X-Git-Url: http://jsfdemo.indexdata.com/?a=blobdiff_plain;f=src%2Forg%2Fz3950%2Fzing%2Fcql%2FCQLParser.java;h=13e505e9bb0f34f863a034005c2bb7f2bdda7958;hb=18405938a115378cc571d697060ac40775c62fb7;hp=79abc31da72e038c1d7efcb3930ac009c8f5a99e;hpb=08b7f3f08d97b9efdd5a3aef7992a359b71910d6;p=cql-java-moved-to-github.git
diff --git a/src/org/z3950/zing/cql/CQLParser.java b/src/org/z3950/zing/cql/CQLParser.java
index 79abc31..13e505e 100644
--- a/src/org/z3950/zing/cql/CQLParser.java
+++ b/src/org/z3950/zing/cql/CQLParser.java
@@ -1,14 +1,18 @@
-// $Id: CQLParser.java,v 1.14 2002-11-03 16:49:38 mike Exp $
+// $Id: CQLParser.java,v 1.23 2003-09-04 21:56:46 mike Exp $
package org.z3950.zing.cql;
import java.io.IOException;
import java.util.Vector;
+import java.util.Properties;
+import java.io.InputStream;
+import java.io.FileInputStream;
+import java.io.FileNotFoundException;
/**
* Compiles CQL strings into parse trees of CQLNode subtypes.
*
- * @version $Id: CQLParser.java,v 1.14 2002-11-03 16:49:38 mike Exp $
+ * @version $Id: CQLParser.java,v 1.23 2003-09-04 21:56:46 mike Exp $
* @see http://zing.z3950.org/cql/index.html
*/
@@ -30,8 +34,10 @@ public class CQLParser {
* data structure) or, more often, simply rendered out in the
* desired form using one of the back-ends. toCQL()
* returns a decompiled CQL query equivalent to the one that was
- * compiled in the first place; and toXCQL() returns an
- * XML snippet representing the query.
+ * compiled in the first place; toXCQL() returns an
+ * XML snippet representing the query; and toPQF()
+ * returns the query rendered in Index Data's Prefix Query
+ * Format.
*
* @param cql The query
* @return A CQLNode object which is the root of a parse
@@ -41,38 +47,38 @@ public class CQLParser {
lexer = new CQLLexer(cql, LEXDEBUG);
lexer.nextToken();
- debug("about to parse_query()");
- CQLNode root = parse_query("srw.serverChoice", new CQLRelation("="));
+ debug("about to parseQuery()");
+ CQLNode root = parseQuery("srw.serverChoice", new CQLRelation("scr"));
if (lexer.ttype != lexer.TT_EOF)
throw new CQLParseException("junk after end: " + lexer.render());
return root;
}
- private CQLNode parse_query(String qualifier, CQLRelation relation)
+ private CQLNode parseQuery(String qualifier, CQLRelation relation)
throws CQLParseException, IOException {
- debug("in parse_query()");
+ debug("in parseQuery()");
- CQLNode term = parse_term(qualifier, relation);
+ CQLNode term = parseTerm(qualifier, relation);
while (lexer.ttype != lexer.TT_EOF &&
lexer.ttype != ')') {
if (lexer.ttype == lexer.TT_AND) {
match(lexer.TT_AND);
- CQLNode term2 = parse_term(qualifier, relation);
+ CQLNode term2 = parseTerm(qualifier, relation);
term = new CQLAndNode(term, term2);
} else if (lexer.ttype == lexer.TT_OR) {
match(lexer.TT_OR);
- CQLNode term2 = parse_term(qualifier, relation);
+ CQLNode term2 = parseTerm(qualifier, relation);
term = new CQLOrNode(term, term2);
} else if (lexer.ttype == lexer.TT_NOT) {
match(lexer.TT_NOT);
- CQLNode term2 = parse_term(qualifier, relation);
+ CQLNode term2 = parseTerm(qualifier, relation);
term = new CQLNotNode(term, term2);
} else if (lexer.ttype == lexer.TT_PROX) {
match(lexer.TT_PROX);
CQLProxNode proxnode = new CQLProxNode(term);
gatherProxParameters(proxnode);
- CQLNode term2 = parse_term(qualifier, relation);
+ CQLNode term2 = parseTerm(qualifier, relation);
proxnode.addSecondSubterm(term2);
term = (CQLNode) proxnode;
} else {
@@ -85,47 +91,50 @@ public class CQLParser {
return term;
}
- private CQLNode parse_term(String qualifier, CQLRelation relation)
+ private CQLNode parseTerm(String qualifier, CQLRelation relation)
throws CQLParseException, IOException {
- debug("in parse_term()");
+ debug("in parseTerm()");
String word;
while (true) {
if (lexer.ttype == '(') {
debug("parenthesised term");
match('(');
- CQLNode expr = parse_query(qualifier, relation);
+ CQLNode expr = parseQuery(qualifier, relation);
match(')');
return expr;
- } else if (lexer.ttype != lexer.TT_WORD &&
- lexer.ttype != lexer.TT_NUMBER &&
- lexer.ttype != '"') {
- throw new CQLParseException("expected qualifier or term, " +
- "got " + lexer.render());
+ } else if (lexer.ttype == '>') {
+ match('>');
+ return parsePrefix(qualifier, relation);
}
debug("non-parenthesised term");
- if (lexer.ttype == lexer.TT_NUMBER) {
- word = lexer.render();
- } else {
- word = lexer.sval;
- }
- match(lexer.ttype);
+ word = matchSymbol("qualifier or term");
if (!isBaseRelation())
break;
qualifier = word;
- relation = new CQLRelation(lexer.render(lexer.ttype, false));
+ relation = new CQLRelation(lexer.ttype == lexer.TT_WORD ?
+ lexer.sval :
+ lexer.render(lexer.ttype, false));
match(lexer.ttype);
while (lexer.ttype == '/') {
match('/');
if (lexer.ttype != lexer.TT_RELEVANT &&
lexer.ttype != lexer.TT_FUZZY &&
- lexer.ttype != lexer.TT_STEM)
+ lexer.ttype != lexer.TT_STEM &&
+ lexer.ttype != lexer.TT_PHONETIC &&
+ lexer.ttype != lexer.TT_WORD)
throw new CQLParseException("expected relation modifier, "
+ "got " + lexer.render());
- relation.addModifier(lexer.sval);
+ if (lexer.ttype == lexer.TT_WORD &&
+ lexer.sval.indexOf('.') == -1)
+ throw new CQLParseException("unknown first-class " +
+ "relation modifier: " +
+ lexer.sval);
+
+ relation.addModifier(lexer.sval.toLowerCase());
match(lexer.ttype);
}
@@ -138,6 +147,21 @@ public class CQLParser {
return node;
}
+ private CQLNode parsePrefix(String qualifier, CQLRelation relation)
+ throws CQLParseException, IOException {
+ debug("prefix mapping");
+
+ String name = null;
+ String identifier = matchSymbol("prefix-name");
+ if (lexer.ttype == '=') {
+ match('=');
+ name = identifier;
+ identifier = matchSymbol("prefix-identifer");
+ }
+ CQLNode term = parseQuery(qualifier, relation);
+ return new CQLPrefixNode(name, identifier, term);
+ }
+
private void gatherProxParameters(CQLProxNode node)
throws CQLParseException, IOException {
for (int i = 0; i < 4; i++) {
@@ -150,9 +174,9 @@ public class CQLParser {
switch (i) {
// Order should be: relation/distance/unit/ordering
// For now, use MA's: unit/relation/distance/ordering
- case 1: gatherProxRelation(node); break;
- case 2: gatherProxDistance(node); break;
- case 0: gatherProxUnit(node); break;
+ case 0: gatherProxRelation(node); break;
+ case 1: gatherProxDistance(node); break;
+ case 2: gatherProxUnit(node); break;
case 3: gatherProxOrdering(node); break;
}
}
@@ -201,15 +225,25 @@ public class CQLParser {
match(lexer.ttype);
}
- private boolean isBaseRelation() {
+ private boolean isBaseRelation()
+ throws CQLParseException {
debug("isBaseRelation: checking ttype=" + lexer.ttype +
" (" + lexer.render() + ")");
+
+ if (lexer.ttype == lexer.TT_WORD &&
+ lexer.sval.indexOf('.') == -1)
+ throw new CQLParseException("unknown first-class relation: " +
+ lexer.sval);
+
return (isProxRelation() ||
lexer.ttype == lexer.TT_ANY ||
lexer.ttype == lexer.TT_ALL ||
- lexer.ttype == lexer.TT_EXACT);
+ lexer.ttype == lexer.TT_EXACT ||
+ lexer.ttype == lexer.TT_SCR ||
+ lexer.ttype == lexer.TT_WORD);
}
+ // Checks for a relation that may be used inside a prox operator
private boolean isProxRelation() {
debug("isProxRelation: checking ttype=" + lexer.ttype +
" (" + lexer.render() + ")");
@@ -234,6 +268,46 @@ public class CQLParser {
" (tmp=" + tmp + ")");
}
+ private String matchSymbol(String expected)
+ throws CQLParseException, IOException {
+
+ debug("in matchSymbol()");
+ if (lexer.ttype == lexer.TT_WORD ||
+ lexer.ttype == lexer.TT_NUMBER ||
+ lexer.ttype == '"' ||
+ // The following is a complete list of keywords. Because
+ // they're listed here, they can be used unquoted as
+ // qualifiers, terms, prefix names and prefix identifiers.
+ // ### Instead, we should ask the lexer whether what we
+ // have is a keyword, and let the knowledge reside there.
+ lexer.ttype == lexer.TT_AND ||
+ lexer.ttype == lexer.TT_OR ||
+ lexer.ttype == lexer.TT_NOT ||
+ lexer.ttype == lexer.TT_PROX ||
+ lexer.ttype == lexer.TT_ANY ||
+ lexer.ttype == lexer.TT_ALL ||
+ lexer.ttype == lexer.TT_EXACT ||
+ lexer.ttype == lexer.TT_pWORD ||
+ lexer.ttype == lexer.TT_SENTENCE ||
+ lexer.ttype == lexer.TT_PARAGRAPH ||
+ lexer.ttype == lexer.TT_ELEMENT ||
+ lexer.ttype == lexer.TT_ORDERED ||
+ lexer.ttype == lexer.TT_UNORDERED ||
+ lexer.ttype == lexer.TT_RELEVANT ||
+ lexer.ttype == lexer.TT_FUZZY ||
+ lexer.ttype == lexer.TT_STEM ||
+ lexer.ttype == lexer.TT_SCR ||
+ lexer.ttype == lexer.TT_PHONETIC) {
+ String symbol = (lexer.ttype == lexer.TT_NUMBER) ?
+ lexer.render() : lexer.sval;
+ match(lexer.ttype);
+ return symbol;
+ }
+
+ throw new CQLParseException("expected " + expected + ", " +
+ "got " + lexer.render());
+ }
+
/**
* Simple test-harness for the CQLParser class.
@@ -291,19 +365,27 @@ public class CQLParser {
* -c option is supplied].
*/
public static void main (String[] args) {
- boolean canonicalise = false;
+ char mode = 'x'; // x=XCQL, c=CQL, p=PQF
+ String pfile = null;
+
Vector argv = new Vector();
for (int i = 0; i < args.length; i++) {
argv.add(args[i]);
}
if (argv.size() > 0 && argv.get(0).equals("-c")) {
- canonicalise = true;
+ mode = 'c';
+ argv.remove(0);
+ } else if (argv.size() > 1 && argv.get(0).equals("-p")) {
+ mode = 'p';
+ argv.remove(0);
+ pfile = (String) argv.get(0);
argv.remove(0);
}
if (argv.size() > 1) {
- System.err.println("Usage: CQLParser [-c] []");
+ System.err.println(
+ "Usage: CQLParser [-c] [-p []");
System.err.println("If unspecified, query is read from stdin");
System.exit(1);
}
@@ -316,7 +398,7 @@ public class CQLParser {
try {
// Read in the whole of standard input in one go
int nbytes = System.in.read(bytes);
- } catch (java.io.IOException ex) {
+ } catch (IOException ex) {
System.err.println("Can't read query: " + ex.getMessage());
System.exit(2);
}
@@ -324,21 +406,51 @@ public class CQLParser {
}
CQLParser parser = new CQLParser();
- CQLNode root;
+ CQLNode root = null;
try {
root = parser.parse(cql);
- debug("root='" + root + "'");
- if (canonicalise) {
- System.out.println(root.toCQL());
- } else {
- System.out.print(root.toXCQL(0));
- }
} catch (CQLParseException ex) {
System.err.println("Syntax error: " + ex.getMessage());
System.exit(3);
- } catch (java.io.IOException ex) {
+ } catch (IOException ex) {
System.err.println("Can't compile query: " + ex.getMessage());
System.exit(4);
}
+
+ try {
+ if (mode == 'c') {
+ System.out.println(root.toCQL());
+ } else if (mode == 'p') {
+ InputStream f = new FileInputStream(pfile);
+ if (f == null)
+ throw new FileNotFoundException(pfile);
+
+ Properties config = new Properties();
+ config.load(f);
+ f.close();
+ System.out.println(root.toPQF(config));
+ } else {
+ System.out.print(root.toXCQL(0));
+ }
+ } catch (IOException ex) {
+ System.err.println("Can't render query: " + ex.getMessage());
+ System.exit(5);
+ } catch (UnknownQualifierException ex) {
+ System.err.println("Unknown qualifier: " + ex.getMessage());
+ System.exit(6);
+ } catch (UnknownRelationException ex) {
+ System.err.println("Unknown relation: " + ex.getMessage());
+ System.exit(7);
+ } catch (UnknownRelationModifierException ex) {
+ System.err.println("Unknown relation modifier: " +
+ ex.getMessage());
+ System.exit(8);
+ } catch (UnknownPositionException ex) {
+ System.err.println("Unknown position: " + ex.getMessage());
+ System.exit(9);
+ } catch (PQFTranslationException ex) {
+ // We catch all of this class's subclasses, so --
+ throw new Error("can't get a PQFTranslationException");
+ }
}
}