Token Macro

This plugin adds reusable macro expansion capability for other plugins to use in job configurations.


In various Jenkins plugins, such as email-ext and description-setter, a user is asked to provide an "expression" that evaluates to text. For example, in the case of the email-ext plugin, the whole email content is one big expression, so is the subject. This often includes some kind of tokens, which evaluates to things like build number, cause, the current branch being built, etc. In the description-setter plugin, the user configures a similar expression that evaluates to the string that becomes a build string. In instant messenger plugin, one might configure the template that controls what messages the user will receive.

All these plugins share the common concept of token expansions, where something like ${SUBVERSION_REVISION} expands to some textual values, and since this notion is useful beyond any single use case, it makes sense to define it elsewhere. And this is exactly what this plugin does — to define the mechanism for plugins to define tokens (and their parameters) and their evaluation, and allow this mechanism to be reused by different plugins.

Token Expansion Model

Much of the definition of this is modeled after the email-ext plugin. In the general form, a token can have an arbitrary number of parameters, with multiple values allowed for a single parameter name:

Example Description
${FOO} no parameter
$FOO alternative syntax for token with no parameter
${FOO,param1=value1,param2=value2} 2 parameters
${FOO,param=v1,param=v2,param=v3} 1 parameter with 3 values

Those plugins that wish to define custom tokens can do so by providing implementations of the TokenMacro class that evaluates this into an arbitrary string.

Using Token Expansion

Those plugins that evaluate an expression can call TokenMacro.expand to have all the occurrences of the macros expanded by the available definitions.

String template = "My revision is ${SUBVERSION_REVISION} at #${BUILD_NUMBER}"
String text = TokenMacro.expand( build, listener, template );
System.out.println(text); // something like "My revision is 125 at #12"

You can also ask to expand all macros managed by the Token Macro plugin but also all environment and build variables using TokenMacro.expandAll.

String template = "My revision is ${SUBVERSION_REVISION} at #${BUILD_NUMBER} and was executed on node ${NODE_NAME}"
String text = TokenMacro.expandAll( build, listener, template );
System.out.println(text); // something like "My revision is 125 at #12 and was executed on node Controller"

Databinding Token Macro

The base TokenMacro class only defines a minimalistic contract between the user of the tokens. To simplify the development of custom tokens, a more convenient subtype called DataBoundTokenMacro is provided to simplify the parameter parsing. In this subtype, you get your parameters injected to your instance, so that your evaluate method can access parameter values in a type-safe manner. The following sample is taken from the Git plugin that defines GIT_REVISION macro that expands to the SHA1 commit ID used for a build, with a parameter that controls the length.

For example, this macro can be used like ${GIT_REVISION,length=8} to only show first 8 letters of the commit.

public class GitRevisionTokenMacro extends DataBoundTokenMacro {
    /** Number of chars to use */
    public int length = 40;

    public boolean acceptsMacroName(String macroName) {
        return macroName.equals("GIT_REVISION");

    public String evaluate(Run<?,?> run, FilePath workspace, TaskListener listener, String macroName)
            throws MacroEvaluationException, IOException, InterruptedException {
        BuildData data = run.getAction(BuildData.class);
        if (data == null) {
            return "";  // shall we report an error more explicitly?

        Revision lb = data.getLastBuiltRevision();
        if (lb == null) {
            return "";

        String s = lb.getSha1String();
        return s.substring(0, Math.min(length, s.length()));

Token Transforms

There are a few different transforms that can be applied on top of the result of the token itself. These are similar to BASH shell transforms.

${#MACRO_NAME} // returns the number of characters in the result of evaluating MACRO_NAME.
${MACRO_NAME:<START>[:LENGTH]} // returns the substring starting at START, and optionally specifying the length. You may use negative numbers similar to BASH.
${MACRO_NAME#<NEEDLE>} // returns the value with <NEEDLE> removed, IF it is at the START of the result value.
${MACRO_NAME%<NEEDLE>} // returns the value with <NEEDLE> removed, IF it is at the END of the result value.