Funktions- und Sprachreferenz
NX, the Ninox query language, is a powerful programming language which allows you to quickly extend Ninox databases with calculations and trigger actions.
While Ninox provides a visual function editor as described in the online manual, it also sports a text-mode for entering more complex expressions. This manual covers the text mode.
Here is an example database covering most of the topics of the tutorial.
NX provides arithmetic operations comparable to most other programming languages:
1 + 2 = 3
3 - 2 = 1
2 * 3 = 6
10 / 5 = 2
Arithmetic operations respect the typical operator precedence:
1 + 2 * 3 = 1 + (2 * 3) = 7
Expressions may be enclosed in parentheses:
(1 + 2) * 3 = 9
Note, that NX will automatically remove insignificant parentheses, i. e. if you enter 1 + (2 * 3) it will store this expression as 1 + 2 * 3.
A string, i. e. a sequence of characters, may be entered enclosed in double-quotes:
"This is a string."
Strings may be concatenated using the + operator:
"Hello World!" + " " + "This is a string."
-- "Hello World! This is a String."
Strings may also be concatenated with other values:
"Hi " + 42 = "Hi 42"
2 + " fast " + 4 + " U" -- "2 fast 4 U"
To enclose a double-quote within a string, it has to be escaped as “”:
"Hi, my name is ""Sally""!"
Working with Fields
Most NX functions operate in the context of a specific record. NX gives you access to all fields of that record. Consider a table Customer consisting of fields First Name, Last Name, Street, ZIP, City. To generate the full address line, you may enter:
Street + ", " + ZIP + " " + City
Some field names may contain special characters like spaces, dots, colons and so on. Also they may collide with a reserved keyword of NX, like if, then, else. To work around this, field names can be enclosed in single quotes:
'First Name' + " " + 'Last Name'
If a field name contains a single quote, this has to be escapes as ”:
'Last year''s total revenue'
Note, that NX is agnostic to field name changes. That is, if you’ve entered an expression referring to a field and you change that field’s name afterwards it will not have an impact to the expression. This is due to the fact that NX internally stores the expression with a reference to the field’s ID (which cannot me modified). Actually, if you open the function editor again, it will reflect the field’s new name.
Working with Table References
NX expressions can navigate along table references as well. Table references are always one-to-may relationships, e.g. one Customer may have many Invoices — where the table reference is defined as from Invoice to Customer (Invoice => Customer).
To display the Customer’s name in an Invoice, you’d write:
Customer.'First Name' + " " + Customer.'Last Name'
Expressions may even jump along multiple hops, consider the case where a Customer has a reference to a Company. You could then display the Company’s in the Invoice as well:
Vice-versa, it’s also possible to get Invoice information for a Customer. Remember, that a Customer may have multiple Invoice’s. Thus, an expression referring to the Customer’s invoices will return multiple values. In order to display those values, some kind of aggregation has to be applied. The most common one is to sum up values:
But there are other aggregations as well:
avg(Invoice.Amount) -- the average, ignoring empty fields
cnt(Invoice.Amount) -- the count of non-empty fields
min(Invoice.Amount) -- the minimum value,
ignoring empty fields
max(Invoice.Amount) -- the maximum value,
ignoring empty fields
first(Invoice.Amount) -- the first value, according to
Ninox' internal sorting
last(Invoice.Amount) -- the first value, according to
Ninox' internal sorting
concat(Invoice.Amount) -- lists all values, separated by ", "
Logical expressions check if something is the case or not. Most often you’ll need them to create specific filter rules or to make decisions within a calculation. A logical expression either returns true or false. The most common form of a logical expression is a comparison of two values, like:
Amount > 100 -- greater than
Amount < 100 -- less than Amount >= 100 -- greater than or equal to
Amount <= 100 -- less than or equal to
Amount = 100 -- equal
Amount != 100 -- not equal
Logical expressions may be combined by and, or, not.
Age > 12 and Age < 18
Status = 1 or Status = 2 or Status = 3
not (Status = 4 or Status = 5)
Status != 4 and Status != 5
Note, the operator precedence is not > and > or.
In the previous chapter you’ve learned about logical expressions. The result of such an expression may be used in a an if / then / else expression:
if Age < 18 then "Child" else "Grown-up"
if Age < 18 then
if Age < 13 then
Please note, that the then and else part have to return values of the same type — see chapter “Understanding Types”.
In the chapter “Working with Table References” you’ve already learned how to access multiple values from a table referring the current one — like sum(Invoice.Amount). With logical expressions, these results may also be filtered. A filter expression has to be enclosed in braces:
sum(Invoice[Status = 2].Amount)
This will only sum up the Invoice Amounts of Invoices with Status = 2.
NX provides a range of built-in functions allowing you to transform values. A function call has the form:
function ( argument1, argument2, … )
age(Birthdate) -- The current age of a person with given Birthdate in years
cos(5) -- Cosinus of 5
Sometimes it can be useful to store the result of a intermediate calculation and do further calculations based on that result. Consider the case where you want to check the age of a person like:
if age(Birthdate) > 18 then "Grown-up"
else if age(Birthdate) > 12 then "Teenager"
Since age(Birthday) is used multiple times, things may be simplified with a variable which stores the result of the age calculation:
let a := age(Birthdate);
if a > 18 then "Grown-up" else if a > 12 then "Teenager" else "Child"
A variable is declared with a let statement:
let variable := expression;
Any expression following that let statement can make use of that variable.
Some expressions in NX may also modify data. Up to the current release 1.5 of Ninox, this is only allowed for trigger expressions (field option Trigger after update). While further enhancements are planned, modifying data is currently restricted to change the value of record fields.
As an example, consider a table Article with a field Price (number) and another table Invoice Item with fields Article (reference to table Article) and Price (number). After assigning the Article, Ninox shall copy the Article’s Price to the Invoice Item’s Price. This can be achieved with a Trigger after update on Article containing following expression:
Price := Article.Price
It is also possible, to update multiple fields, using a semicolon:
Price := Article.Price;
Description := Article.'Article No' + " " + Article.Name
With the create statement it’s possible to create new records for a table. The following example creates a new record for the table Person and stores a reference to that record in the variable p. It then assigns a Name to the newly created Person.
let p := create Person;
p.Name := "Sarah Conner"
Typically, formulas will access data from related tables like descibed above (“Working with Table References”). However, using the select statement there’s also a way to access data from an unrelated tables.
Will look up all Person records, this statement returns an array of record references.
select Person where ZIP="10000"
This statement selects only the Persons with the given ZIP value 10000.
Selecting records is useful in many situations, one common case is to calculate the next Invoice ID.
'Invoice ID' :=
max((select Invoices).'Invoice ID') + 1
With the “delete” command, it’s possible to remove one or more records. Be careful in combination with select, because this may affect a wide range of records with a single call. Some examples:
delete first(select Person where 'First Name' like "Hans")
This will look up the first record from the Person table where the name contains “Hans” and remove it.
delete select Person where 'First Name' like "Hans"
This will look up the all records from the Person table where the name contains “Hans” and remove all matching records.
delete select Person
Will remove any record in the person table.
When issued on an invoice record, this will remove the customer that is linked to the current invoice.
When issued on a customer record, this will remove all invoices which are linked to the current customer.
The for statement can be used to loop over arrays of values – e.g. the result of a select.
for p in (select Person)
p.Haircolor := "red"
This exmple would lookup all Person records (select Person). For each record (stored in the variable p) it would then change the Person‘s Haircolor to red.
Use the range(From, To) function to create counting loops:
for i in range(0, 10) do ... end
Please note, that for range(From, To), From in inclusive and To is exclusive. To state an example:
=> 0, 1, 2, 3
concat(for i in range(0, 4) do i*i end)
=> 0, 1, 4, 9
Some expressions will return arrays, that is lists of values, as a result. E. g. the reverse end of a relation will return an array of record ids:
=> [ID1, ID2, ID3]
It is also possible to create an array explicitly with the following syntax:
let MyArray1 := [1, 2, 81];
let MyArray2 := [1, 2, 9 * 9];
Tries to interpret the given value as a number.
- If value stems from a choice field, this will be the choice’s internal id. Use number(text(choice)) to get a numeric representation of the choice’s text.
- If value is a date, time or timestamp, this will be the the number of milliseconds between midnight of January 1, 1970 and the specified date.
- If value is an appointment, it is treated like the appointment’s begin date.
number("10") => 10
number(5) => 5
number(now()) => 1440681777046
Converts any value to a string representation possibly reflecting the format option of its field settings. If there’s no value, the result will be the empty string: “”.
text("Hello") => "Hello"
text(2.34) => "2.34"
text(MyCurrencyField) => "1,234.56 $"
text(today()) => "08/27/2015"
round(x) – Rounds a number to the nearest integer.
round(x, digits) – Rounds a number with the given amount of digits.
E.g. round(1.234, 2) = 1.23
floor(x) – Rounds a number DOWNWARDS to the nearest integer.
ceil(x) – Rounds a number UPWARDS to the nearest integer.
sqrt(x) – The square root of x.
sqr(x) – The square of x: x2
sign(x) – The signum of x:
sign(-2.5) => -1
sign(2.5) => 1
sign(0) => 1
abs(x) – The absolute value of x: abs(-5) = abs(5) = 5
sin(x) – Sinus of x (in radians).
cos(x) – Cosinus of x (in radians).
tan(x) – Tangens of x (in radians).
asin(x) – Arcus sinus of x (in radians).
acos(x) – Arcus cosinus of x (in radians).
atan(x) – Arcus tangens of x (in radians).
atan2(x) – Arcus tangens of x (in radians), squared.
random() – A random number between 0 (incl.) and 1 (excl.)
pow(x, y) – x to the power of y: xy
exp(x) – 10 to the power of x: 10x
log(x) – Logarithm of x to the base of 10.
log(x, y) – Logarithm of x to the base of y.
ln(x) – Natural logarithm of x.
length(string) – Returns the string’s length (the number of characters).
trim(string) – Removes leading and trailing white-space of string.
lower(string) – Converts a string to lower case.
upper(s) – Converts a string to upper case.
lpad(string, length, padding)
If string’s length is smaller than the given length, the missing space is filled up with the given padding at the start of the string.
rpad(string, length, padding)
If string’s length is smaller than the given length, the missing space is filled up with the given padding at the end of the string.
substring(string, start, end) – Extracts a part of the string. Start and end are zero-based.
substr("Hello World!", 0, 5) => "Hello"
substr("Hello World!", 6, 10) => "World"
substr(string, start, length) – Extracts a part of the string. Start is zero-based.
substring("Hello World!", 0, 5) => "Hello"
substring("Hello World!", 6, 5) => "World"
contains(string, match) – Checks if string contains the given match string by exact comparison.
contains("Hello World!", "World") => true
contains("Hello World!", "world") => false
index(string, match) – Finds the start index of the given match string within string.
index("Hello World!", "World") => 6
index("Hello World!", "world") => -1 (not found)
replace(string, find, replace) – Replaces any occurance of find with replace.
replace("Hello World", "l", "X") => "HeXXo WorXd"
string.replace(new RegExp(pattern, flags), replace)
replacex("Hello World", "l", "g", "X") => "HeXXo WorXd"
date(year, month, day) – Returns a date value, e.g. date(2016, 9, 27).
date(number) – Converts a number to a date value. The number represents the number of milliseconds since the UNIX epoch ( 00:00:00 UTC on 1 January 1970).
date(datetime) – Converts a date+time value to a date value (leaving out the time value).
year(date) – Full year of the given date.
year(datetime) – Full year of the given date+time value.
year(appointment) – Full year of the start date of the given appointment.
month(date) – Month of the given date (1 = January, … 12 = December).
month(datetime) – Month of the given date+time value.
month(appointment) – Month of the start date of the given appointment.
monthName(number) – Name of the given month number, e.g. monthName(1) = “January”.
monthIndex(text) – Number of the given month name, e.g. monthIndex(“January”) = 1.
day(date) – Day of month of the given date (between 1 and 31).
day(datetime) – Day of month of the given date+time value.
day(appointment) – Day of month of the start date of the given appointment
weekday(date) – Weekday of the given date (0 = Sunday, 1 = Monday, … 6 = Saturday)
weekday(datetime) – Weekday of the given date+time value
weekday(appointment) – Weekday of the start date of the given appointment
weekdayName(number) – Name of the given weekday number, e.g. weekdayName(0) = “Sunday”
weekdayIndex(text) – Number value of the given weekday name, e.g. weekdayIndex(“Sunday”) =0
today() – The current date (without time).
now() – The current timestamp.
age(date) – Number of full year’s between now and the given date (e.g. a person’s age).
Formats a date as a string. The format expression is a string which may contain following tokens (example for 9th of August 2015, 01:02:03 am):
Token Description Example
YY two digit year 15
YYYY four digit year 2015
M one or two digit month 8
MM two digit month 08
MMM abbreviated month name Aug
MMMM full month name August
D one digit day 9
DD two digit day 09
Do day ordinal 9th
dd day name short Sat
ddd day name short Sat.
dddd day name full Saturday
h one digit hour 1
hh tow digit hour 01
m one digit minute 2
mm two digit minute 02
s one digit second 3
ss two digit second 03
a am or pm am
Year and month of a date as a string, e.g. "2015/08". Useful for grouping records per month.
Year and quarter of a date as a string, e.g. "2015/03". Useful for grouping records per quarter.
Year and week of year of a date as a string, e.g. "2015/32". Useful for grouping records per calendar week.
Calendar week of a date as a number.
start(appointment) – Start timestamp of an appointment.
endof(appointment) – End timestamp of an appointment.
duration(appointment) – Duration of an appointment in milliseconds.
days(start, end) – Number of days between two dates.
workdays(start, end) – Number of working days between two dates. This function does consider any Monday to Friday to be working days, it does not respect holidays.
time(number) – converts a number value (milliseconds) to a time value.
time(hours, minutes, seconds)
time(hours, minutes, seconds, millis)
Returns a time value with given hours, minutes, seconds and milliseconds.
datetime(year, month, day)
datetime(year, month, day, hour)
datetime(year, month, day, hour, minute)
datetime(year, month, day, hour, minute, second)
datetime(year, month, day, hour, minute, second, millis)
Returns a date+time value with given year, month, date, hour, minute, second and millisecond.
datetime(date) – converts a date value to a date+time value, time is set to 00:00.
datetime(date, time) – converts a date and a time value to a date+time value.
datetime(number) – converts a number value to a date+time value, the number value represents the number of milliseconds since the UNIX epoch.
timeinterval(number) – converts a number (milliseconds) to a timeinterval value.
timeinterval(time) – converts a time value to a timeinterval value.
color(text) – converts a text to a color value, e.g. color("red"), color("#f00"), color("ff0000"), color("rgb(255,0,0)"). color accepts any valid HTML/CSS color identifier.
color(red, green, blue) – creates a color value from rgb-components. red, green and blue a numbers between 0 and 255.
color(red, green, blue, alpha) – creates a color value from rgb- and alpha-components. red, green and blue a numbers between 0 and 255, alpha is between 0 and 1.
color(choice) – returns the color of a choice value.
icon(name) – returns an icon, e.g. icon(“warn”)
icon(choice) – returns the icon of a choice value
A full list of all icons and its corresponding names can be found here.
longitude(Location) – the longitude value of a location field
latitude(Location) – the latitude value of a location field
location(title, latitude, longitude) – creates a location value with given title (text), latitude (number) and longitude (number).
User Interface Automation
User interface automation has been added with Ninox 1.7.0. It’s not yet available for Ninox Server and Cloud.
Creates a PDF from the given record and print layout name.
printRecord(record(Invoices, 1), "My Invoice Template")
Opens the print layout editor.
openPrintLayout(record(Invoices, 1), "My Invoice Template")
Opens the record and jumps to the related table.
Opens the record in a popup form without switching to another table.
Opens a table view.
openTable("Invoices", "All Invoices")
User Management Functions
Pls. note that the following functions do only work for Ninox Server and Cloud installations. They won’t return useful results for Ninox App.
user() – returns the current user value
userId(id) – returns the user value with given id
userId() – returns the id of the current user.
userId(user) – returns the id of the given user value.
userName() – returns the name of the current user.
userName(user) – returns the name of the given user value.
userFirstName() – first name of the current user
userFirstName(user) – first name of the given user value
userLastName() – last name of the current user
userLastName(user) – last name of the given user value
userFullName() – full name of the current user
userFullName(user) – full name of the given user value
userEmail() – email address of the current user
userEmail(user) – email address of the given user value
raw(value) – returns the internal / raw representation of a value as a text.
Returns a rich text representation of a text.