opLEM: reject wildcards with error message.

doTestQuery

@@ -3,15 +3,15 @@
 # 06.12.24/FB
 
 # current Koral lib:
-KJAR=./target/Koral-0.42.0.jar
+KJAR=./target/Koral-0.44.0.jar
 # Calling that main class:
 MCL=de.ids_mannheim.korap.query.serialize.QuerySerializer
 # Query Language:
 QL="cosmas2"
 
 Query="&Fes-&Mann"
 Query="&F+&Mann+"
-Query="ab /s0 &F+&Mann+"
+Query="ab /s0 &?&+&*&Mann"
 
 echo -e "Query = '$Query' in QL='$QL'..."
 

src/main/antlr/cosmas/c2ps.g

@@ -14,7 +14,11 @@
 //    - more generally: comma at end of searchword, which is not enclosed by "..." is
 //      excluded from searchword now.
 //    - a comma inside a searchword is accepted if enclosed by "...".
-//
+//  10.12.24/FB
+//    - skip wildcards [?*+] in lemma search expression, as regex/wildcards are not allowed
+//      in &opts&lemma, but instead they may appear as options in 'opts'.
+//      E.g. &F+&Prüfung -> lemma with F+ as an option.
+//    - test added for F+.
 // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
 
 grammar c2ps;
@@ -236,8 +240,8 @@ searchExpr
 		-> $se1+ ;
 
 searchExpr1
-	:	op1 		   -> {$op1.tree}
-	| 	searchWord 	   -> {$searchWord.tree}
+	:	op1 			   -> {$op1.tree}
+	| 	searchWord 		   -> {$searchWord.tree}
 	| 	searchLemma 	   -> {$searchLemma.tree}
 	| 	searchAnnot 	   -> {$searchAnnot.tree}
 	| 	searchLabel        -> {$searchLabel.tree}
@@ -250,13 +254,13 @@ searchWord
 	:	word1
 	|	word2;
 
-word1	:	SEARCHWORD1 -> {c2ps_opWF.check($SEARCHWORD1.text, false, false, $SEARCHWORD1.index)} ; 
+word1	:	SEARCHWORD1 -> {c2ps_opWF.check($SEARCHWORD1.text, false, false, $SEARCHWORD1.pos)} ; 
 
-word2	:	SEARCHWORD2 -> {c2ps_opWF.check($SEARCHWORD2.text, true, false, $SEARCHWORD2.index)} ;
+word2	:	SEARCHWORD2 -> {c2ps_opWF.check($SEARCHWORD2.text, true, false, $SEARCHWORD2.pos)} ;
 
 // Suchbegriff = Lemma:
 searchLemma
-	:	SEARCHLEMMA -> {c2ps_opWF.check($SEARCHLEMMA.text, false, true, $SEARCHLEMMA.index)} ; 
+	:	SEARCHLEMMA -> {c2ps_opWF.check($SEARCHLEMMA.text, false, true, $SEARCHLEMMA.pos)} ; 
 
 // Suchbegriff = Annotationsoperator:
 // (damit Lexer den richtige Token erzeugt, muss OP_ELEM den gesamten

src/main/antlr/cosmas/c2ps_opWF.g

@@ -47,14 +47,14 @@ WS	:	(' ')+ {skip();};
 //
 // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
 
-searchWFs
-	:	searchWF+;
+searchWFs[int pos]
+	:	searchWF[pos]+;
 
-searchWF:	optCase? wordform tpos?
+searchWF[int pos] :	optCase? wordform[OPWF,pos] tpos?
 
 		-> ^(OPWF wordform optCase? tpos? ) ;
 
-wordform:	WF -> {c2ps_opWF.encode($WF.text, OPWF)};
+wordform[int type, int pos]:	WF -> {c2ps_opWF.encode($WF.text, $type, $pos)};
 
 // Case Options:
 optCase	:	Case 
@@ -67,8 +67,8 @@ tpos	:	TPos
 		-> ^(TPOS {c2ps_opBED.checkTPos($TPos.text, $TPos.index)});
 
 // analog für Lemmata, kein optCase:
-searchLEM
-	:	wordform tpos?
-	
-		-> ^(OPLEM wordform tpos?);
-		
+// todo: check wordform (=lemma) for wildcards, which are not allowed in the lemma expr.
+
+searchLEM[int pos]
+   : wordform[OPLEM,pos] tpos?
+   -> ^(OPLEM wordform tpos?);

src/main/java/de/ids_mannheim/korap/query/parse/cosmas/c2ps_opPROX.java

@@ -22,7 +22,7 @@ public class c2ps_opPROX
 	public static final int MLANG_GERMAN  = 1;
 
 	public static int
-		messLang = MLANG_ENGLISH; // default.
+		messLang = MLANG_GERMAN; // default.
 
 	// type of an Error CommonToken:
 	final static int 
@@ -53,9 +53,9 @@ private static String getErrMessEN(int errCode, String text)
 
 	case StatusCodes.ERR_PROX_WRONG_CHARS:
 		return String.format("Proximity operator at '%s': unknown proximity options!", text);
-		
-	case StatusCodes.UNKNOWN_QUERY_ERROR:
-		return String.format("Unknown error!");
+
+	case StatusCodes.ERR_LEM_WILDCARDS:
+		return String.format("Lemma operator at '%s': wildcards (?*+) are not allowed inside a lemma.", text);
 
 	default:
 		return String.format("Proximity operator at '%s': unknown error. The correct syntax looks like this: E.g. ' /+w2 ' or ' /w10,s0 '.", text);
@@ -86,33 +86,15 @@ private static String getErrMessGE(int errCode, String text)
 	case StatusCodes.ERR_PROX_WRONG_CHARS:
 		return String.format("Abstandsoperator an der Stelle '%s': unbekannte Abstandsoption(en)!", text);
 
-	case StatusCodes.UNKNOWN_QUERY_ERROR:
-		return String.format("Unbekannter Fehler!");
+	case StatusCodes.ERR_LEM_WILDCARDS:
+		return String.format("Lemma-Suchbegriff an der Stelle '%s': Platzhalter (?*+) können im gesuchten Lemma nicht eingesetzt werden.", text);
 
 	default:
 		return String.format("Abstandsoperator an der Stelle '%s': unbekannter Fehler. Korrekte Syntax z.B.: ' /+w2 ' oder ' /w10,s0 '.", text);
 		}
 	}
 
-	private static String getErrMess(int errCode, int messLang, String text)
-
-	{
-	if( messLang == c2ps_opPROX.MLANG_GERMAN )
-		return getErrMessGE(errCode, text);
-	else
-		return getErrMessEN(errCode, text);	
-	}
-
-
-	/**
-	 * in this version, the pre-stored message language is used.
-	 * @param errCode
-	 * @param text
-	 * @return
-	 * 10.06.24/FB
-	 */
-
-	public static String getErrMess(int errCode, String text)
+	public static String getErrMess(int errCode, int messLang, String text)
 
 	{
 	if( messLang == c2ps_opPROX.MLANG_GERMAN )
@@ -183,7 +165,7 @@ public static Object encodeDIST(int typeDIST, int typeDIR, Object ctDir, Object
 		CommonTree tree3 = (CommonTree)ctVal;
 
 		if( bDebug )
-			System.err.printf("Debug: encodeDIST: scanned input='%s' countM=%d countD=%d countV=%d pos=%d.\n", 
+			System.out.printf("Debug: encodeDIST: scanned input='%s' countM=%d countD=%d countV=%d pos=%d.\n", 
 					text, countM, countD, countV, pos);
 
 		if( countM == 0 )
@@ -203,7 +185,7 @@ public static Object encodeDIST(int typeDIST, int typeDIR, Object ctDir, Object
 			treeDIR.addChild(treeBOTH);
 
 			if( bDebug )
-				System.err.printf("Debug: encodeDIST: tree for DIR: '%s'.\n", treeDIR.toStringTree());
+				System.out.printf("Debug: encodeDIST: tree for DIR: '%s'.\n", treeDIR.toStringTree());
 			tree1 = treeDIR;
 			}
 		else if( countD > 1 )
@@ -218,7 +200,7 @@ else if( countD > 1 )
 		tree.addChild(tree2);
 
 		if( bDebug )
-			System.err.printf("Debug: encodeDIST: returning '%s'.\n", tree.toStringTree());
+			System.out.printf("Debug: encodeDIST: returning '%s'.\n", tree.toStringTree());
 
 		return tree;
 	} // encodeDIST

src/main/java/de/ids_mannheim/korap/query/parse/cosmas/c2ps_opREG.java

@@ -27,7 +27,7 @@
 public class c2ps_opREG
 
 {
-	private static boolean DEBUG = false;
+	private static boolean DEBUG = true;
 
 	/* 
 	 * encode():

src/main/java/de/ids_mannheim/korap/query/parse/cosmas/c2ps_opWF.java

@@ -2,6 +2,9 @@
 
 import org.antlr.runtime.*;
 import org.antlr.runtime.tree.*;
+import de.ids_mannheim.korap.query.parse.cosmas.c2ps_opPROXLexer;
+import de.ids_mannheim.korap.query.parse.cosmas.c2ps_opPROX;
+import de.ids_mannheim.korap.query.serialize.util.StatusCodes;
 
 /*
  * parses prefixed and suffixed options of a search wordform.
@@ -11,6 +14,8 @@
 public class c2ps_opWF
 
 {
+	//static final int OPWF  = 5; // must be same value than OPWF in c2ps_opWF.g
+	//static final int OPLEM = 7; // must be same value than OPLEM in c2ps_opWF.g
     /* Arguments:
      * bStrip: true: 'input' contains "wort" -> strip " away -> wort.
      *        false: 'input' contains no " -> nothing to strip.
@@ -19,9 +24,16 @@ public class c2ps_opWF
      * input: may be a single Lemma or Wform or a list of Wforms.
      */
 
-    public static Tree check (String input, boolean bStrip, boolean bLem,
-            int index) {
-        if (bStrip)
+    public static Tree check (String input, boolean bStrip, boolean bLem, int pos) 
+    {
+    	if( bLem )
+    		{
+    		System.out.printf("c2ps_opWF.check: input='%s' bStrip=%b bLem=%b pos=%d.\n", 
+    				 input, bStrip, bLem, pos);
+    		System.out.flush();
+    		}
+
+    	if (bStrip)
             input = input.substring(1, input.length() - 1);
 
         if (bLem && input.charAt(0) == '&') {
@@ -43,18 +55,23 @@ public static Tree check (String input, boolean bStrip, boolean bLem,
 
         try {
             if (bLem)
-                c2PQLEMReturn = g.searchLEM();
+                c2PQLEMReturn = g.searchLEM(pos);
             else
-                c2PQWFReturn = g.searchWFs();
+                c2PQWFReturn = g.searchWFs(pos);
         }
         catch (RecognitionException e) {
             e.printStackTrace();
         }
 
         // AST Tree anzeigen:
-        Tree tree = bLem ? (Tree) c2PQLEMReturn.getTree() : (Tree) c2PQWFReturn
-                .getTree();
-        // System.out.println(bLem? "opLEM: " : "opWF: " + tree.toStringTree() );
+        Tree tree = bLem ? (Tree)c2PQLEMReturn.getTree() : (Tree)c2PQWFReturn.getTree();
+
+        if( bLem )
+        	 {
+        	 System.out.printf("c2ps_opWF.check: %s: '%s'.\n", bLem ? "opLEM" : "opWF", 
+        			 tree.toStringTree() );
+        	 System.out.flush();
+        	 }
 
         return tree;
     }
@@ -63,24 +80,108 @@ public static Tree check (String input, boolean bStrip, boolean bLem,
     /* Wordform Encoding, e.g. to insert a Wordform into an AST.
      * a) wf ->  "wf".
      * b) remove escape char before ':': abc\: -> abc:.
-     * Returns a Tree.
+     * Args:
+     * wf			: wordform or lemma (expected lemma : "lemma" or "opts&lemma",
+     * 				  the starting '&' has been removed before entering this function).
+     * tokenType	: either OPWF or OPLEM.
+     * pos			: start position of wf.
+     * Notes:
+     *  - &opts&lemma : may contain wildcards as options in the &opts& section only.
+     *    reject if wildcards appear in the &lemma section.
+     * Returns a Tree or an ErrorTree.
      */
-    public static Tree encode (String wf, int tokenType)
+    public static Tree encode (String wf, int tokenType, int pos)
 
     {
+    	//System.out.printf("c2ps_opWF.encode: wf='%s' tokenType=%d pos=%d.\n",  wf, tokenType, pos);
+
         // b)
         StringBuffer sbWF = new StringBuffer(wf);
 
-        for (int i = 0; i < sbWF.length() - 1; i++) {
+        for (int i = 0; i < sbWF.length()-1; i++) 
+        	{
             if (sbWF.charAt(i) == '\\' && sbWF.charAt(i + 1) == ':')
                 sbWF.deleteCharAt(i);
-        }
-
-        return new CommonTree(new CommonToken(tokenType, "\"" + sbWF.toString()
-                + "\""));
+        	}
+
+        if( tokenType == c2ps_opWFLexer.OPLEM )
+        	{
+        	boolean hasOpts  = false; // true if a '&' occurs: e.g. "Fes+C&lemma"
+        	boolean hasFound = false; // false for all wildcards found to the left of '&', true in all other cases.
+
+        	for(int i=0; i< sbWF.length(); i++)
+	        	{
+        		if( sbWF.charAt(i) == '&' )
+        			{
+        			hasOpts  = true;
+        			hasFound = false;
+        			}
+        		else if (sbWF.charAt(i) == '?' || sbWF.charAt(i) == '*' || sbWF.charAt(i) == '+' )
+        			{
+        			hasFound = true;
+        			}
+        		}
+
+        	// error if hasFound==true:
+        	if( hasFound )
+	        	{
+        		System.out.printf("c2ps_opWF.encode: Syntax error: '%s' contains wildcards inside lemma expression!\n", wf);
+        		return buildErrorTree(wf, StatusCodes.ERR_LEM_WILDCARDS, pos);
+        		}
+        	}
+
+        return new CommonTree(new CommonToken(tokenType, "\"" + sbWF.toString() + "\""));
     }
 
-
+    /**
+	 * buildErrorTree(): 
+	 * @param text = part of the query that contains an error.
+	 * @param errCode
+	 * @param typeDIST
+	 * @param pos
+	 * @return
+	 */
+
+	//private static CommonTree buildErrorTree(String text, int errCode, int typeDIST, int pos)
+
+    private static CommonTree buildErrorTree(String text, int errCode, int pos)
+	{
+	/*
+	 CommonTree
+	//errorTree = new CommonTree(new CommonToken(typeDIST, "DIST"));
+	errorTree = new CommonTree(new CommonToken(c2ps_opPROX.typeERROR, "Fehlercherchen"));
+	*/
+	CommonTree
+		errorNode = new CommonTree(new CommonToken(c2ps_opPROX.typeERROR, "ERROR"));
+	CommonTree
+		errorPos  = new CommonTree(new CommonToken(c2ps_opPROX.typeERROR, String.valueOf(pos)));
+	CommonTree
+		errorCode = new CommonTree(new CommonToken(c2ps_opPROX.typeERROR, String.valueOf(errCode)));
+	CommonTree
+		errorMes;
+	String
+		mess;
+
+	mess 	 = c2ps_opPROX.getErrMess(errCode, c2ps_opPROX.messLang, text);
+	errorMes = new CommonTree(new CommonToken(c2ps_opPROX.typeERROR, mess));
+
+	// new:
+	errorNode.addChild(errorPos);
+	errorNode.addChild(errorCode);
+	errorNode.addChild(errorMes);
+
+	return errorNode;
+
+	/* old, no need for errorTree(typeXY).
+	errorTree.addChild(errorNode);
+	errorNode.addChild(errorPos);
+	errorNode.addChild(errorCode);
+	errorNode.addChild(errorMes);
+	
+	return errorTree;
+	*/
+	}
+
     /*
      * main testprogram:
      */