root / tags / 1.0.8 / xapian-core / backends / flint / flint_termlisttable.cc

/** @file flint_termlisttable.cc
 * @brief Subclass of FlintTable which holds termlists.
 */
/* Copyright (C) 2007 Olly Betts
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA
 */

#include <config.h>

#include <xapian/document.h>
#include <xapian/error.h>
#include <xapian/termiterator.h>

#include "flint_termlisttable.h"
#include "flint_utils.h"
#include "omassert.h"
#include "omdebug.h"
#include "stringutils.h"
#include "utils.h"

#include <string>

using namespace std;

void
FlintTermListTable::set_termlist(Xapian::docid did,
                                 const Xapian::Document & doc,
                                 flint_doclen_t doclen)
{
    DEBUGCALL(DB, void, "FlintTermListTable::set_termlist",
              did << ", " << doc << ", " << doclen);

    string tag = pack_uint(doclen);

    Xapian::doccount termlist_size = doc.termlist_count();
    if (termlist_size == 0) {
        // doclen is sum(wdf) so should be zero if there are no terms.
        Assert(doclen == 0);
        Assert(doc.termlist_begin() == doc.termlist_end());
        add(flint_docid_to_key(did), string());
        return;
    }

    Xapian::TermIterator t = doc.termlist_begin();
    if (t != doc.termlist_end()) {
        tag += pack_uint(termlist_size);
        string prev_term = *t;

        // Previous database versions encoded a boolean here, which was
        // always false (and pack_bool() encodes false as a '0').  We can
        // just omit this and successfully read old and new termlists
        // except in the case where the next byte is a '0' - in this case
        // we need keep the '0' so that the decoder can just skip any '0'
        // it sees in this position (this shouldn't be a common case - 48
        // character terms aren't very common, and the first term
        // alphabetically is likely to be shorter than average).
        // FIXME: If we have an incompatible database version bump we should
        // drop this completely.
        if (prev_term.size() == '0') tag += '0';

        tag += prev_term.size();
        tag += prev_term;
        tag += pack_uint(t.get_wdf());
        --termlist_size;

        while (++t != doc.termlist_end()) {
            const string & term = *t;
            // If there's a shared prefix with the previous term, we don't
            // store it explicitly, but just store the length of the shared
            // prefix.  In general, this is a big win.
            size_t reuse = common_prefix_length(prev_term, term);

            // reuse must be <= prev_term.size(), and we know that value while
            // decoding.  So if the wdf is small enough that we can multiply it
            // by (prev_term.size() + 1), add reuse and fit the result in a
            // byte, then we can pack reuse and the wdf into a single byte and
            // save ourselves a byte.  We actually need to add one to the wdf
            // before multiplying so that a wdf of 0 can be detected by the
            // decoder.
            size_t packed = 0;
            Xapian::termcount wdf = t.get_wdf();
            // If wdf >= 128, then we aren't going to be able to pack it in so
            // don't even try to avoid the calculation overflowing and making
            // us think we can.
            if (wdf < 127)
                packed = (wdf + 1) * (prev_term.size() + 1) + reuse;

            if (packed && packed < 256) {
                // We can pack the wdf into the same byte.
                tag += char(packed);
                tag += char(term.size() - reuse);
                tag.append(term.data() + reuse, term.size() - reuse);
            } else {
                tag += char(reuse);
                tag += char(term.size() - reuse);
                tag.append(term.data() + reuse, term.size() - reuse);
                // FIXME: pack wdf after reuse next time we rejig the format
                // incompatibly.
                tag += pack_uint(wdf);
            }

            prev_term = *t;
            --termlist_size;
        }
    }
    Assert(termlist_size == 0);
    add(flint_docid_to_key(did), tag);
}

flint_doclen_t
FlintTermListTable::get_doclength(Xapian::docid did) const
{
    DEBUGCALL(DB, flint_doclen_t, "FlintTermListTable::get_doclength", did);

    string tag;
    if (!get_exact_entry(flint_docid_to_key(did), tag))
        throw Xapian::DocNotFoundError("No termlist found for document " +
                                       om_tostring(did));

    if (tag.empty()) RETURN(0);

    const char * pos = tag.data();
    const char * end = pos + tag.size();

    flint_doclen_t doclen;
    if (!unpack_uint(&pos, end, &doclen)) {
        const char *msg;
        if (pos == 0) {
            msg = "Too little data for doclen in termlist";
        } else {
            msg = "Overflowed value for doclen in termlist";
        }
        throw Xapian::DatabaseCorruptError(msg);
    }

    RETURN(doclen);
}