[jira] [Commented] (TOMEE-2039) Closing Curly Brace Removed from Custom Resource Properties

["Benedikt Schellhaas (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TOMEE-2039?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17202040#comment-17202040 ] 

Benedikt Schellhaas commented on TOMEE-2039:
--------------------------------------------

We recently ran into this issue while trying to configure a MS SQL data source in resources.xml. The MS SQL JDBC driver does [escaping of special chars in the JDBC URL|https://docs.microsoft.com/en-us/sql/connect/jdbc/building-the-connection-url?view=sql-server-ver15#escaping-values-in-the-connection-url] by enclosing them in curly braces {} ... and the closing brace always gets removed, leading to an invalid URL.

Some way to do escaping (or deactivating this "aggressive" behaviour, as proposed above) would be highly appreciated.

> Closing Curly Brace Removed from Custom Resource Properties
> -----------------------------------------------------------
>
>                 Key: TOMEE-2039
>                 URL: https://issues.apache.org/jira/browse/TOMEE-2039
>             Project: TomEE
>          Issue Type: Bug
>    Affects Versions: 7.0.3
>            Reporter: Steve Beatty
>            Priority: Major
>
> In migrating a project from OpenEJB 4.7 to 7, I started getting errors in a custom resource due the '}' character being removed from a property value.  Looking in the source it appears that issue is in org/apache/openejb/util/PropertyPlaceHolderHelper.java which is trying to do variable substitution but is very aggressive about it - all $\{ and } Strings are being removed in the process.  I believe the issue is the simpleValueAsStringOrCharArray method and the snippet below should fix it.
> {code}
> public static Object simpleValueAsStringOrCharArray(final String raw) {
> 	if (raw == null) {
> 		return null;
> 	}
> 	
> 	// This will replace ${ or } if either is encountered instead of matching pair
> 	// I don't know why this is useful
> 	//
> 	//if (!raw.contains(PREFIX) || !raw.contains(SUFFIX)) {
> 	//	return decryptIfNeeded(raw.replace(PREFIX, "").replace(SUFFIX, ""), true);
> 	//}
> 	
> 	if (!raw.contains(PREFIX) && !raw.contains(SUFFIX)) {
> 		return decryptIfNeeded(raw, true);
> 	}
> 	String value = SUBSTITUTOR.replace(raw);
> 	if (!value.equals(raw) && value.startsWith("java:")) {
> 		value = value.substring(5);
> 	}
> 	
> 	// not sure the replaces here are useful either
> 	// if there are still markup characters now then the substitution didn't work
> 	// and it's probably more useful for debugging purposes to see those characters
> 	//
> 	// return decryptIfNeeded(value.replace(PREFIX, "").replace(SUFFIX, ""), true);
> 	return decryptIfNeeded(value, true);
> }
> {code}



--
This message was sent by Atlassian Jira
(v8.3.4#803005)