-// $Id: CQLParser.java,v 1.10 2002-10-30 09:19:26 mike Exp $
+// $Id: CQLParser.java,v 1.28 2007-06-27 22:44:40 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 a CQL string into a parse tree.
- * ###
+ * Compiles CQL strings into parse trees of CQLNode subtypes.
*
- * @version $Id: CQLParser.java,v 1.10 2002-10-30 09:19:26 mike Exp $
+ * @version $Id: CQLParser.java,v 1.28 2007-06-27 22:44:40 mike Exp $
* @see <A href="http://zing.z3950.org/cql/index.html"
* >http://zing.z3950.org/cql/index.html</A>
*/
System.err.println("PARSEDEBUG: " + str);
}
+ /**
+ * Compiles a CQL query.
+ * <P>
+ * The resulting parse tree may be further processed by hand (see
+ * the individual node-types' documentation for details on the
+ * data structure) or, more often, simply rendered out in the
+ * desired form using one of the back-ends. <TT>toCQL()</TT>
+ * returns a decompiled CQL query equivalent to the one that was
+ * compiled in the first place; <TT>toXCQL()</TT> returns an
+ * XML snippet representing the query; and <TT>toPQF()</TT>
+ * 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
+ * tree representing the query. */
public CQLNode parse(String cql)
throws CQLParseException, IOException {
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 index, CQLRelation relation)
throws CQLParseException, IOException {
- debug("in parse_query()");
+ debug("in parseQuery()");
- CQLNode term = parse_term(qualifier, relation);
+ CQLNode term = parseTerm(index, 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(index, 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(index, 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(index, relation);
term = new CQLNotNode(term, term2);
} else if (lexer.ttype == lexer.TT_PROX) {
- // ### Handle "prox"
+ match(lexer.TT_PROX);
+ CQLProxNode proxnode = new CQLProxNode(term);
+ gatherProxParameters(proxnode);
+ CQLNode term2 = parseTerm(index, relation);
+ proxnode.addSecondSubterm(term2);
+ term = (CQLNode) proxnode;
} else {
throw new CQLParseException("expected boolean, got " +
lexer.render());
return term;
}
- private CQLNode parse_term(String qualifier, CQLRelation relation)
+ private CQLNode parseTerm(String index, 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(index, relation);
match(')');
return expr;
- } else if (lexer.ttype != lexer.TT_WORD && lexer.ttype != '"') {
- throw new CQLParseException("expected qualifier or term, " +
- "got " + lexer.render());
+ } else if (lexer.ttype == '>') {
+ match('>');
+ return parsePrefix(index, relation);
}
debug("non-parenthesised term");
- word = lexer.sval;
- match(lexer.ttype);
+ word = matchSymbol("index or term");
if (!isBaseRelation())
break;
- qualifier = word;
- relation = new CQLRelation(lexer.render(lexer.ttype, false));
+ index = word;
+ relation = new CQLRelation(lexer.ttype == lexer.TT_WORD ?
+ lexer.sval :
+ lexer.render(lexer.ttype, false));
match(lexer.ttype);
while (lexer.ttype == '/') {
match('/');
- // ### could insist on known modifiers only
- if (lexer.ttype != lexer.TT_WORD)
+ if (lexer.ttype != lexer.TT_RELEVANT &&
+ lexer.ttype != lexer.TT_FUZZY &&
+ 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);
- match(lexer.TT_WORD);
+ 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);
}
- debug("qualifier='" + qualifier + ", " +
+ debug("index='" + index + ", " +
"relation='" + relation.toCQL() + "'");
}
- CQLTermNode node = new CQLTermNode(qualifier, relation, word);
+ CQLTermNode node = new CQLTermNode(index, relation, word);
debug("made term node " + node.toCQL());
return node;
}
- boolean isBaseRelation() {
+ private CQLNode parsePrefix(String index, 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(index, relation);
+ return new CQLPrefixNode(name, identifier, term);
+ }
+
+ private void gatherProxParameters(CQLProxNode node)
+ throws CQLParseException, IOException {
+ for (int i = 0; i < 4; i++) {
+ if (lexer.ttype != '/')
+ return; // end of proximity parameters
+
+ match('/');
+ if (lexer.ttype != '/') {
+ // not an omitted default
+ switch (i) {
+ // Order should be: relation/distance/unit/ordering
+ // For now, use MA's: unit/relation/distance/ordering
+ case 0: gatherProxRelation(node); break;
+ case 1: gatherProxDistance(node); break;
+ case 2: gatherProxUnit(node); break;
+ case 3: gatherProxOrdering(node); break;
+ }
+ }
+ }
+ }
+
+ private void gatherProxRelation(CQLProxNode node)
+ throws CQLParseException, IOException {
+ if (!isProxRelation())
+ throw new CQLParseException("expected proximity relation, got " +
+ lexer.render());
+ node.addModifier("relation", null, lexer.render(lexer.ttype, false));
+ match(lexer.ttype);
+ debug("gPR matched " + lexer.render(lexer.ttype, false));
+ }
+
+ private void gatherProxDistance(CQLProxNode node)
+ throws CQLParseException, IOException {
+ if (lexer.ttype != lexer.TT_NUMBER)
+ throw new CQLParseException("expected proximity distance, got " +
+ lexer.render());
+ node.addModifier("distance", null, lexer.render(lexer.ttype, false));
+ match(lexer.ttype);
+ debug("gPD matched " + lexer.render(lexer.ttype, false));
+ }
+
+ private void gatherProxUnit(CQLProxNode node)
+ throws CQLParseException, IOException {
+ if (lexer.ttype != lexer.TT_pWORD &&
+ lexer.ttype != lexer.TT_SENTENCE &&
+ lexer.ttype != lexer.TT_PARAGRAPH &&
+ lexer.ttype != lexer.TT_ELEMENT)
+ throw new CQLParseException("expected proximity unit, got " +
+ lexer.render());
+ node.addModifier("unit", null, lexer.render());
+ match(lexer.ttype);
+ }
+
+ private void gatherProxOrdering(CQLProxNode node)
+ throws CQLParseException, IOException {
+ if (lexer.ttype != lexer.TT_ORDERED &&
+ lexer.ttype != lexer.TT_UNORDERED)
+ throw new CQLParseException("expected proximity ordering, got " +
+ lexer.render());
+ node.addModifier("ordering", null, lexer.render());
+ match(lexer.ttype);
+ }
+
+ 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_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() + ")");
return (lexer.ttype == '<' ||
lexer.ttype == '>' ||
lexer.ttype == '=' ||
lexer.ttype == lexer.TT_LE ||
lexer.ttype == lexer.TT_GE ||
- lexer.ttype == lexer.TT_NE ||
- lexer.ttype == lexer.TT_ANY ||
- lexer.ttype == lexer.TT_ALL ||
- lexer.ttype == lexer.TT_EXACT);
+ lexer.ttype == lexer.TT_NE);
}
private void match(int token)
" (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
+ // indexes, 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());
+ }
+
- // Test harness.
- //
- // e.g. echo '(au=Kerninghan or au=Ritchie) and ti=Unix' |
- // java org.z3950.zing.cql.CQLParser
- // yields:
- // <triple>
- // <boolean>and</boolean>
- // <triple>
- // <boolean>or</boolean>
- // <searchClause>
- // <index>au<index>
- // <relation>=<relation>
- // <term>Kerninghan<term>
- // </searchClause>
- // <searchClause>
- // <index>au<index>
- // <relation>=<relation>
- // <term>Ritchie<term>
- // </searchClause>
- // </triple>
- // <searchClause>
- // <index>ti<index>
- // <relation>=<relation>
- // <term>Unix<term>
- // </searchClause>
- // </triple>
- //
+ /**
+ * Simple test-harness for the CQLParser class.
+ * <P>
+ * Reads a CQL query either from its command-line argument, if
+ * there is one, or standard input otherwise. So these two
+ * invocations are equivalent:
+ * <PRE>
+ * CQLParser 'au=(Kerninghan or Ritchie) and ti=Unix'
+ * echo au=(Kerninghan or Ritchie) and ti=Unix | CQLParser
+ * </PRE>
+ * The test-harness parses the supplied query and renders is as
+ * XCQL, so that both of the invocations above produce the
+ * following output:
+ * <PRE>
+ * <triple>
+ * <boolean>
+ * <value>and</value>
+ * </boolean>
+ * <triple>
+ * <boolean>
+ * <value>or</value>
+ * </boolean>
+ * <searchClause>
+ * <index>au</index>
+ * <relation>
+ * <value>=</value>
+ * </relation>
+ * <term>Kerninghan</term>
+ * </searchClause>
+ * <searchClause>
+ * <index>au</index>
+ * <relation>
+ * <value>=</value>
+ * </relation>
+ * <term>Ritchie</term>
+ * </searchClause>
+ * </triple>
+ * <searchClause>
+ * <index>ti</index>
+ * <relation>
+ * <value>=</value>
+ * </relation>
+ * <term>Unix</term>
+ * </searchClause>
+ * </triple>
+ * </PRE>
+ * <P>
+ * @param -c
+ * Causes the output to be written in CQL rather than XCQL - that
+ * is, a query equivalent to that which was input, is output. In
+ * effect, the test harness acts as a query canonicaliser.
+ * @return
+ * The input query, either as XCQL [default] or CQL [if the
+ * <TT>-c</TT> option is supplied].
+ */
public static void main (String[] args) {
- if (args.length != 0) {
- System.err.println("Usage: " + args[0]);
+ char mode = 'x'; // x=XCQL, c=CQL, p=PQF
+ String pfile = null;
+
+ Vector<String> argv = new Vector<String>();
+ for (int i = 0; i < args.length; i++) {
+ argv.add(args[i]);
+ }
+
+ if (argv.size() > 0 && argv.get(0).equals("-d")) {
+ DEBUG = true;
+ argv.remove(0);
+ }
+
+ if (argv.size() > 0 && argv.get(0).equals("-c")) {
+ 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 [-d] [-c] [-p <pqf-properties> [<CQL-query>]");
+ System.err.println("If unspecified, query is read from stdin");
System.exit(1);
}
- byte[] bytes = new byte[10000];
- try {
- // Read in the whole of standard input in one go
- int nbytes = System.in.read(bytes);
- } catch (java.io.IOException ex) {
- System.err.println("Can't read query: " + ex.getMessage());
- System.exit(2);
+ String cql;
+ if (argv.size() == 1) {
+ cql = (String) argv.get(0);
+ } else {
+ byte[] bytes = new byte[10000];
+ try {
+ // Read in the whole of standard input in one go
+ int nbytes = System.in.read(bytes);
+ } catch (IOException ex) {
+ System.err.println("Can't read query: " + ex.getMessage());
+ System.exit(2);
+ }
+ cql = new String(bytes);
}
- String cql = new String(bytes);
+
CQLParser parser = new CQLParser();
- CQLNode root;
+ CQLNode root = null;
try {
root = parser.parse(cql);
- debug("root='" + root + "'");
- System.out.println(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 (UnknownIndexException ex) {
+ System.err.println("Unknown index: " + 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");
+ }
}
}