Skip to content

Latest commit

 

History

History
93 lines (76 loc) · 3.36 KB

Elements-of-the-configuration-file.md

File metadata and controls

93 lines (76 loc) · 3.36 KB
Raw

Configuration files are INI-like files that use notion of the sections ([section]) and relevant subsections ([[subsection]]). Sections and subsections specify the relationship between various parts of the system. Main top-level sections used in the configuration are:

  • group - all settings related to a specific component or group of components
  • table and sysTable - data flow and data model settings
  • user, technicalUser, and userGroup – access related parameters

Subsections are context specific but always describe component within the system. Each configuration variable has to be declared in the schema definition file as a tuple of type and (optionally) the default value, for example:

cfg.timeout = <type(INT), default(1000)>
  • type – mandatory part; it is an equivalent of the standard data type declaration; (see below)
  • default – optional; default value to be used if variable definition is not present in appropriate configuration file

Available data types

These are just the typical data types known from the other programming languages that can be used in type declarations in qsd files (with their Java equivalents, where possible):

  • BOOLEAN – possible values: TRUE, FALSE, 1, 0

  • SHORT – Short in Java

  • INT – Integer in Java

  • LONG – Long in Java

  • REAL - Float in Java

  • FLOAT - Double in Java)

  • GUID – UUID in Java

  • TIME – Time in Java

  • TIMESTAMP – Timestamp in Java

  • TIMESPAN – Timespan in Java

  • DATETIME – Timestamp in Java

  • DATE – Date in Java

  • STRING - Character array in Java

  • SYMBOL - String in Java

  • PATH – path as understood by the kdb+, for example:

    eodPath = ${EC_SYS_PATH}/data/rtr.hdb

    where referenced EC_SYS_PATH variable is resolved to its value

  • PORT – integers, such as:

    port = 5001
port = ${basePort} + 10 # simple arithmetic operations are allowed

  • LIST - a list of atomic types. For example, the following line in a schema (.qsd) file declares a mandatory configuration variable named cfg.ports that is a list of integers:

    cfg.ports = <type(LIST INT)>

    To assign value to that variable the configuration file needs to contain a line like the following:

    cfg.ports = 17024,17004,17009,17020,17005,17021,17001,17006

  • TABLE - the following dataflow.qsd definition:

    model = <type(TABLE), col1(TYPESPEC), col2(TYPESPEC)>

    Example four-column table definition to use in dataflow.cfg:

    model = time(TIME), sym(SYMBOL), price(FLOAT), size(LONG)

  • ARRAY – array of TABLE entries (or a TABLE with data). ARRAY variables can be defined using following syntax:

    ARRAY = (row1,row2,row3, ... ,rowN)

    where:

    rowN = (col1(val1),... , colM(valM))

    is similar to TABLE definition. Example definition of ARRAY variable:

    tab = ((timezone(UTC), descr(Coordinated Universal Time), offset(00:00:00.000)),
       (timezone(CET), descr(Central European Time),      offset(01:00:00.000)))

    will yield following q table:

    timezone descr offset 
    UTC    | Coordinated Universal Time  | 00:00:00.000
CET    | Central European Time       | 01:00:00.000