1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
|
// Copyright Vespa.ai. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
#include "mcand.h"
#include "Matcher.h"
#include "juniperdebug.h"
#include <vespa/vespalib/util/stringfmt.h>
#include <vespa/log/log.h>
LOG_SETUP(".juniper.mcand");
// Invariant: elms has room for query->_arity match element pointers
MatchCandidate::MatchCandidate(QueryExpr* m, MatchElement** elms, off_t ctxt_start) :
MatchElement(0, 0),
element(elms),
_match(m),
_nelems(0),
_elems(std::max(m->_arity, 1)),
_endpos(0),
_endtoken(0),
_docid(0),
_ctxt_start(ctxt_start),
_elem_weight(0),
_options(m->_options),
_overlap(0),
_refcnt(1),
_klist()
{
for (int i = 0; i < _elems; i++)
element[i] = NULL;
if (LOG_WOULD_LOG(debug)) {
std::string s;
dump(s);
LOG(debug, "new %s", s.c_str());
}
}
MatchCandidate::~MatchCandidate()
{
delete[] element;
}
void MatchCandidate::dump(std::string& s)
{
int i;
s.append("MC<");
for (i = 0; i < _elems; i++)
{
if (i > 0) s.append(";");
_match->AsNode()->_children[i]->Dump(s);
s.append(":");
if (element[i])
{
s.append(vespalib::make_string("%" PRId64,
static_cast<int64_t>
(element[i]->starttoken())));
if (element[i]->starttoken() + 1 < element[i]->endtoken())
s.append(vespalib::make_string("-%" PRId64,
static_cast<int64_t>
(element[i]->endtoken())));
}
else
s.append("<nil>");
}
s.append(">");
}
void MatchCandidate::set_valid()
{
for (int j = 0; j < _elems; j++)
if (element[j])
element[j]->set_valid();
_valid = true;
}
void MatchCandidate::make_keylist()
{
add_to_keylist(_klist);
}
void MatchCandidate::add_to_keylist(keylist& kl)
{
if (kl.size() > 0) return; // already made list
for (int i = 0; i < _elems; i++)
{
MatchElement* me = element[i];
if (me) me->add_to_keylist(kl);
}
}
MatchCandidate::accept_state MatchCandidate::accept(MatchElement* k, QueryExpr* mexp)
{
if (element[mexp->_childno]) {
if (_overlap) return M_OVERLAP;
return M_EXISTS;
} else {
if (order()) {
// Ensure that overlapping matches are not considered in ordered mode..
if (k->startpos() < _endpos) {
_overlap++;
return M_OVERLAP;
} else {
_overlap--; // Found overlap..
}
}
element[mexp->_childno] = k;
// Note that in 2.1.x match elements are no longer arriving in position order!
// They may also overlap (because they may be complex candidates themselves)
if (!_nelems || (k->startpos() < _startpos)) {
_startpos = k->startpos();
_starttoken = k->starttoken();
}
_nelems++;
// Update 2.0 term weight/element count/combined element word length
_elem_weight += weight(k, mexp);
if (!_nelems || (k->endpos() > _endpos)) {
_endpos = k->startpos() + k->length();
_endtoken = k->starttoken() + k->word_length();
}
if (LOG_WOULD_LOG(spam)) {
std::string s("(accept:"); k->dump(s);
s.append(") "); dump(s);
LOG(spam, "%s", s.c_str());
}
return M_OK;
}
}
int MatchCandidate::weight(MatchElement* me, QueryExpr* mexp)
{
QueryTerm* texp = mexp->AsTerm();
if (texp) return mexp->_weight;
MatchCandidate* m = reinterpret_cast<MatchCandidate*>(me);
return m->weight();
}
bool MatchCandidate::complete()
{
if (_nelems < _elems) return false;
for (int i = 0; i < _elems; i++)
if (!element[i]->complete()) return false;
return true;
}
void MatchCandidate::log(std::string& logobj)
{
char buf[200];
for (int i = 0; i < _elems; i++)
{
if (element[i])
{
snprintf(buf, sizeof(buf), "<td align=left>%" PRId64 "</td>",
static_cast<int64_t>(element[i]->starttoken()));
logobj.append(buf);
}
else
logobj.append("<td></td>");
}
snprintf(buf, sizeof(buf), "<td align=right>%d</td><td align=right>%d</td>", word_distance(),rank());
logobj.append(buf);
}
// Check optional WITHIN(limit) constraints:
bool MatchCandidate::matches_limit()
{
if (!match()->HasLimit()) return true;
// completeness check:
if (!complete()) return false;
int limit = match()->Limit();
size_t elem_word_len = element[0]->word_length();
for (int i = 1; i < _elems; i++)
{
int prev_term = i - 1;
elem_word_len += element[i]->word_length();
// Order check:
if (order() && element[prev_term]->starttoken() >= element[i]->starttoken())
return false;
}
// Then check that within total limit:
if (((int)word_length() - (int)elem_word_len) > limit * (_elems - 1)) return false;
return true;
}
bool gtematch_cand::gtDistance(const MatchCandidate* m1, const MatchCandidate* m2) const
{
int m1d(m1->word_distance()), m2d(m2->word_distance());
return (m1d < m2d)
? true
: (m1d > m2d)
? false
: m1->startpos() < m2->startpos();
}
// A suitable comparator for MatchCandidates
bool gtematch_cand::operator()(const MatchCandidate* m1, const MatchCandidate* m2) const
{
// replace return m1->rank() > m2->rank();
// which does return (_elem_weight << 11) - (word_distance() << 8) - (_startpos >> 8);
return (m1->weight() > m2->weight())
? true
: (m1->weight() < m2->weight())
? false
: gtDistance(m1, m2);
}
|
