Content Management and the Apache Jelly engine

  • Release version: Zurich
  • Updated July 31, 2025
  • 4 minutes to read
  • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Content Management and the Apache Jelly engine

    Apache Jelly is a Java and XML-based scripting engine used within ServiceNow to dynamically render UI components such as forms, lists, and UI Pages. Its syntax resembles XML and supports developers familiar with JavaScript, XML, XHTML, or HTML. Jelly scripts are primarily used within dynamic content blocks, with some limitations in static blocks.

    Show full answer Show less

    Jelly scripting in ServiceNow uses multiple namespaces and supports two parsing phases, enabling efficient script execution and caching. The engine processes native Jelly tags prefixed with j, ServiceNow-specific tags prefixed with g, and their corresponding phase two counterparts j2 and g2. Selecting the two-phase option on content forms is necessary when using phase two tags.

    Key Features

    • Two-phase parsing: Jelly scripts run in two phases, improving performance by caching phase one results and executing phase two on subsequent calls.
    • Namespaces and prefixes: The j and g prefixes represent native and ServiceNow-specific tags, respectively, while j2 and g2 are their second phase equivalents.
    • Dynamic content rendering: Jelly scripts can embed XML and HTML seamlessly, supporting complex UI logic and data display.
    • Integration with Glide API: Jelly supports executing server-side scripts via g:evaluate tags, enabling access to GlideRecord queries and other platform APIs.
    • Output escaping: Expressions like ${HTML:variable} ensure safe rendering by escaping output according to context (e.g., HTML, JavaScript).

    Practical Usage

    ServiceNow customers can leverage Jelly scripting to create customized content blocks, reports, and dynamic UI elements. For example, you can query records such as user groups and incidents using GlideRecord within a g:evaluate script block, loop through results using Jelly control tags like j:while, and render results in HTML tables.

    When working with Jelly scripts:

    • Always include the root Jelly tag with the correct namespaces (j:jelly trim="false" xmlns:j="jelly:core" xmlns:g="glide" xmlns:j2="null" xmlns:g2="null") at the start.
    • Ensure proper closing of tags, including self-closing tags like <br /> and <img />.
    • Use the two-phase option on content forms if utilizing phase two tags to enable deferred parsing and caching.
    • Understand output escaping to securely display data and prevent injection issues.

    Key Outcomes

    • Enable creation of flexible, data-driven UI components and reports by combining Jelly scripting with Glide API access.
    • Improve script performance by leveraging the two-phase parsing mechanism.
    • Maintain secure and well-formed HTML output through explicit escaping and proper tag usage.
    • Customize content rendering dynamically based on page context and data conditions.

    Apache Jelly is a Java-based and XML-based scripting and processing engine for turning XML into executable code.

    The Apache Jelly engine closely resembles XML and should be comfortable for developers familiar with JavaScript, XML, XHTML, or HTML. In the ServiceNow instance, the Apache Jelly engine renders items such as forms, lists, and UI Pages. Apache Jelly code renders well within a dynamic content block, but can have issues when used in static blocks. You can use Jelly tags, calls, and statements, but HTML acts just like XHTML.

    <?xml version= "1.0" encoding= "utf-8" ?>
    <j:jelly trim = "false" xmlns:j = "jelly:core" xmlns:g = "glide" xmlns:j2 = "null" xmlns:g2 = "null" >
     
    	<j:if test = "${current_page.getName()=='Solutions'}" >
    	<h1 class = "page_name" > <b> <a href = "solutions.do?" title = "${gs.getMessage('Solutions')}" >${gs.getMessage('Solutions')}</a> </b> </h1>
    	<p class = "page_description" >
    	 	    ${current_page.getDescription()}
    	</p> <br />
    	</j:if>
    	<j:if test = "${current_page.getName()=='IT 3.0'}" >
    	   <h1 class = "page_name" > <b> <a href = "solutions.do?" title = "${gs.getMessage('Solutions')}" >${gs.getMessage('Solutions')}</a> </b> | ${current_page.getName()}</h1>
    	   <p class = "page_description" >
    	 	    ${current_page.getDescription()}
    		</p> <br />
    		</j:if>
    		</j:jelly>

    Ensure that all tags are closed. If the tag is not a naturally closing tag, then place a forward slash before the end bracket. For example, a <BR /> or an <IMG src="cms.png" />.

    If you are unfamiliar with Jelly scripting, review these topics:

    Include the following tag with all Apache Jelly scripts.

    <j:jelly trim= "false" xmlns:j= "jelly:core" xmlns:g= "glide" xmlns:j2= "null" xmlns:g2= "null" >
    The tag looks complex, but keep the following information in mind.
    • Apache Jelly script uses multiple namespaces.
    • There are two types of prefixes in tags: j and g. The j prefix is used for tags that are natively part of Apache Jelly. The g prefix is used for tags that the ServiceNow platform created and is using for platform purposes.

    The j2 and g2 prefixes are just like j and g, except that they are processed in a second phase. The Apache Jelly script parser runs through each j and g tag respectively. For example:

    <j:jelly trim="false" xmlns:j="jelly:core" xmlns:g="glide" xmlns:j2="null" xmlns:g2="null">
    	<j:set var="jvar_phase1" value="Hello" />
    	<j2:set var="jvar_phase2" value="World" />
    	${jvar_phase1} $[jvar_phase2]
    </j:jelly>

    In phase 1, the parser runs through all the j and g tags. It then caches the result. Before it runs the second phase, it takes the j and g namespaces and moves the namespaces to the second phase. It looks something like the following code.

    <j:jelly trim="false" xmlns:j="jelly:core" xmlns:g="glide" xmlns:j2="jelly:core" xmlns:g2="glide">
    	<j2:set var="jvar_phase2" value="World" />
    	Hello $[jvar_phase2]
    </j:jelly>
    For subsequent calls of this script, only phase 2 is parsed.
    Note:
    If you plan to use phase 2 Jelly tags (g2 and j2) on the Content Type [content_type] or Dynamic Content [content_block_programmatic] tables, select the Two phase option on the content form.

    Another example is to create a report of all open incidents assigned to each group. For this purpose, you could use a report and save time, but it is a good example for learning Jelly. Start with the Jelly tag:

    <j:jelly trim= "false" xmlns:j= "jelly:core" xmlns:g= "glide" xmlns:j2= "null" xmlns:g2= "null" >
    </j:jelly>

    First, you need a list of open incidents. Use a g2:evaluate tag. The evaluate tag runs the script. Anything inside the tag is parsed like a business rule, so, for example, you can call global business rules, script includes, and gliderecord.

    <j:jelly trim="false" xmlns:j="jelly:core" xmlns:g="glide" xmlns:j2="null" xmlns:g2="null">
    	<g:evaluate var="jvar_groups" object="true">
    		var now_GR = new GlideRecord("sys_user_group");
    		gr.orderBy('name');
    		gr.query();
    		gr;
    	</g:evaluate>
    </j:jelly>

    This script is in phase 1 because frequent changes to incident assignment groups are not expected. Also notice the var attribute on the evaluate tag. This attribute specifies what variable is set from this block. At the end of the script, there is a gr on a line by itself. That last line is what sets the variable.

    You can omit the jvar_groups variable, but then all the variables in the evaluate tag become Apache Jelly variables. The object=true specifies that the variable is not a primitive data type. If object=true is omitted, the script would break because jvar_groups would only be able to hold items like integers and strings.

    After the evaluate tag, loop through these groups and find the incidents for each one.

    <j:jelly trim="false" xmlns:j="jelly:core" xmlns:g="glide" xmlns:j2="null" xmlns:g2="null">
    	<g:evaluate var="jvar_groups" object="true">
    		var now_GR = new GlideRecord("sys_user_group");
    		gr.orderBy('name');
    		gr.query();
    		gr;
    	</g:evaluate>
    	<table>
    		<tr>
    			<th>Name</th>
    			<th>Incidents</th>
    		</tr>
    		<j:while test="${jvar_groups.next()}">
    			<tr>
    				<td>${HTML:jvar_groups.getValue('name')}</td>
    				<td></td>
    			</tr>
    		</j:while>
    	</table>
    </j:jelly>
    You can include normal XML in the Apache Jelly script at any time. Since there is no namespace, the Apache Jelly script does not try to parse the XML tags. Notice the j:while loop. It is a normal while loop and can iterate through a GlideRecord object. Also notice that you output a value with ${HTML:jvar_groups.getValue('name')}. Here are the important elements:
    • The outer brackets, ${}, specify the output of the variable and the phase in which the variable is output: ${} means first phase, $[] means second phase.
    • HTML before the expression is for escaping the output. The expression jvar_groups.getValue('name') is being escaped for HTML. For other types of escaping, there are JS (Javascript), NS (No Script), and some other options.

    To select only one record and not iterate through many records, the code looks like the following example:

    <j:jelly trim="false" xmlns:j="jelly:core" xmlns:g="glide" xmlns:j2="null" xmlns:g2="null">
    	<g:evaluate var="jvar_groups" object="true">
    		var now_GR = new GlideRecord("sys_user_group");
    		gr.orderBy('name');
    		gr.query();
    		gr;
    	</g:evaluate>
    	<j:if test="${jvar_groups.next()}">
    		We found ${HTML:jvar_groups.getValue('name')}
    	</j:if>
    </j:jelly>