Xeditor: Inferring and Applying XML Consistency Rules
2019
XML files are frequently used in Java programming and when building Web application
implementation. However, it is a challenge to maintain XML files since these files should
follow various domain-specific rules and the existing program analysis tools seldom check
XML files. In this thesis, we introduce a new approach to XML file debugging called Xeditor
that extracts XML consistency rules from open-source projects and uses these rules to detect
XML bugs. To extract the rules, Xeditor first looks at working XML files and finds all the
pairs of entities A and B, which entities coexist in one file and have the same value on at least
one occasion. Then Xeditor will check when A occurs, what is the probability that B also
occurs. If the probability is high enough, Xeditor infers a rule that A is associated with B.
To apply the rule, Xeditor checks XML files with errors. If a file violates the rules that were
previously inferred, Xeditor will report the violation and suggest a change. Our evaluation
shows that Xeditor inferred the correct rules with high precision 83%. More importantly,
Xeditor identified issues in previous versions of XML files, and many of those issues were
fixed by developers in later versions. Therefore, Xeditor is able to help find and fix errors
when developers write their XML files.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
0
Citations
NaN
KQI