Created
December 4, 2013 09:38
Revisions
-
rsimon created this gist
Dec 4, 2013 .There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters. Learn more about bidirectional Unicode charactersOriginal file line number Diff line number Diff line change @@ -0,0 +1,43 @@ import scala.xml.XML import java.io.FileWriter import scala.xml.transform.RewriteRule import scala.xml.Node import scala.xml.NodeSeq import scala.xml.Elem import scala.xml.transform.RuleTransformer import scala.xml.Text object TEI extends App { val xml = XML.loadFile("/home/simonr/Workspaces/pelagios/pelagios3-scripts/tei/Perseus_text_1999.02.0137.xml") val books = xml \\ "div1" books.foreach(book => { val bookNumber = (book \ "@n").text.toInt // We're only interested in books 3 - 6 (the geographical ones) if (bookNumber > 2 && bookNumber < 7) { (book \\ "div2").foreach(chapter => { val writer = new FileWriter("book" + bookNumber + "_chapter" + (chapter \ "@n").text + ".txt") // Remove 'note' tags val removeNotes = new RewriteRule { override def transform(n: Node): NodeSeq = n match { case e: Elem if (e.label.equals("note")) => Text("") case e: Elem if (e.label.equals("head")) => Text(e.text + "\n\n") case e: Text => Text(e.text.replace("\n", " ")) case n => n } } val cleaned = new RuleTransformer(removeNotes).transform(chapter).text.replaceAll(" +", " ").replace("\n ", "\n").trim() writer.write(cleaned) writer.flush() writer.close() }) } }) }