Deploying

Codeflow projects are typically tested from within the designer using the embedded Codeflow engine. However, to deploy projects to production or to run from a command line, the engine can be called directly from a command line.

The engine can be installed directly as an NPM package. The engine should run on any system that runs Node.js version 4.0 and greater.

To begin, install engine as a global NPM package.

npm install -g codeflow

Running engine

After installing the engine, Codeflow commands can be executed from a command line terminal as below:

codeflow [command] <options> <command params>

Commands

Option	Description
run [params]	Command to execute a single flow or the whole project.
install [params]	Command to install dependency packages. Should be run from the root of a Codeflow project. Omit the parameter to install all the dependent packages from command line.
uninstall [params]	Uninstall the specified package from the project. Should be run from the root of a Codeflow project.

Running flows

The run command is used to run a flow or entire project:

codeflow run <options> [path]

Where path can be a path to an individual flow, or the root folder of a Codeflow project. When the folder is passed, the project is started as whole, where all the trigger files inside the project are started by the engine automatically during startup.

Run options

Short option	Long option	Description
-i	--input	Pass a json as the input data. Example: codeflow run -i '{"age":50}' compute.flw
-w	--working-dir	Set the current working directory. Paths passed are resolved relative to the working directory. Example: codeflow run -w /user/project flow.flw
-o	--output	Print the output returned. Example: codeflow run hello.flw -o will print whatever returned by hello.flw to the console
-c	--optimize	Run optimized. If set, this will make Codeflow compile the flows into JavaScript for optimum performance. Example: codeflow run hello.flw -c
-e	--env	Set the environment. Value should be one of dev, qa or prod. This option is used to load the context specific configuration file.
-d	--debug	Set the engine in debug mode. Can pass either ipc or tcp. Defaults to tcp. This is used by Codeflow designer.
-p	--debug-port	Change the default debugger port.

Installing packages

In Codeflow, the installed package dependencies (packages installed under {project-root}/.packages folder) are not be typically added to the version control. Just tracking the package.json that includes the list of dependent packages is enough. The install command can then be used to install the dependent packages from the command line during deployment.

codeflow install <namespace/name@version>,<namespace/name@version>...

Example:

To install all the dependent packages specified in the package.json, simply run install without any parameters at the project's root directory:

cd <project directory>
codeflow install

To install a specific package that may or may not be listed in the package.json, run:

codeflow install codeflow/handlebars 

Specific versions of a package can also be installed by appending @<version> as below:

codeflow install codeflow/handlebars@0.1.1

Uninstalling packages

The uninstall command can be used to remove an unused package from the current project.

codeflow uninstall <namespace/name>,<namespace/name>...

Examples

Use case	Example command(s)
Run a flow	codeflow run hello.flw
Run a flow with an input JSON	codeflow run update.flw -i '{"id": 1, "name": "Jack Sparrow"}'
Run a flow with an input JSON and print the output	codeflow run calculate.flw -i '{"age": 30}' -o
Start as project from the current directory	codeflow run .
Start as project from a specific path	codeflow run . -w /path-to-project
Installing all the dependent packages	codeflow install
Install or update to the latest version of a package	codeflow install codeflow/handlebars
Install or update to a specific version of a package	codeflow install codeflow/handlebars@0.1.10
Uninstall a package	codeflow uninstall codeflow/archiver