better stationxml vs seed detection

[crotwell]

See #95

I think the input file detection still does not work the way it should. This is a StationXML validator, and so the default case should be to parse a StationXML file, not to parse a SEED file.

Currently the file reading looks like:

if(isStationXml(file)) {
...read station xml file
} else {
...read as SEED file
}

But, the else has no checking is done to see if the file is a SEED file. An xml file that has a mistake in the root xml element will be processed as if it is a SEED volume giving hard to understand errors.

For example if the root element is:

<TypoFDSNStationXML xmlns="http://www.fdsn.org/xml/station/1" schemaVersion="1.1">

then the error generated by the validator looks like:

[2021-07-28 09:36:13] [SEVERE] edu.iris.dmc.Application exitWithError: java.io.IOException: Error calculation record length!edu.iris.dmc.seed.blockette.util.BlocketteItrator.hasNext(BlocketteItrator.java:58)
edu.iris.dmc.IrisUtil.readSeed(IrisUtil.java:62)
edu.iris.dmc.IrisUtil.readSeed(IrisUtil.java:54)
edu.iris.dmc.Application.read(Application.java:214)
edu.iris.dmc.Application.run(Application.java:178)
edu.iris.dmc.Application.run(Application.java:168)
edu.iris.dmc.Application.run(Application.java:147)
edu.iris.dmc.Application.main(Application.java:118)

which tells you nothing about why the file is not valid.

Would be better if the code looked like:

if(isXml(file)) {
    if(isStationXml(file)) {
       ...read station xml file
    } else {
        "warning, file looks like XML but is not StationXML: "+reason
    }
} else if (isSEED(file)) {
   ...read as SEED file
} else {
   warning "input file doesn't look like StationXML or SEED: "+reason
}

Where isXml() and isSeed() did basic checking.

Even better IMHO would be if the command line args were -x <xmlfile> and -s <seedfile> instead of a blanket -input arg. The person running the code usually knows what type of file they want processed and you might as well give them a way to tell the program.