The original Greek word "model" means "misshapen ball of clay", and I try to think about that every time I go in front of the camera. - Derek Zoolander
Clay is a template focused code generator. Used to take a model represented in JSON and konvert it to actual code matching the model through a set of generators using handlebar templates, shell commands and file operations.
The goal is to have a consistent implementation of all model related operations in a system. Similar to how Ruby On Rails scaffolds work but backed using files, so you can with confident change the model and have all files updated.
I use clay as a globally installed commandline tool.
> npm install -g clay-generator
Running clay will display its available commands.
> clay
Usage: clay [options] [command]
Options:
-V, --version output the version number
-v, --verbose ignore test hook
-h, --help output usage information
Commands:
test-path <model_path> <json_path> test a json-path selector using your model
clean <model_path> <output_path> cleans up the output of the generators
generate <model_path> <output_path> runs the generators
watch <model_path> <output_path> runs the generators on filechanges in the models directory
The first command to know in clay is the generate command.
> clay generate <model_path> <output_path>
It takes the model to generate as the first input and the second input is where the model should be generated too.
The test path command will allow you to test jsonpath against your specified model. This is usefull when you are creating generators and templates. To see what you will get from a model given a specific path.
The command will simply print the resulting model objects your path returns.
> clay test-path <model_path> $.model.types[*].actions[*].parameters[?\(@.type==\"array\"\)]
The clean command will clear up all generated code based on the generators specified in the
model.
It will not undo any commandline executions that has ben run by a generator.
> clay clean <model_path> <output_path>
If you start using clay extensivly you will notice that you want to rerun clay as soon as you make changes to the model or any of the custom generators you might be using.
For this there is a built in watch command that follows the same structure as the generate command.
> clay <model_path> <output_path>
Using this will make clay listen for changes under the directory specified in the model.
We have a directory called clay where we keep our model and our custom generators.
.
├── dataaccess
│ ├── generator.json
│ └── template
│ └── db.js
├── frontend
│ ├── components
│ │ ├── {{pascalCase\ name}}Components.js
│ ├── generator.json
│ ├── partials
│ │ └── form-fields.js
│ └── reducers
│ └── {{kebabCase\ name}}.reducer.js
├── mixins
│ └── default_actions.mixin.js
└── model.json
This makes it simple to run clay watch on the "clay directory" and work continously on the generators and the model.
Clay takes a structured specification of a domain model and generates code based on the domain model and a given set of generators.
Clay tries to stay clear of exactly how the domain model i structured, but in order to successfully generate from the model there needs to be some structure.
The model can have includes to split it up into several files.
There is also something called mixins that allow you to make modifications to models using simple js functions. These are usefull for convention based model properties.
The model holds a list of generators to be run for this specific model.
Generators holds a list of steps to run. Steps can do one of the following:
- Generate a set of files using a handlebar template and the model
- Run commands using the model and handlebar to modify executed commands
- Copy existing files
the order of execution is defined in the generator
Clays models are described in a file called models.json.
The structure for describing a model in clay is as follows:
-
A model contains:
-
a list of generators.
-
a name
-
a list of mixins
-
the model as a json structure
mixin and include are reserved properties on a clay model.
mixin is a list of functions to execute on a part of the model
include will replace the current part of the model with the content if a
specified file.
Example of a models.json file:
{
name: 'mymodel'
generators: [
'documentation',
'backend.json'
]
mixins: [
'has_created.js',
'has_error.js'
]
model: {
events: [
{include: 'standardevents.json'}
],
types:[
{
name: 'order',
mixin: [has_created, has_errors],
commands: [
{
name: 'finish_order',
raise: 'order_finished',
parameters: [
{
name: 'finished'
}
]
}
],
fields: [
{
name: 'test'
}
]
}
]
}
}
Generators describe a sequence of steps that will be taken with the model as input. Generators are generally installed globaly so they can be used by clay.
a generator contains:
- a list of steps
- a list of partials
- a list of formatters
Currently available generators:
- (Clay Model Generator)[https://github.com/morkeleb/clay-model-documentation]
The partials are handlebar partials that can be used in the templates within the generator. These are usefull for header and footer like operations.
The steps describe what the generator does and in which order. Steps can be one of:
- Copy
- generate handlebar template
- run a command
Steps use jsonpath to filter out what part of the model that is interesting for a particular step.
To ensure the generated files are as pretty as possible, Clay has support for external formatters.
Formatters are used to format the output before it is written to disk. For example it could be to ensure propper indentation before writing the file. Formatters are installed globally and the generator will require the specified formatter.
Example:
{
"partials":[],
"formatters:["clay-generator-formatter-prettier"],
"steps":[
{
"runCommand": "yo newservice"
},
{
"generate": "templates/html",
"select": "jsonpath statement"
},
{
"runCommand": "yo servicepart {{service.name}}",
"select": "jsonpath statement"
},
{
"copy": "foundation",
"select": "jsonpath statement",
"target": "{{microservice}}"
}
]
}
Clay uses the handlebars template engine to generate a files.
The paths are also handlebar template, allowing you to use values from the model or view data to write the file to the correct location.
Example of a template:
var command = function (user, {{ parameters }}) {
var {{lowerCase type.name}} = repository.find('{{type.name}}', id);
if({{lowerCase type.name}}){
eventsource.raise('{{raises}}', [user, {{parameters}}]);
}
}
It's tedious to rewrite parts of code even in template files. To help with that clay include all partials listed in a generator.
in the domain-documentation.html template I use the header partial as shown:
{{>header }}
<h1>{{name}}</h1>
This helper is very usefull when creating templates. It will pretty print the json input given to it. Allowing you to check what objects you have received to the template as you were generating.
<p>
{{{json this}}}
</p>
The markdown helper method shown in the following example.
<p>
{{markdown description}}
</p>
lobars and lodash string manipulation as helpers to the handlebar templating system.
var command = function (user, {{ parameters }}) {
var {{lowerCase type.name}} = repository.find('{{type.name}}', id);
if({{lowerCase type.name}}){
eventsource.raise('{{raises}}', l[user, {{parameters}}]);
}
}
- support casing help in helpers
- inc one helper for indexes plus 1
- use chalk to make pretty output... pretty output is... pretty
- update generator and allow loading of node modules as generators
- built in watch support
- add usage instructions to readme
- clean up command that will remove files instead of writing them
- add a .clay file that keeps inventory of what has been generated
- make sure regeneration of files only write changes
- make clean command work with the .clay file
- make the generate command work with the .clay file so it doesnt require arguments
- make the clean command work with the .clay file so it doesnt require arguments
- make the watch command work with the .clay file so it doesnt require arguments
- validations on models and generators
- add an option to make dry runs
- tests on handlebar templating system to prevent regressions and broken generators
- option to clear directories