ess you night nullsafe opeFollow these rules:
⚠ Copy pasted the code is forbidden; to share functionality use functions.
All Groovy source code (“element” folders) should be auto-formatted by IDEA, see /wiki/spaces/LEARN/pages/2651979915.
Note: logic.json should be excluded! It is formatted by Packaging Tool.Always use curly brackets for
if
,for
,while
, e.g.:if (isCondition) { return null }
Reformatting of big amount of old code should be done within a separate commit, otherwise it is almost impossible to validate the merge request.
Follow the Recommended Logic Structure. Try to keep all input elements in the beginning of the logic.
Use
out.SomeElement
instead of (deprecated)api.getElement("SomeElement")
.Use
input.SomeInputName
instead of (deprecated)api.input("SomeInputName")
.Avoid api.local, api.global where possible. Using them in functions is forbidden except for caching. For small amount of data, use a dedicated element – this will help you trace the logic more easily. Big data cause a slowdown in line item logics.
Use field names in filters or use
api.namedEntities()
if possible.Use real data type when declaring the variable instead of
def
. Mandatory use in functions – both parameters and the return type.Prevent api.isSyntaxCheck Issues by having the AbortSyntaxCheck element in the logic.
Keep elements/functions small and simple. Shorter methods are always better than big ones. Try to break bigger element code / function code into smaller pieces.
Where to place functions?
1. In the element – if used only once.
2. In the Library element – if used twice or more, but only within the same logic.
3. In Groovy Library – if used between logics.Use Input Builders – mandatory for configurators and header logics, but recommended also for ordinary logics.
Use
@Field
constants (no use for fields, table names, configurator names in customer projects). Mandatory use for important values used in conditions or repeatedly used constant Strings or used in Accelerators.Use
api.findLookupTableValues("TableName")
instead ofapi.find("MLTVx", ...)
= you will get all rows.Use consistently one approach for adding items to a list, do not mix various approaches. It is recommended to use
List << value
rather thanList.add(value)
(unless you need nullsafe operator).Use the same approach for accessing a List item, do not mix various approaches. Use either
list[index]
orlist?.getAt(index)
. Notlist.get(index)
.Use the same approach for accessing Map values, do not mix various approaches. It is recommended to use
Map.key
rather thanMap[key]
orMap.get(key)
unless there are spaces in 'key'.BigDecimal
should be the most commonly used data type for amount or percentage variables (never use floats for amounts or percentages). UseBigDecimal
constants (e.g.0.0
instead of0
) so you can rely it will be aBigDecimal
.If casting is needed, e.g.
?.toBigDecimal()
for amounts, do this at the first occurrence of the variable.Use closures (
collect
,collectEntries
,each
,eachWithIndex
,find
,findAll
,first
,last
,max
,min
,inject
,sort
, etc.) whenever applicable.
Note: For better performance, use thesortBy
parameter inapi.find()
rather than thesort()
closure.E.g. instead of for LOOP:
for (item in items) { ... }
use
items.each { item -> ... }
Instead of FOR loop:
def list for (item in items) { list.add(x) }
use collect:
items.collect { item -> x }
Instead of FOR loop:
def map for (item in items) { map.put(x, y) }
use collectEntries:
items.collectEntries { item -> [(x) : y ] }
Use "find" (resp. "query") prefix for names of functions that query Master Data tables (resp. Analytics Data Sources / Datamarts). Do not mix data querying and calculation in the functions.
Use the library element alias when calling multiple Groovy library functions in an element or function:
def CurrencyUtils = libs.MainLib.CurrencyUtils
Casting: instead of
def xxx = out.Cost as BigDecimal
or?.toBigDecimal
move the cast to thereturn
of the element.For performance reasons, avoid returning big data in an element. Remember: in Groovy, the expression on the last line is the value returned!
Initialize
Map
directly. E.g. instead ofdef map = [:] map["key1"] = "value1" map["key2"] = "value2" map["key3"] = "value3"
directly use (and following auto-format):
def map = [ key1: "value1" key2: "value2" key3: "value3" ]
Analogically, initialize
List
directly. E.g. instead ofdef list = [] list << "value1" list << "value2" list << "value3"
directly use (and following auto-format):
def list = [ "value1", "value2", "value3" ]
For assignments based on IF
def result if (condition) { result = x } else { result = y }
directly use:
def result = condition ? x : y
or for longer x, y expressions:
def result = condition ? x : y
Split suuuupeeeeeerlooooong lines. If the function has many parameters, put a group of parameters on a separate line.
Dot not commit commented code and
api.trace
in GIT. Avoidapi.log
unless it is really needed for the PROD run.Resolve the TODOs.