WSO2 api context growing too big - api

Question

I have a situation where my api file sizes keeps growing as I keep adding more and more switch mediator cases... what's the best way to split and organize them in wso2. Below is a sample example of one of the files...
The problem is the cases are growing, and the file size is becoming too large.
Sometimes one of the 'CASE' can be like 100 lines! Here in the example, I'm obfuscating by using (...) of course.
I would appreciate any help. Ideally there is a way how we can chunk this into smaller file sizes. It's surprisingly hard to find after googling. I'm pretty sure I'm not the first person to run into this.
Also you can see here that function fixDate(given_time) is being repeated in this file, what's the best way to include this function once and use it throughout the file.
Thanks for any insights!
The file is \synapse-config\api\Applications.xml ... and it's growing in size...
<?xml version="1.0" encoding="UTF-8"?>
<api context="/applications" name="Applications" xmlns="http://ws.apache.org/ns/synapse">
<resource methods="GET">
<inSequence>
<property name="DISABLE_CHUNKING" scope="axis2" type="STRING" value="true"/>
....
...
...
<property expression="$url:page" name="uri.var.page" scope="default" type="STRING"/>
<script language="js"><![CDATA[
...
></script>
<payloadFactory description="PayloadFactory" media-type="json">
<format>
$1
</format>
<args>
<arg evaluator="xml" expression="get-property('uri.var.ats_credentials')"/>
</args>
</payloadFactory>
<property expression="json-eval($.auth_code)" name="uri.var.auth_code" scope="default" type="STRING"/>
<log level="full">
</log>
<switch source="get-property('uri.var.ats_type')">
<case regex="CASE 1">
<payloadFactory media-type="json">
<format>{
"filters": [
{
"name": "applicantworkflow.updateddate",
"value": [
"$1"
],
"secondaryValue": [
"$2"
]
},
{
"name":
"applicantworkflow.id",
"value": [
"$3"
],
"operator": ">"
}
]
.......
<case regex="CASE 2">
<script language="js"><![CDATA[function fixDate(given_time) {
......
<case regex="CASE 3">
<script language="js"><![CDATA[function fixDate(given_time) {
......
<case regex="CASE 4">
<script language="js"><![CDATA[function fixDate(given_time) {
......
//FILE SIZE KEEPS GROWING..... with different cases......
<default>
<log description="Fault Log" level="custom">
<property expression="fn:concat('Invalid ATS - ', get-property('uri.var.ats_type'))" name="message"/>
</log>
<payloadFactory media-type="json">
<format>{
"Error": "Invalid ATS"
}
</format>
<args/>
</payloadFactory>
<respond description="Respond"/>
</default>
</switch>
</inSequence>
<outSequence>
<send/>
</outSequence>
<faultSequence>
<send/>
</faultSequence>
</resource>
<handlers>
<handler class="com.tti.security.OpenSourceAuthHandler"/>
</handlers>
</api>

Answer 1

The easiest solution for this is to use separate sequences. Create a separate sequence for each case (or at least the ones that are getting really big). And from the switch mediator call the sequences using the sequence mediator.
First save the sequence in a sequence file that you add to your project:
<sequence name="case1_sequence" xmlns="http://ws.apache.org/ns/synapse">
<case regex="CASE 1">
<payloadFactory media-type="json">
<format>{
"filters": [
{
"name": "applicantworkflow.updateddate",
"value": [
"$1"
],
"secondaryValue": [
"$2"
]
},
{
"name":
"applicantworkflow.id",
"value": [
"$3"
],
"operator": ">"
}
]
..
</sequence>
Then use the sequence mediator inside your switch statement to call the sequences.
<switch source="get-property('uri.var.ats_type')">
<case regex="CASE 1">
<sequence key="case1_sequence"/>
</case>
..
</switch>