--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/software/eXist/mpdl-modules/src/de/mpg/mpiwg/berlin/mpdl/lt/analyzer/lang/MpdlNormalizerLexDE.lex	Tue Feb 22 16:03:45 2011 +0100
@@ -0,0 +1,117 @@
+/*
+ * Normalization rules for German text
+ * [this is a JFlex specification]
+ *
+ * Wolfgang Schmidle 
+ * version 0.96
+ * 2011-02-21
+ *
+ */
+
+package de.mpg.mpiwg.berlin.mpdl.lt.analyzer.lang;
+
+%%
+
+%public
+%class MpdlNormalizerLexDE
+%type java.lang.String
+%unicode
+
+// German: de, deu, ger
+
+%states DISP, DICT, SEARCH
+%state CELEX, GRIMM
+
+%{
+	private String original = "";
+	private String normalized = "";
+	private int problem = 0;
+	
+	private void add (String norm) {
+		original += yytext(); 
+		normalized += norm;
+	}
+%}
+
+END = \n
+
+Alphabet = [abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ]
+
+%%
+
+ſ { add("s"); }
+
+// Fraktur
+
+<DISP, DICT, SEARCH,
+GRIMM> {
+
+uͦ {add("u"); }
+aͤ {add("ä"); }
+oͤ {add("ö"); }
+uͤ {add("ü"); }
+
+}
+
+<CELEX> {
+
+// normalize ä ö ü ß only for Celex!
+
+ä | Ä | aͤ { add("ae"); }
+ö | Ö | oͤ { add("oe"); }
+ü | Ü | uͤ { add("ue"); }
+uͦ {add("u"); }
+ß { add("ss"); }
+
+{Alphabet} { add(yytext()); }
+
+. { problem = 1; add(yytext()); }
+
+}
+
+<GRIMM> {
+
+ß { add("sz"); }
+
+}
+
+
+// default
+
+@ { problem = 1; add(yytext()); }
+. { add(yytext()); }
+
+
+<DISP, SEARCH> {
+
+{END} {
+		switch (problem) {
+			case 1: return original;
+			default: return normalized;
+		}
+	}
+}
+
+<DICT, CELEX, GRIMM> {
+
+{END} {
+		switch (problem) {
+			case 1: return "";
+			default: return normalized;
+		}
+	}
+}
+
+
+/*
+
+Annahmen:
+- die Routine wird wortweise aufgerufen, mit einem \n am Ende des Strings
+- Wörter mit Zeilenumbrüchen wurden bereits wieder zusammengesetzt
+
+TO DO:
+
+DE: Trennung von Deutsch und Fraktur?
+DE: Celex: hyphens weg?
+
+*/