Propose a fix to #1033 and add a new test for TemplateMatcher #1034
Conversation
…ith an or instead of an and. Now the test is failing, proof that the assertion is checked :)
… a local variable in the same block or by a field (or a variable in another block). First detection implemented but patch not satisfying.
Solve in using version of inria/master
…Show another bug with templateMatcher matching itself.
…ate matcher must find itself to prove it runs correctly.
|
|
||
| scanner.scan(templateRoot); | ||
| if (!finds.contains(templateRoot)) { | ||
| throw new SpoonException("TemplateMatcher was unable to find itself, it certainly express a problem somewhere in the template."); |
There was a problem hiding this comment.
TemplateMatcher was unable to find itself, it certainly indicates a bug. Please revise your template or report an issue.
what do you mean? |
|
so is there a bug behind #1025 ? |
The signature of the method But I don't really see the point of returning a
Frankly I don't know: I tried to show the bug by taking elements from the stackoverflow question to create a test but apparently it works fine: I ask the guy to have a look on my test and to try the last Spoon version to check if it's ok for him... |
good idea |
Fix #1033 by looking if the TemplateMatcher has been detected in the find and remove it if it the case. Moreover this PR add a new dynamic contract by checking before doing any scan, if the TemplateMatcher is able to match itself.
However I still wonder why the
findmethod in TemplateMatcher does return aListand not aSetor at least aCollection?