form.

The variants follow a pattern: i ignores case, n negates, and ni does both, so the full set is SAME, SAMEi, SAMEn, SAMEni, CONTAIN, CONTAINi, CONTAINn and CONTAINni. EMPTY selects rows with an empty cell, and EMPTYT also counts cells that contain only spaces.

The value given to GREATER, GREATERE, LESS, LESSE and EQUAL says how to compare:

• #32: a number;
• [2026-09-10] or [-30 days]: a date, including relative dates;
• "Jones": text, compared alphabetically.

BETWEEN takes two values of the same kind: #15 #25, [2026-09-08] [2026-09-10] or "S" "T". For numbers and dates both ends are included. For text the second value is not, so "S" "T" selects text from S up to, but not including, T.

Minimum, maximum and median

MIN, MAX and MEDIAN select the row, column or cell holding the lowest, highest or middle value inside a search area, written as [row search range],[column search range]:

The search ranges accept numbers, lists, from-to, >, >=, <, <=, FIRST, LAST, ALL and the L notation, plus three relative values: SAME for the row or column currently being checked, P1, P2... for previous ones and N1, N2... for next ones. MAX [SAME],[2-L1] on a cell rule, for example, looks for the highest value within each cell's own row.

Cell ranges

Cells also accept > and >=, the other text operators (SAMEi, SAMEn, SAMEni, CONTAIN, CONTAINn, CONTAINni), EMPTY, EMPTYT, GREATER, GREATERE, LESS, LESSE, EQUAL, MIN and MEDIAN. Decimal values use a dot, as in 0-3.5.

Table ranges

For the table element the range is a condition: the rule applies to the whole table only when it is true.

ROWS, COLUMNS and CELLS accept >, >=, < and <=, and all the SAME and CONTAIN variants work too. A typical use: turn on pagination only when ROWS > 25, so short tables stay simple.

3. Range key: counting groups instead of rows

When cells are merged (autospan), a group of rows can look like one entry. In the sample table, Monday and Wednesday each have two activities. Zebra stripes with ODD would then change colour in the middle of a day.

Range Key fixes that. Set it to the column that defines a group, here 1 for Day, and the range counter moves on only when that column's value changes, not on every row. With Range = ODD and Range Key = 1, whole days are striped: both Monday rows share one colour, Tuesday the other, both Wednesday rows the first colour again.

4. Element section: only part of each selected element

A range picks whole rows or whole columns. Element Section narrows each of them down to part of it. For a row rule it lists the columns to keep, and for a column rule the rows.

For example, Element = row, Range = 1-5 and Element Section = 1,3 applies the rule only to the 1st and 3rd cells of the first five rows. Another example, on the sample table: to make the places in the busiest sessions bold without touching the other columns, use Element = row, Range = GREATER IzIw [4] (more than 20 places) and Element Section = 4.

5. Priority: the order rules run in

Order matters. If one rule replaces "Y" with "In stock" and another highlights cells containing "In stock", the replacement has to run first.

By default Tabulizer decides the order itself, using the kind of work each rule does:

When that order is not what you want, give a rule a Priority. Priority is a whole number, and the higher it is, the earlier the rule runs. The explicit priority is compared first; only rules with equal priority, including the default 0, fall back to the automatic order above. So a rule with priority 1 runs before every rule left at 0, whatever it does.

Common questions