Aviator is a tool to template & merge YAML files in a convenient fashion based on a configuration file called aviator.yml
. Aviator utilizes Spruce for templating and merging and therefore enables you to use all the Spruce operators in your YAML files.
If you have to handle rather complex YAML files (for Kubernetes, Concourse, or Bosh), you just provide the flight plan (aviator.yml
), the Aviator flies you there.
Reads:
$ wget -O /usr/local/bin/aviator https://github.com/JulzDiverse/aviator/releases/download/v1.9.0/aviator-darwin-amd64 && chmod +x /usr/local/bin/aviator
Via Homebrew
$ brew tap julzdiverse/tools
$ brew install aviator
$ wget -O /usr/bin/aviator https://github.com/JulzDiverse/aviator/releases/download/v1.9.0/aviator-linux-amd64 && chmod +x /usr/bin/aviator
https://github.com/JulzDiverse/aviator/releases/download/v1.9.0/aviator-win
To run Aviator navigate to a directory that contains an aviator.yml
and run:
$ aviator
OR
Specify an AVIATOR YAML file with the [--file|-f] option:
$ aviator -f myAviatorFile.yml
Aviator provides a verbose style of configuration. It is the result of configuring a spruce merge plan and optionally an execution plan (e.g fly
).
Example for a simple aviator file:
spruce:
- base: path/to/base.yml
merge:
- with:
files:
- top.yml
to: result.yml
The spruce
section is an array of merge steps. It provides different parameters to provide high flexibility when merging YAML files. You can:
- specify specific files to include into your merge
- specify a specific directory to include into your merge
- specify a specific directory including all subdirectories to include into your merge
However, this is not enough. Additionally you can use regular expressions, environment-variables, and more. Read about all parameters and what they do in this section.
The base
property specifies the path to the base YAML file. All other YAML files will be merged on top of this YAML file.
prune
defines YAML properties which will be pruned during the merge. For more information check the spruce
merge semantics.
Example:
spruce:
- base: base.yml
prune:
- meta
- properties
merge:
- with:
files:
- top.yml
to: result.yml
In this case meta
and properties
will be pruned during merge.
Enables Spruce cherry pick
option: With the cherry_pick
property you can specify specific YAML subtrees you want to have in your restulting YAML file (opposite of prune
)
Example:
spruce:
- base: path/to/base.yml
cherry_pick:
- properties
merge:
- with_in: path/to/dir/
- with:
files:
- top.yml
regexp: ".*.(yml)"
skip_eval: true
to: result.yml
To use spruce in conjuction with the go-patch
format it can be enabled within the aviator spruce
section as a toplevel bool property:
spruce:
- base: some.yml
go_patch: true
merge:
- with:
files:
- some/ops/file.yml
to: result.yml
Read more about it here
You can configure three different merge types inside the merge
section: with
, with_in
, with_all_in
:
with
with
specifies specific files you want to include into the merge.
-
files
(required): List of paths to YAML files -
in_dir
(optional): If all of the files you want to include into the merge are in one specific directory, you can specify the directoyr path and list only file names in thefiles
list. Note: Whenever a directory is defined, the path requires a trailing "/"!!! -
skip_non_existing
(optional): Setting this property totrue
will skip non existing files that are specified in thefiles
list rather then returning an error. This is useful, if a file is not necessarely there.
Example:
spruce:
- base: path/to/base.yml
merge:
- with:
files:
- top.yml
- top2.yml
- top3.yml
in_dir: path/to/
skip_non_existing: true
to: result.yml
with_in (string
)
with_in
specifies a path (do not forget the trailing "/") to a directory. All files within this directory (but not subdirectories) will be included in the merge.
Example:
spruce:
- base: path/to/base.yml
merge:
- with_in: path/to/dir/
to: result.yml
except
(array
)
With except
you can specify a list of files you want to exclude from the path specified in with_in
Example:
spruce:
- base: path/to/base.yml
merge:
- with_in: path/to/dir/
except:
- file1
- file2
to: result.yml
This will exclude path/to/dir/file1
and path/to/dir/file2
from the merge.
with_all_in
with_all_in
specifies a path (do not forget the trailing "/") to a directory. All files within this directory -including all subdirectories - will be included in the merge.
Example:
spruce:
- base: path/to/base.yml
merge:
- with_all_in: path/to/dir/
except:
- someFiles.yml
- youWant.yml
- toExclude.yml
to: result.yml
NOTE: except
also works for with_all_in
regexp (string
(quoted))
Only files matching the regular expression will be included in the merge. It can be specified for all three merge types with
, with_in
, and with_all_in
. This could be required if the target directory contains other then only YAML files.
Example:
spruce:
- base: path/to/base.yml
merge:
- with_in: path/to/dir/
regexp: ".*.(yml)"
- with:
files:
- top.yml
regexp: ".*.(yml)"
- with_all_in: path/to/another/dir/
regexp: ".*.(yml)"
to: result.yml
Enabling this skip-eval will merge without resolve spruce expressions. For more information check Spruce doc
Example:
spruce:
- base: path/to/base.yml
merge:
- with_in: path/to/dir/
- with:
files:
- top.yml
regexp: ".*.(yml)"
skip_eval: true
to: result.yml
to
specifies the target file, where the merged files should be saved to. It can be used only in combination with the basic merge types files
, with_in
, and with_all_in
.
On top of the basic merge
you can do more complex merges with for_each
. More precisely, you can execute the basic merge
for multiple files specified in for_each
. When specifying files with for_each
you need to use to_dir
instead of to
to specify a target directory instead of a target file.
files
files
specifies a list of files that will be included in your merge seperately.
Example:
spruce:
- base: path/to/base.yml
merge:
- with:
files:
- top.yml
regexp: ".*.(yml)"
for_each:
files:
- env.yml
- env2.yml
to_dir: results/
This merge step will execute two merges and generate two files. It will merge base.yml
and top.yml
with env.yml
, write it to results/
and do the same with env2.yml
.
in
in
is basically the same as files
with the difference that it will merge all files for a given path sperately
Example:
spruce:
- base: path/to/base.yml
merge:
- with:
files:
- top.yml
regexp: ".*.(yml)"
for_each:
in: path/to/dir/
to_dir: results/
Except
except
works in combination with in
: list of files that you want to exclude from the merge.
Example:
spruce:
- base: path/to/base.yml
merge:
- with:
files:
- top.yml
regexp: ".*.(yml)"
for_each_in: path/to/dir/
except:
- some.yml
to_dir: results/
include_sub_dirs
include_sub_dirs
includes all files including files in all subdirectories of a directory into the merge seperately.
spruce:
- base: path/to/base.yml
merge:
- with:
files:
- top.yml
regexp: ".*.(yml)"
for_each:
in: path/to/dir
include_sub_dirs: true
enable_matching: true
copy_parents: true
to_dir: results/
When include_sub_dirs
is defined you can specify further properties:
-
enable_matching
: this will only include files in the merge, that contains the same substring as the parent directory. -
copy_parents
: setting this property totrue
(defaultfalse
) will copy the parent folder of a file to the target directory (in the above exampleresults/
)
regexp
The regexp
property can also be set in combination with for_each
, for_each_in
, and walk_through
to only include files matching the regular expression.
spruce:
- base: path/to/base.yml
merge:
- with:
files:
- top.yml
regexp: ".*.(yml)"
for_each:
in: path/to/dir
include_sub_dirs: true
enable_matching: true
copy_parents: true
regexp: ".*.(yml)"
to_dir: results/
Sometimes it is required to do more than one merge step, which creates intermediate YAML files. In this case you can save merge results to internal datastore/cache which you can write/read by surrounding your location with double courly braces {{file|dir}}
. Internal cache also work as directories and can be used with to_dir
.
Example:
spruce:
- base: path/to/base.yml
merge:
- with_in: path/to/dir/
to: {{result}}
- base: {{result}}
merge:
- with_in: another/path/
to: final.yml
Aviator supports to read Environment Variables. Environment variables can be set with $VAR
or ${VAR}
at an arbitrary place in the aviator.yml
.
Example:
spruce:
- base: $BASE_PATH/app-${NUMBER}.yml
merge:
- with_in: path/to/dir/
to: {{result}}
- base: {{result}}
merge:
- with_in: $TARGET_PATH
to: $RESULT_YAML
Executing aviator
as follows:
$ BASE_PATH=/tmp/ NUMBER=1 RESULT_YAML=result.yml aviator
will resolve:
spruce:
- base: /tmp/app-1.yml
merge:
- with_in: path/to/dir/
to: {{result}}
- base: {{result}}
merge:
- with_in: $TARGET_PATH
to: result.yml
You can provide variables to aviator files using the --var
flag. Basic CLI usage:
$ aviator --var key=value
In you aviator file you need to specify the name of the variable you want to interpolate. The syntax for a variable is the following (( varName ))
(note the space before and after the variable name!)
Example:
---
spruce:
- base: (( key ))
...
In this example the variable key
will be replaced by the value value
. So the result would look like this:
---
spruce:
- base: value
...
Values for aviator variables can be multi-line
With modifier you can modify the resulting (merged) YAML file. You can either delete, set, or update a property. The modifier will always be applied on the result. If you use for_each
it will be applied on each for_each
merge step.
Consider a resulting YAML from a merge process result.yml
, which has a property person.name
:
---
person:
name: Julz
- the property can be deleted
spruce:
- base: base.yml
merge:
- with:
files:
- top.yml
modify:
delete:
- "person.name"
to: result.yml
It deletes a property only if it exists. There will be no error if a proerty does NOT exist.
- the property can be updated:
spruce:
- base: base.yml
merge:
- with:
files:
- top.yml
modify:
update:
- path: person.name
value: newName
to: result.yml
Using update will update existing properties only.
- Other properties can be added/updated with set:
spruce:
- base: base.yml
merge:
- with:
files:
- top.yml
modify:
set:
- path: person.name
value: NewName
to: result.yml
Set updates or adds a property to an array. If a property exists it will be overwritten, if the property does not exist it will be added (works only for maps not arrays).
Aviator uses goml as YAML modifier. If you want to read more about update
, delete
, and set
, check the README.
You can squash multiple files into one single YAML file using the squash
section.
squash:
contents:
- files:
- deployment.yml
- service.yml
to: app.yml
squash:
contents:
- dir: my/dir/
to: app.yml
Executors execute executables installed on the OS that Aviator is running on. The following three executors are currently supported by Aviator:
kubectl
Executorfly
Executor- Generic Executor: Runs an any specified executable.
Execute kubectl apply
with the following options:
Supported flags
- file: Name of the file to
kubectl apply
- force: calls
kubectl apply
with the--force
flag - dry_run: calls
kubectl apply
with the--dry-run
flag - overwrite: calls
kubectl apply
with the--overwrite
flag - recursive: calls the
kubectl apply
with the--recursive
flag - output: calls the
kubectl apply
with the--output=<desired-ouput>
parameter - kustomize: calls the
kubectl apply
with the--kustomazation/-k
flag rather than with--filename/-f
.
You can read about the details of the flags of kubectl apply
here
Example:
kubectl
apply:
file: deployment.yml
force: true
dry_run: true
overwrite: true
recursive: true
output: yaml
Kustomize Example:
When kustomize
flag in aviator is set to true it will call kubectl apply
with the --kustomization/-k
flag, which expects a directory path. The directory paht needs to be specified in the file
property of kubectl.apply
:
kubectl:
apply:
file: kustomization/path/
kustomize: true
An executor for the Concourse Fly CLI. The supported commands are set-pipeline
, validate-pipeline
, format-pipeline
, and expose-pipeline/hide-pipeline
.
The set-pipeline
and hide-pipeline
commands are executed by default. Here is the list of commands and flags that can be used with aviator
:
- name: Name of the pipeline
- target: Target short name (
fly
target) - config (string): the pipeline config file (yml)
- load_vars_from (array): List of all property files (-l)
- vars (map): Map of variables (--var)
- non_interactive (bool): Enables non-interactive mode (-n)
- expose (bool): Exposes the pipeline (expose-pipeline)
- validate_pipeline (bool): Validate local pipeline configuration (failes on errors)
- strict (bool): causes
validate_pipeline
to fail on errors AND warnings - format_pipeline (bool): format a pipeline config in a "canonical" form (prints to stdout).
- write: update the formatted pipeline config file in-place.
- team_name (string): the team name to fly for set-pipeline
More detailed description of the flags can be found here
Example - set-pipeline:
fly:
name: myPipelineName
target: myFlyTarget
team_name: concourse-team
config: pipeline.yml
non_interactive: true
check_creds: true
load_vars_from:
- credentials.yml
vars:
var1: myvar
var2: myvar2
expose: true
Example - validate-pipeline:
fly:
config: myconfig.yml
validate_pipeline: true
strict: true
Example - format-pipeline:
fly:
config: myconfig.yml
format_pipeline: true
write: true
NOTE: You will need to fly login first, before executing aviator
The Generic Executor executes any specified executable. Here is how to define an Generic Executor in the aviator.yml
:
exec:
- executable: some-executable # required
global_options: # optional
- name: --option
value: option-value
command: # optional
name: command-name
options:
- name: --command-option
value: command-option-value
args:
- arg1
- arg2
This will executed as $ some-executable --option option-value command-name --command-option command-option-value arg1 arg2
Example: cp
exec:
- executable: cp
global_options:
- name: -r
args:
- dir/
- destination/
This calls cp
as follows: $ cp -r dir/ destination/
Some YAML based tools (like concourse in the past) are using {{}}
sytnax. This is not YAML conform. Using the --curly-braces
option you can allow this syntax.
This option will output no infromation to stdout.
This option prints which files are excluded from a merge.
This option prints contents to stdout
rather than writing it to files. This flag also omits any defined executor.
You can provide variables to the aviator file.
$ go get github.com/JulzDiverse/aviator
To build the aviator binary you can navigate to cmd/aviator
:
$ cd ./cmd/aviator`
Now you can build with the go cli:
$ go build
For more information on the $ go build
command run $ go help build
.