/** * Extracts entries from the archive. * * @param ctx query context * @return text entries * @throws QueryException query exception */ private TokenList extract(final QueryContext ctx) throws QueryException { final B64 archive = (B64) checkType(checkItem(expr[0], ctx), AtomType.B64); TokenSet hs = null; if (expr.length > 1) { // filter result to specified entries hs = new TokenSet(); final Iter names = ctx.iter(expr[1]); for (Item en; (en = names.next()) != null; ) { hs.add(checkElmStr(en).string(info)); } } final TokenList tl = new TokenList(); final ArchiveIn in = ArchiveIn.get(archive.input(info), info); try { while (in.more()) { final ZipEntry ze = in.entry(); if (ze.isDirectory()) continue; if (hs == null || hs.delete(token(ze.getName())) != 0) tl.add(in.read()); } } catch (final IOException ex) { Util.debug(ex); ARCH_FAIL.thrw(info, ex); } finally { in.close(); } return tl; }
/** * Caches and returns all unique tokens specified in a query. * * @param list token list * @return token set */ private TokenSet unique(final TokenList list) { // cache all query tokens in a set (duplicates are removed) final TokenSet ts = new TokenSet(); switch (mode) { case ALL: case ANY: for (final byte[] t : list) ts.add(t); break; case ALL_WORDS: case ANY_WORD: final FTLexer l = new FTLexer(ftt.opt); for (final byte[] t : list) { l.init(t); while (l.hasNext()) ts.add(l.nextToken()); } break; case PHRASE: final TokenBuilder tb = new TokenBuilder(); for (final byte[] t : list) tb.add(t).add(' '); ts.add(tb.trim().finish()); } return ts; }