* huge change to weave new addrs.Provider into addrs.ProviderConfig
* terraform: do not include an empty string in the returned Providers /
Provisioners
- Fixed a minor bug where results included an extra empty string
Now we'll consider any message that doesn't start with a digit as a
continuation. This is _definitely_ in loose, best-effort territory now
since a continuation line could very well start with a digit, but
ultimately this is still better than the totally broken behavior that
preceded it.
Just to make sure that none of these heuristics interfere with producing
complete, accurate logs for TRACE, we'll skip the LevelFilter altogether
in that case.
The filtering for other log levels is unreliable and glitchy because it's
trying to infer information from the log stream that isn't reliably
represented.
Although the previous commit has improved the situation somewhat, it is
still a tricky and unreliable heuristic, so worth a warning to anyone who
is reading such a log that if they see something confusing it could be
a result of the heuristic not working fully.
In order to make this work reasonably we can't avoid using some funny
heuristics, which are somewhat reasonable to apply within the context of
Terraform itself but would not be good to add to the general "logutils".
Specifically, this is adding the additional heuristic that lines starting
with spaces are continuation lines and so should inherit the log level
of the most recent non-continuation line.
If a state mv target happens to be a resource that doesn't exist, allow
the creation of the new resource inferring the EachMode from the target
address.
* terraform/context: use new addrs.Provider as map key in provider factories
* added NewLegacyProviderType and LegacyString funcs to make it explicit that these are temporary placeholders
This PR introduces a new concept, provider fully-qualified name (FQN), encapsulated by the `addrs.Provider` struct.
Add deprecation warning for references from destroy provisioners or
their connections to external resources or values. In order to ensure
resource destruction can be completed correctly, destroy nodes must be
able to evaluate with only their instance state.
We have sufficient information to validate destroy-time provisioners
early on during the config loading process. Later on these can be
converted to hard errors, and only allow self, count.index, and each.key
in destroy provisioners. Limited the provisioner and block evaluation
scope later on is tricky, but if the references can never be loaded,
then they will never be encountered during evaluation.
NoEach and Each list both have this check, but it was missing in
EachMap. Refactor the EachList check to remove a level of indentation,
and make the check consistently near the start of the block.
James Bardin