X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=src%2Forg%2Fz3950%2Fzing%2Fcql%2FCQLParser.java;h=58e03269c8aa41bc9196ba9649b9f7e191f1fbbc;hb=af9fc1ccd0a8d9048d0b469a72b11bead020b719;hp=f03c6aa025931f47ed2137f27df2f4a40f8eed64;hpb=9a22e4f1c6e5bdf6ff5cd24ef4035e4de164cafb;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 f03c6aa..58e0326 100644 --- a/src/org/z3950/zing/cql/CQLParser.java +++ b/src/org/z3950/zing/cql/CQLParser.java @@ -1,4 +1,4 @@ -// $Id: CQLParser.java,v 1.29 2007-06-28 00:00:53 mike Exp $ +// $Id: CQLParser.java,v 1.39 2007-08-06 15:54:48 mike Exp $ package org.z3950.zing.cql; import java.io.IOException; @@ -12,15 +12,44 @@ import java.io.FileNotFoundException; /** * Compiles CQL strings into parse trees of CQLNode subtypes. * - * @version $Id: CQLParser.java,v 1.29 2007-06-28 00:00:53 mike Exp $ + * @version $Id: CQLParser.java,v 1.39 2007-08-06 15:54:48 mike Exp $ * @see http://zing.z3950.org/cql/index.html */ public class CQLParser { private CQLLexer lexer; + private int compat; // When false, implement CQL 1.2 + public static int V1POINT1 = 12368; + public static int V1POINT2 = 12369; + public static int V1POINT1SORT = 12370; + static private boolean DEBUG = false; static private boolean LEXDEBUG = false; + /** + * The new parser implements a dialect of CQL specified by the + * compat argument: + * + */ + public CQLParser(int compat) { + this.compat = compat; + } + + /** + * The new parser implements CQL 1.2 + */ + public CQLParser() { + this.compat = V1POINT2; + } + private static void debug(String str) { if (DEBUG) System.err.println("PARSEDEBUG: " + str); @@ -48,39 +77,66 @@ public class CQLParser { lexer.nextToken(); debug("about to parseQuery()"); - CQLNode root = parseQuery("srw.serverChoice", new CQLRelation("scr")); + CQLNode root = parseTopLevelPrefixes("cql.serverChoice", + new CQLRelation(compat == V1POINT2 ? "=" : "scr")); if (lexer.ttype != lexer.TT_EOF) throw new CQLParseException("junk after end: " + lexer.render()); return root; } + private CQLNode parseTopLevelPrefixes(String index, CQLRelation relation) + throws CQLParseException, IOException { + debug("top-level prefix mapping"); + + if (lexer.ttype == '>') { + return parsePrefix(index, relation, true); + } + + CQLNode node = parseQuery(index, relation); + if ((compat == V1POINT2 || compat == V1POINT1SORT) && + lexer.ttype == lexer.TT_SORTBY) { + match(lexer.ttype); + debug("sortspec"); + + CQLSortNode sortnode = new CQLSortNode(node); + while (lexer.ttype != lexer.TT_EOF) { + String sortindex = matchSymbol("sort index"); + ModifierSet ms = gatherModifiers(sortindex); + sortnode.addSortIndex(ms); + } + + if (sortnode.keys.size() == 0) { + throw new CQLParseException("no sort keys"); + } + + node = sortnode; + } + + return node; + } + private CQLNode parseQuery(String index, CQLRelation relation) throws CQLParseException, IOException { debug("in parseQuery()"); CQLNode term = parseTerm(index, relation); while (lexer.ttype != lexer.TT_EOF && - lexer.ttype != ')') { - if (lexer.ttype == lexer.TT_AND) { - match(lexer.TT_AND); + lexer.ttype != ')' && + lexer.ttype != lexer.TT_SORTBY) { + if (lexer.ttype == lexer.TT_AND || + lexer.ttype == lexer.TT_OR || + lexer.ttype == lexer.TT_NOT || + lexer.ttype == lexer.TT_PROX) { + int type = lexer.ttype; + String val = lexer.sval; + match(type); + ModifierSet ms = gatherModifiers(val); CQLNode term2 = parseTerm(index, relation); - term = new CQLAndNode(term, term2); - } else if (lexer.ttype == lexer.TT_OR) { - match(lexer.TT_OR); - CQLNode term2 = parseTerm(index, relation); - term = new CQLOrNode(term, term2); - } else if (lexer.ttype == lexer.TT_NOT) { - match(lexer.TT_NOT); - CQLNode term2 = parseTerm(index, 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 = parseTerm(index, relation); - proxnode.addSecondSubterm(term2); - term = (CQLNode) proxnode; + term = ((type == lexer.TT_AND) ? new CQLAndNode(term, term2, ms) : + (type == lexer.TT_OR) ? new CQLOrNode (term, term2, ms) : + (type == lexer.TT_NOT) ? new CQLNotNode(term, term2, ms) : + new CQLProxNode(term, term2, ms)); } else { throw new CQLParseException("expected boolean, got " + lexer.render()); @@ -91,6 +147,33 @@ public class CQLParser { return term; } + private ModifierSet gatherModifiers(String base) + throws CQLParseException, IOException { + debug("in gatherModifiers()"); + + ModifierSet ms = new ModifierSet(base); + while (lexer.ttype == '/') { + match('/'); + if (lexer.ttype != lexer.TT_WORD) + throw new CQLParseException("expected modifier, " + + "got " + lexer.render()); + String type = lexer.sval.toLowerCase(); + match(lexer.ttype); + if (!isRelation()) { + // It's a simple modifier consisting of type only + ms.addModifier(type); + } else { + // It's a complex modifier of the form type=value + String comparision = lexer.render(lexer.ttype, false); + match(lexer.ttype); + String value = matchSymbol("modifier value"); + ms.addModifier(type, comparision, value); + } + } + + return ms; + } + private CQLNode parseTerm(String index, CQLRelation relation) throws CQLParseException, IOException { debug("in parseTerm()"); @@ -104,8 +187,7 @@ public class CQLParser { match(')'); return expr; } else if (lexer.ttype == '>') { - match('>'); - return parsePrefix(index, relation); + return parsePrefix(index, relation, false); } debug("non-parenthesised term"); @@ -114,20 +196,12 @@ public class CQLParser { break; index = word; - relation = new CQLRelation(lexer.ttype == lexer.TT_WORD ? - lexer.sval : - lexer.render(lexer.ttype, false)); + String relstr = (lexer.ttype == lexer.TT_WORD ? + lexer.sval : lexer.render(lexer.ttype, false)); + relation = new CQLRelation(relstr); match(lexer.ttype); - - while (lexer.ttype == '/') { - match('/'); - if (lexer.ttype != lexer.TT_WORD) - throw new CQLParseException("expected relation modifier, " - + "got " + lexer.render()); - relation.addModifier(lexer.sval.toLowerCase()); - match(lexer.ttype); - } - + ModifierSet ms = gatherModifiers(relstr); + relation.setModifiers(ms); debug("index='" + index + ", " + "relation='" + relation.toCQL() + "'"); } @@ -137,10 +211,12 @@ public class CQLParser { return node; } - private CQLNode parsePrefix(String index, CQLRelation relation) + private CQLNode parsePrefix(String index, CQLRelation relation, + boolean topLevel) throws CQLParseException, IOException { debug("prefix mapping"); + match('>'); String name = null; String identifier = matchSymbol("prefix-name"); if (lexer.ttype == '=') { @@ -148,74 +224,14 @@ public class CQLParser { 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 + CQLNode node = topLevel ? + parseTopLevelPrefixes(index, relation) : + parseQuery(index, relation); - 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; - } - } - } + return new CQLPrefixNode(name, identifier, node); } - private void gatherProxRelation(CQLProxNode node) - throws CQLParseException, IOException { - if (!isRelation()) - 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); - } - - // Checks for a relation that may be used inside a prox operator + // Checks for a relation private boolean isRelation() { debug("isRelation: checking ttype=" + lexer.ttype + " (" + lexer.render() + ")"); @@ -224,7 +240,8 @@ public class CQLParser { lexer.ttype == '=' || lexer.ttype == lexer.TT_LE || lexer.ttype == lexer.TT_GE || - lexer.ttype == lexer.TT_NE); + lexer.ttype == lexer.TT_NE || + lexer.ttype == lexer.TT_EQEQ); } private void match(int token) @@ -256,12 +273,7 @@ public class CQLParser { lexer.ttype == lexer.TT_OR || lexer.ttype == lexer.TT_NOT || lexer.ttype == lexer.TT_PROX || - 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_SORTBY) { String symbol = (lexer.ttype == lexer.TT_NUMBER) ? lexer.render() : lexer.sval; match(lexer.ttype); @@ -320,6 +332,10 @@ public class CQLParser { * </triple> * *

+ * @param -1 + * CQL version 1.1 (default version 1.2) + * @param -d + * Debug mode: extra output written to stderr. * @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 @@ -337,6 +353,12 @@ public class CQLParser { argv.add(args[i]); } + int compat = V1POINT2; + if (argv.size() > 0 && argv.get(0).equals("-1")) { + compat = V1POINT1; + argv.remove(0); + } + if (argv.size() > 0 && argv.get(0).equals("-d")) { DEBUG = true; argv.remove(0); @@ -353,7 +375,8 @@ public class CQLParser { } if (argv.size() > 1) { - System.err.println("Usage: CQLParser [-d] [-c] [-p []"); + System.err.println("Usage: CQLParser [-1] [-d] [-c] " + + "[-p []"); System.err.println("If unspecified, query is read from stdin"); System.exit(1); } @@ -373,7 +396,7 @@ public class CQLParser { cql = new String(bytes); } - CQLParser parser = new CQLParser(); + CQLParser parser = new CQLParser(compat); CQLNode root = null; try { root = parser.parse(cql);