Class XmlSessionReaderV1_2

java.lang.Object
ca.phon.session.io.xml.v1_2.XmlSessionReaderV1_2
All Implemented Interfaces:
IPluginExtensionPoint<SessionReader>, SessionReader, XMLObjectReader<Session>
@XMLSerial(namespace="http://phon.ling.mun.ca/ns/phonbank", elementName="session", bindType=Session.class) @SessionIO(group="ca.phon", id="phonbank", version="1.2", mimetype="application/xml", extension="xml", name="Phon <3.6 (.xml)") public class XmlSessionReaderV1_2 extends Object implements SessionReader, XMLObjectReader<Session>, IPluginExtensionPoint<SessionReader>
Session XML reader for session files with version 'PB1.2' When reading in this older format alignment may be fixed. Fixing alignment Small pinchers for phonetic groups have been added to the IPA parser. This was necessary for one-to-one alignment and is our solution for upgrading Phon data. Currently, I don't do anything when reading the older version 1.3 files when alignment is not complete in groups. We previously discussed using ',' but I think now that is a mistake. Here are some siuations. In each example, I'm using upper case letters for words in orthography and smaller case for ipa. Old Phon: ``` Orthography: [A] [B C] [D E F] [ ] [G H] IPA Target: [1] [1 ] [1 2 3] [ ] [1 ] IPA Actual: [a] [ ] [b c ] [d] [e f] ``` New Phon fixed alignment (spaces added for clarity): ``` Orthography: A ‹B C› ‹D E F› yyy G H . IPA Actual: a * ‹b c› d e f ``` New Phon don't fix alignment: ``` Orthography: A B C D E F G H IPA Actual: a b c d e f ``` Rules: * If # of words match between Orthography and IPA - just add content (group 1/5 above) * If there is content in Orthography and not IPA, wrap orthography in phonetic group markers if more than one word and add * to IPA tier (group 2 above) * If there is content in IPA and not orthography, wrap IPA in phonetic group markers if more than one word and add yyy to Orthography (group 4 above) * If there is content in both IPA and orthography but the number of words do not match, wrap Orthography and IPA in phonetic group markers if they have more than one word (group 3 above) * If no terminator specified one will be added (e.g., '.')