E08266: an injected component is always set, so checking that it is set is always true - use the injected value directly; for a fallback, use a new variable ('chosen <- supplied' then 'chosen :=? injected')

Category: E08xxx Pre-IR checks (flow analysis) · Reported in phase: FULL_RESOLUTION · Classification: INJECTED_COMPONENT_SET_CHECK · All error codes

What went wrong

You are checking whether an INJECTED component is set. An injected variable ('name as T!') is always set: the compiler proves that every injection a program can reach is registered in its application (E08210, E08220), so the application always supplies the component. The check is always true - it is dead code that reads as though the injection might not arrive, and any 'else' it guards never runs.

Every form of the check is rejected: 'logger?' as an if/while condition, in 'require' or 'assert', as a returned value or as a ternary's control; the coalescing operator '?:' with the injected variable on the left; a guard on the injected variable itself ('if log <- logger'); and a re-point checked in the head ('if log := logger then log?').

Worked examples

Compilable examples demonstrating this error and the correct fix:

  • ek9 -q 669 (Why can't I reassign an injected component variable?)

How to fix it

  1. Remove the check and use the injected value directly - it is always there.
  2. If you wanted a fallback, the fallback runs the other way round: keep the injected component as the default and assign it to a NEW variable only when another value is not set (good_fallback.ek9 in the [EXAMPLE] section below). The injected variable itself is never assigned (E08170).
  3. If the component may genuinely be absent, it should not be injected: pass it as a parameter, or hold it in an ordinary variable or property that can be unset.

Common causes

  • Defensive null-checking habit from Spring/Guice, where an optional injection can be null
  • 'require injected?' written to "touch" the variable - use it instead
  • The idiom 'x <- injected? <- injected else Default()', which looks like a fallback but never takes it
  • Checking a component property declared 'as T!' before calling it

See also

  • E08266: checking that an injected component is set (this error)
  • E08170: an injected variable is never assigned, ':=?' included
  • E08210: an injection reachable from a program is not registered in its application
  • E08220: a program reaches an injection but has no application
  • E08088: an isSet check that data flow proves is always true

Examples

Every example is compiled by the EK9 compiler: the first fails exactly as captioned, the others fix it.

Fails with E08266: an injected component is always set, so checking that it is set is always true - use the injected value directly; for a fallback, use a new variable ('chosen <- supplied' then 'chosen :=? injected')

#!ek9
defines module error.e08266.bad

  defines function

    report()
      logger as Logger!
      if logger?
        logger.log("report")

  defines component

    Logger as abstract
      log() as abstract
        -> message as String

Fixed: Use the injected component directly - it is always set

#!ek9
defines module error.e08266.good

  defines function

    report()
      logger as Logger!
      logger.log("report")

  defines component

    Logger as abstract
      log() as abstract
        -> message as String

Fixed: For a fallback, start from the other value and fill in with :=?

#!ek9
defines module error.e08266.good.fallback

  defines component

    Logger as abstract
      log() as abstract
        -> message as String

  defines function

    choose()
      -> supplied as Logger
      logger as Logger!
      chosen <- supplied
      chosen :=? logger
      chosen.log("choose")

Questions and answers

These Q&A entries show E08266 and its fix:

Found a problem?

If EK9 reported E08266 for a program you think is valid, or did not report it where it should have, use the compiler form with the smallest program that shows it. If this page is wrong or unclear, use the documentation form. Both are at github.com/ek9lang/ek9-issues.