The binary of Helm to use for dynamic completion should be the same
as the actual Helm binary being used. For example, if PATH points
to a version of helm v3, but the user calls a binary named helm2 to
use a renamed v2 version, then dynamic completion should also use helm2.
If not, in this example, the dynamic completion will use the
information returned by helm v3.
This improvement is particularly useful for users that will run both
helm v2 and helm v3 at the same time.
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
This cuts down on the number of duplicate json/yaml marshaling blocks and
simplifies how to add printing to a function by using simple "container" structs
that know how to massage the data
Signed-off-by: Taylor Thomas <taylor.thomas@microsoft.com>
unit test adapted for output json or yaml in install_test.go and search_test.go.
docs modified for new flag helm_install.md, helm_repo_list.md, helm_search.md and helm_upgrade.md
Signed-off-by: Michael Schaefer <michael.schaefer@sew-eurodrive.de>
* This was initiated by new versions from Masterminds
* Linting was set to check for the same chart version as the other
checks of this were. >0.0.0-0. The -0 is important to pull in
prereleases. The latest version of semver Validate now finds
errors here properly where some were missed before
* 0.0.0 is a valid semantic version. Instead of expecting it to be
invalid in the tests to check validation now using an invalid
semantic version
Signed-off-by: Matt Farina <matt@mattfarina.com>
When waiting for resources use the `ListWatchUntil` instead of
`UntilWithoutRetry` so that if the connection drops between tiller and
the API while waiting the operation can still succeed.
Signed-off-by: Richard Connon <richard.connon@oracle.com>
The original is explaining the negation, when the body would not be
included. Which would happen in the complement of the if expression,
ie. flipped by De Morgan's law's:
```
not (or .Values.anUnsetVariable (not .Values.aSetVariable))
==
and (not .Values.anUnsetVariable) .Values.aSetVariable
```
> unset variables evaluate to false
is equivalent to `not .Values.anUnsetVariable`.
> and
is equivalent to `and`.
> .Values.setVariable was negated with the not function
doesn't seem to match `.Values.aSetVariable`.
To me, that would be `not .Values.aSetVariable` instead.
Anyway, explaining the `if` expression as-is and not the negation is
clearer and parallels the first `if` operator.
Signed-off-by: Kevin Lau <kelau1993@gmail.com>