+ * 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. toCQL() + * returns a decompiled CQL query equivalent to the one that was + * 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 + * tree representing the query. */ + public CQLNode parse(String cql) + throws CQLParseException, IOException { + lexer = new CQLLexer(cql, LEXDEBUG); - st = new StreamTokenizer(new StringReader(cql)); - st.wordChars('/', '/'); - st.wordChars('0', '9'); // ### but 1 is still recognised as TT_NUM - st.wordChars('.', '.'); - st.wordChars('-', '-'); - st.ordinaryChar('='); - st.ordinaryChar(','); - st.ordinaryChar('('); - st.ordinaryChar(')'); - -// int token; -// while ((token = st.nextToken()) != st.TT_EOF) { -// System.out.println("token=" + token + ", " + -// "nval=" + st.nval + ", " + -// "sval=" + st.sval); -// } - - st.nextToken(); - String ret; - try { - ret = parse_expression(); - } catch (CQLParseException ex) { - System.err.println("### Oops: " + ex); - return null; - } + lexer.nextToken(); + 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()); - if (st.ttype != st.TT_EOF) { - System.err.println("### Extra bits: " + render(st)); - return null; - } - - // Interpret attributes as BIB-1 unless otherwise specified - return "@attrset bib-1 " + ret; + return root; } - private String parse_expression() + private CQLNode parseQuery(String index, CQLRelation relation) throws CQLParseException, IOException { - String term = parse_term(); + debug("in parseQuery()"); - while (st.ttype == st.TT_WORD) { - String op = st.sval.toLowerCase(); - if (!st.sval.equals("and") && - !st.sval.equals("or") && - !st.sval.equals("not")) - break; - match(st.TT_WORD); - String term2 = parse_term(); - term = "@" + op + " " + term + " " + term2; + 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 = 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; + } else { + throw new CQLParseException("expected boolean, got " + + lexer.render()); + } } + debug("no more ops"); return term; } - private String parse_term() + private CQLNode parseTerm(String index, CQLRelation relation) throws CQLParseException, IOException { - if (st.ttype == '(') { - match('('); - String expr = parse_expression(); - match(')'); - return expr; - } + debug("in parseTerm()"); - String word = null; - String attrs = ""; + String word; + while (true) { + if (lexer.ttype == '(') { + debug("parenthesised term"); + match('('); + CQLNode expr = parseQuery(index, relation); + match(')'); + return expr; + } else if (lexer.ttype == '>') { + match('>'); + return parsePrefix(index, relation); + } - // ### We treat ',' and '=' equivalently here, which isn't quite right. - while (st.ttype == st.TT_WORD) { - word = st.sval; - match(st.TT_WORD); - if (st.ttype != '=' && st.ttype != ',') { - // end of qualifer list + debug("non-parenthesised term"); + word = matchSymbol("index or term"); + if (!isBaseRelation()) break; - } - String attr = qualsetProperties.getProperty(word); - if (attr == null) { - throw new CQLParseException("unrecognised qualifier: " + word); + index = word; + 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_PHONETIC && + lexer.ttype != lexer.TT_WORD) + throw new CQLParseException("expected relation modifier, " + + "got " + lexer.render()); + 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); } - attrs = attrs + attr + " "; - match(st.ttype); - word = null; // mark as not-yet-read + + debug("index='" + index + ", " + + "relation='" + relation.toCQL() + "'"); } - if (word == null) { - // got to the end of a "foo,bar=" sequence - word = st.sval; - if (st.ttype != '\'' || st.ttype != '"') { - word = "\"" + word + "\""; - match(st.ttype); - } else { - match(st.TT_WORD); + CQLTermNode node = new CQLTermNode(index, relation, word); + debug("made term node " + node.toCQL()); + return node; + } + + 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; + } } } + } - return attrs + word; + 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 match(int token) + 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 (st.ttype != token) - throw new CQLParseException("expected " + render(st, token, null) + - ", " + "got " + render(st)); - st.nextToken(); + 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); } - // ### This utility should surely be a method of the StreamTokenizer class - private static String render(StreamTokenizer st) { - return render(st, st.ttype, null); + 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 static String render(StreamTokenizer st, int token, String str) { - String ret; + private boolean isBaseRelation() + throws CQLParseException { + debug("isBaseRelation: checking ttype=" + lexer.ttype + + " (" + lexer.render() + ")"); - switch (token) { - case st.TT_EOF: return "EOF"; - case st.TT_EOL: return "EOL"; - case st.TT_NUMBER: return "number"; - case st.TT_WORD: ret = "word"; break; - case '"': case '\'': ret = "string"; break; - default: return "'" + String.valueOf((char) token) + "'"; + 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); + } + + private void match(int token) + throws CQLParseException, IOException { + debug("in match(" + lexer.render(token, true) + ")"); + if (lexer.ttype != token) + throw new CQLParseException("expected " + + lexer.render(token, true) + + ", " + "got " + lexer.render()); + int tmp = lexer.nextToken(); + debug("match() got token=" + lexer.ttype + ", " + + "nval=" + lexer.nval + ", sval='" + lexer.sval + "'" + + " (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; } - if (str != null) - ret += "(\"" + str + "\")"; - return ret; + throw new CQLParseException("expected " + expected + ", " + + "got " + lexer.render()); } - // ### Not really the right place for this test harness. - // - // e.g. java uk.org.miketaylor.zoom.CQLCompiler - // '(au=Kerninghan or au=Ritchie) and ti=Unix' qualset.properties - // yields: - // @and - // @or - // @attr 1=1 @attr 4=1 Kerninghan - // @attr 1=1 @attr 4=1 Ritchie - // @attr 1=4 @attr 4=1 Unix - // + + /** + * Simple test-harness for the CQLParser class. + *
+ * Reads a CQL query either from its command-line argument, if + * there is one, or standard input otherwise. So these two + * invocations are equivalent: + *
+ * CQLParser 'au=(Kerninghan or Ritchie) and ti=Unix' + * echo au=(Kerninghan or Ritchie) and ti=Unix | CQLParser + *+ * The test-harness parses the supplied query and renders is as + * XCQL, so that both of the invocations above produce the + * following output: + *
+ * <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> + *+ *
+ * @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
+ * -c option is supplied].
+ */
public static void main (String[] args) {
- if (args.length != 2) {
- System.err.println("Usage: CQLQuery