Side effect when indicator matrices have the wrong format

If no actual indicator matrices are submitted, but instead strings (e.g. a list of file paths), the web service does not throw an error. Instead, it ignores the submitted L parameter by using the default L parameter and proceeds with the calculations.

Example:
#assuming there is an own webservice running locally
url <- "http://localhost:8080/kpm-web/"

ATTACHED_TO_ID <- paste(sample(c(LETTERS[1:6],0:9),32,replace=TRUE),collapse="")

inputData <- list("/path/to/example/file.csv", "/path/to/second/file.csv")
ws.results <- call.KPM(inputData, ATTACHED_TO_ID, url=url,
async=FALSE, strategy=INES,
algorithm=Greedy, removeBENs=TRUE,
graphID=4, graph.file=NULL,
range=FALSE, linkType=OR,
computed.pathways=20,
Kmin=10,
l_same_percentage = TRUE,
same_percentage=10)