Legacy providers expect Terraform to be able to clean up invalid plans
and computed attributes. Add a special case for the LegacyTypeSystem to
revert `ignore_changes = all` to the complete prior state.
When handling ignore_changes=all, we must filter computed attributes
from the prior state to prevent them showing in the configuration. Since
it's not valid for the user to have set computed attributes in the
config, the provider should expect to never see any values there. The
oversight has only now become apparent, as more providers adopt the
plugin-framework which has direct access to the plan-time configuration
value.
Go 1.19's "fmt" has some awareness of the new doc comment formatting
conventions and adjusts the presentation of the source comments to make
it clearer how godoc would interpret them. Therefore this commit includes
various updates made by "go fmt" to acheve that.
In line with our usual convention that we make stylistic/grammar/spelling
tweaks typically only when we're "in the area" changing something else
anyway, I also took this opportunity to review most of the comments that
this updated to see if there were any other opportunities to improve them.
By observing the sorts of questions people ask in the community, and the
ways they ask them, we've inferred that various different people have been
confused by Terraform reporting that a value won't be known until apply
or that a value is sensitive as part of an error message when that message
doesn't actually relate to the known-ness and sensitivity of any value.
Quite reasonably, someone who sees Terraform discussing an unfamiliar
concept like unknown values can assume that it must be somehow relevant to
the problem being discussed, and so in that sense Terraform's current
error messages are giving "too much information": information that isn't
actually helpful in understanding the problem being described, and in the
worst case is a distraction from understanding the problem being described.
With that in mind then, here we introduce an explicit annotation on
diagnostic objects that are directly talking about unknown values or
sensitive values, and then the diagnostic renderer will react to that to
avoid using the terminology "known only after apply" or "sensitive" in the
generated diagnostic annotations unless we're rendering a message that is
explicitly related to one of those topics.
This ends up being a bit of a cross-cutting concern because the code that
generates these diagnostics and the code that renders them are in separate
packages and are not directly aware of each other. With that in mind, the
logic for actually deciding for a particular diagnostic whether it's
flagged in one of these special ways lives inside the tfdiags package as
an intermediation point, which both the diagnostic generator (in the core
package) and the diagnostic renderer can both depend on.
When a data resource is used for the purposes of verifying a condition
about an object managed elsewhere (e.g. if the managed resource doesn't
directly export all of the information required for the condition) it's
important that we defer the data resource read to the apply step if the
corresponding managed resource has any changes pending.
Typically we'd expect that to come "for free" but unfortunately we have
a pragmatic special case in our handling of data resources where we
normally defer to the apply step only if a _direct_ dependency of the data
resource has a change pending, and allow a plan-time read if there's
a pending change in an indirect dependency. This allowed us to preserve
some compatibility with the questionable historical behavior of always
reading data resources proactively unless the configuration contains
unknown values, since the arguably-more-correct behavior would've been a
regression for anyone who had been depending on that before.
Since preconditions and postconditions didn't exist until now, we are not
constrained in the same way by backward compatibility, and so we can adopt
the more correct behavior in the case where a data resource has conditions
specified. This does unfortunately make the handling of data resources
with conditions subtly inconsistent with those that don't, but this is
a better situation than the alternative where it would be easy to get into
a trapped situation where the remote system is invalid and it's impossible
to plan the change that would make it valid again because the conditions
evaluate too soon, prior to the fix being applied.
We have two different reasons why a data resource might be read only
during apply, rather than during planning as usual: the configuration
contains unknown values, or the data resource as a whole depends on a
managed resource which itself has a change pending.
However, we didn't previously distinguish these two in a way that allowed
the UI to describe the difference, and so we confusingly reported both
as "config refers to values not yet known", which in turn led to a number
of reasonable questions about why Terraform was claiming that but then
immediately below showing the configuration entirely known.
Now we'll use our existing "ActionReason" mechanism to tell the UI layer
which of the two reasons applies to a particular data resource instance.
The "dependency pending" situation tends to happen in conjunction with
"config unknown", so we'll prefer to refer that the configuration is
unknown if both are true.
Previously we had three different layers all thinking they were
responsible for substituting a default value for an unset root module
variable:
- the local backend, via logic in backend.ParseVariableValues
- the context.Plan function (and other similar functions) trying to
preprocess the input variables using
terraform.mergeDefaultInputVariableValues .
- the newer prepareFinalInputVariableValue, which aims to centralize all
of the variable preparation logic so it can be common to both root and
child module variables.
The second of these was also trying to handle type constraint checking,
which is also the responsibility of the central function and not something
we need to handle so early.
Only the last of these consistently handles both root and child module
variables, and so is the one we ought to keep. The others are now
redundant and are causing prepareFinalInputVariableValue to get a slightly
corrupted view of the caller's chosen variable values.
To rectify that, here we remove the two redundant layers altogether and
have unset root variables pass through as cty.NilVal all the way to the
central prepareFinalInputVariableValue function, which will then handle
them in a suitable way which properly respects the "nullable" setting.
This commit includes some test changes in the terraform package to make
those tests no longer rely on the mergeDefaultInputVariableValues logic
we've removed, and to instead explicitly set cty.NilVal for all unset
variables to comply with our intended contract for PlanOpts.SetVariables,
and similar. (This is so that we can more easily catch bugs in callers
where they _don't_ correctly handle input variables; it allows us to
distinguish between the caller explicitly marking a variable as unset vs.
not describing it at all, where the latter is a bug in the caller.)
Revert the evaluation change from #29862.
While returning a dynamic value for all expanded resources during
validation is not optimal, trying to work around this using unknown maps
and lists is causing other undesirable behaviors during evaluation.
Allow `GetResource` to return correct types values during validation,
rather than relying on `cty.DynamicVal` as a placeholder. This allows
other dependent expressions to be more correctly evaluated.
Historically the responsibility for making sure that all of the available
providers are of suitable versions and match the appropriate checksums has
been split rather inexplicably over multiple different layers, with some
of the checks happening as late as creating a terraform.Context.
We're gradually iterating towards making that all be handled in one place,
but in this step we're just cleaning up some old remnants from the
main "terraform" package, which is now no longer responsible for any
version or checksum verification and instead just assumes it's been
provided with suitable factory functions by its caller.
We do still have a pre-check here to make sure that we at least have a
factory function for each plugin the configuration seems to depend on,
because if we don't do that up front then it ends up getting caught
instead deep inside the Terraform runtime, often inside a concurrent
graph walk and thus it's not deterministic which codepath will happen to
catch it on a particular run.
As of this commit, this actually does leave some holes in our checks: the
command package is using the dependency lock file to make sure we have
exactly the provider packages we expect (exact versions and checksums),
which is the most crucial part, but we don't yet have any spot where
we make sure that the lock file is consistent with the current
configuration, and we are no longer preserving the provider checksums as
part of a saved plan.
Both of those will come in subsequent commits. While it's unusual to have
a series of commits that briefly subtracts functionality and then adds
back in equivalent functionality later, the lock file checking is the only
part that's crucial for security reasons, with everything else mainly just
being to give better feedback when folks seem to be using Terraform
incorrectly. The other bits are therefore mostly cosmetic and okay to be
absent briefly as we work towards a better design that is clearer about
where that responsibility belongs.
There are a few different reasons why a resource instance tracked in the
prior state might be considered an "orphan", but previously we reported
them all identically in the planned changes.
In order to help users understand the reason for a surprising planned
delete, we'll now try to specify an additional reason for the planned
deletion, covering all of the main reasons why that could happen.
This commit only introduces the new detail to the plans.Changes result,
though it also incidentally exposes it as part of the JSON plan result
in order to keep that working without returning errors in these new
cases. We'll expose this information in the human-oriented UI output in
a subsequent commit.
Previously terraform.Context was built in an unfortunate way where all of
the data was provided up front in terraform.NewContext and then mutated
directly by subsequent operations. That made the data flow hard to follow,
commonly leading to bugs, and also meant that we were forced to take
various actions too early in terraform.NewContext, rather than waiting
until a more appropriate time during an operation.
This (enormous) commit changes terraform.Context so that its fields are
broadly just unchanging data about the execution context (current
workspace name, available plugins, etc) whereas the main data Terraform
works with arrives via individual method arguments and is returned in
return values.
Specifically, this means that terraform.Context no longer "has-a" config,
state, and "planned changes", instead holding on to those only temporarily
during an operation. The caller is responsible for propagating the outcome
of one step into the next step so that the data flow between operations is
actually visible.
However, since that's a change to the main entry points in the "terraform"
package, this commit also touches every file in the codebase which
interacted with those APIs. Most of the noise here is in updating tests
to take the same actions using the new API style, but this also affects
the main-code callers in the backends and in the command package.
My goal here was to refactor without changing observable behavior, but in
practice there are a couple externally-visible behavior variations here
that seemed okay in service of the broader goal:
- The "terraform graph" command is no longer hooked directly into the
core graph builders, because that's no longer part of the public API.
However, I did include a couple new Context functions whose contract
is to produce a UI-oriented graph, and _for now_ those continue to
return the physical graph we use for those operations. There's no
exported API for generating the "validate" and "eval" graphs, because
neither is particularly interesting in its own right, and so
"terraform graph" no longer supports those graph types.
- terraform.NewContext no longer has the responsibility for collecting
all of the provider schemas up front. Instead, we wait until we need
them. However, that means that some of our error messages now have a
slightly different shape due to unwinding through a differently-shaped
call stack. As of this commit we also end up reloading the schemas
multiple times in some cases, which is functionally acceptable but
likely represents a performance regression. I intend to rework this to
use caching, but I'm saving that for a later commit because this one is
big enough already.
The proximal reason for this change is to resolve the chicken/egg problem
whereby there was previously no single point where we could apply "moved"
statements to the previous run state before creating a plan. With this
change in place, we can now do that as part of Context.Plan, prior to
forking the input state into the three separate state artifacts we use
during planning.
However, this is at least the third project in a row where the previous
API design led to piling more functionality into terraform.NewContext and
then working around the incorrect order of operations that produces, so
I intend that by paying the cost/risk of this large diff now we can in
turn reduce the cost/risk of future projects that relate to our main
workflow actions.
This is part of a general effort to move all of Terraform's non-library
package surface under internal in order to reinforce that these are for
internal use within Terraform only.
If you were previously importing packages under this prefix into an
external codebase, you could pin to an earlier release tag as an interim
solution until you've make a plan to achieve the same functionality some
other way.
