[jira] [Commented] (CXF-5719) NoSuchElementException in ClientFaultConverter when stack trace message contains * or #

["Alban Deconinck (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CXF-5719?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13985329#comment-13985329 ] 

Alban Deconinck commented on CXF-5719:
--------------------------------------

an example to show you why StringTokenizer is not relevant :

{code:title=Test.java|borderStyle=solid}
import java.util.StringTokenizer;
import java.util.regex.Pattern;

public class Test {

	// precompiled (even if not compiled...) regex (to avoid String.split regex compilation)
	static final Pattern	causeSuffixSplitter	= Pattern.compile("#*#"/* Message.EXCEPTION_CAUSE_SUFFIX */,
													Pattern.LITERAL | Pattern.MULTILINE);

	public static void main(String[] args) {
		System.out.println("wrong way");
		String s = "this\nis#sparta#*#oh yeah!";
		StringTokenizer st = new StringTokenizer(s, "#*#");// same as "#*"
		while (st.hasMoreElements()) {
			System.out.println("-----\n" + st.nextElement());
		}

		System.out.println("\nregex way");
		for (String nextElement : causeSuffixSplitter.split(s)) {
			System.out.println("-----\n" + nextElement);
		}

		// out put in console :

		// "I did not read JDoc" way
		// -----
		// this
		// is
		// ----- <========== should not be splitted in our case !
		// sparta
		// -----
		// oh yeah!
		//
		// regex way (avoid : (parse all String in 1 time)
		// -----
		// this
		// is#sparta
		// -----
		// oh yeah!

	}
}
{{code}}

> NoSuchElementException in ClientFaultConverter when stack trace message contains * or #
> ---------------------------------------------------------------------------------------
>
>                 Key: CXF-5719
>                 URL: https://issues.apache.org/jira/browse/CXF-5719
>             Project: CXF
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 2.6.10, 2.7.7, 3.0.0-milestone1
>            Reporter: Saad Benbouzid
>              Labels: exception-handling, exceptions, soapfault, stacktrace
>   Original Estimate: 24h
>  Remaining Estimate: 24h
>
> Following stack trace shows up whenever there is a # or a * character in one of the 'Caused By:' clause.
> {code}
> Caused by: java.util.NoSuchElementException
> 	at java.util.StringTokenizer.nextToken(StringTokenizer.java:349)
> 	at org.apache.cxf.interceptor.ClientFaultConverter.parseStackTrackLine(ClientFaultConverter.java:287)
> 	at org.apache.cxf.interceptor.ClientFaultConverter.getCause(ClientFaultConverter.java:278)
> 	at org.apache.cxf.interceptor.ClientFaultConverter.setStackTrace(ClientFaultConverter.java:246)
> 	at org.apache.cxf.interceptor.ClientFaultConverter.handleMessage(ClientFaultConverter.java:79)
> 	at org.apache.cxf.phase.PhaseInterceptorChain.doIntercept(PhaseInterceptorChain.java:272)
> 	at org.apache.cxf.interceptor.AbstractFaultChainInitiatorObserver.onMessage(AbstractFaultChainInitiatorObserver.java:113)
> 	at org.apache.cxf.binding.soap.interceptor.CheckFaultInterceptor.handleMessage(CheckFaultInterceptor.java:69)
> 	at org.apache.cxf.binding.soap.interceptor.CheckFaultInterceptor.handleMessage(CheckFaultInterceptor.java:34)
> 	at org.apache.cxf.phase.PhaseInterceptorChain.doIntercept(PhaseInterceptorChain.java:272)
> 	at org.apache.cxf.endpoint.ClientImpl.onMessage(ClientImpl.java:845)
> {code}
> For example : 
> {code}
> org.springframework.jdbc.UncategorizedSQLException: Error setting null for parameter #1 with JdbcType OTHER . Try setting a different JdbcType for this parameter or a different jdbcTypeForNull configuration property. Cause: java.sql.SQLException: Type de colonne non valide: 1111
> ; uncategorized SQLException for SQL []; SQL state [99999]; error code [17004]; Type de colonne non valide: 1111; nested exception is java.sql.SQLException: Type de colonne non valide: 1111
> 	at org.springframework.jdbc.support.AbstractFallbackSQLExceptionTranslator.translate(AbstractFallbackSQLExceptionTranslator.java:83)
> 	at org.springframework.jdbc.support.AbstractFallbackSQLExceptionTranslator.translate(AbstractFallbackSQLExceptionTranslator.java:80)
> 	at org.springframework.jdbc.support.AbstractFallbackSQLExceptionTranslator.translate(AbstractFallbackSQLExceptionTranslator.java:80)
> 	at org.mybatis.spring.MyBatisExceptionTranslator.translateExceptionIfPossible(MyBatisExceptionTranslator.java:73)
> 	at org.mybatis.spring.SqlSessionTemplate$SqlSessionInterceptor.invoke(SqlSessionTemplate.java:371)
> 	at com.sun.proxy.$Proxy74.selectOne(Unknown Source)
> {code}
> Because it considers the # (sharp) character as a delimiter.
> That means actual CSF stack trace deserialization does not work for any 'Caused By' clause with a sharp (#) or a star character (#)... which is really unpleasant for customization (see stack trace Apache MyBatis stack trace example above).
> Please avoid *StringTokenizer* splitting in favor of a  *String.split(...)* or a plain old iterator. Because *StringTokenizer* considers not only *Message.EXCEPTION_CAUSE_SUFFIX* as a delimiter, but also each character in it, which are sharp # and star *.
> {quote}
> StringTokenizer is a legacy class that is retained for compatibility reasons although its use is discouraged in new code. It is recommended that anyone seeking this functionality use the split method of String or the java.util.regex package instead.
> {quote}
> {color:red}
> *org.apache.cxf.interceptor.ClientFaultConverter._setStackTrace_*
> {color}
> also see CXF-5231



--
This message was sent by Atlassian JIRA
(v6.2#6252)