Parent: [be9be8] (diff)

Download this file

plgwithslave.cxx    715 lines (653 with data), 23.6 kB

  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
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
/* Copyright (C) 2016 J.F.Dockes
* 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.,
* 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*/
#include "config.h"
#include "plgwithslave.hxx"
#include <string>
#include <vector>
#include <sstream>
#include <functional>
#include <memory>
#include <string.h>
#include <fcntl.h>
#include <upnp/upnp.h>
#include <json/json.h>
#include <libupnpp/log.hxx>
#include "cmdtalk.h"
#include "pathut.h"
#include "smallut.h"
#include "conftree.h"
#include "sysvshm.h"
#include "main.hxx"
#include "streamproxy.h"
#include "netfetch.h"
#include "curlfetch.h"
#ifdef ENABLE_SPOTIFY
#include "spotify/spotiproxy.h"
#endif
using namespace std;
using namespace std::placeholders;
//using json = nlohmann::json;
using namespace UPnPProvider;
class StreamHandle {
public:
StreamHandle(PlgWithSlave::Internal *plg) {
}
~StreamHandle() {
clear();
}
void clear() {
plg = 0;
path.clear();
media_url.clear();
len = 0;
opentime = 0;
}
PlgWithSlave::Internal *plg;
string path;
string media_url;
long long len;
time_t opentime;
};
// Timeout seconds for reading data from plugins. Be generous because
// streaming services are sometimes slow, but we don't want to be
// stuck forever.
static const int read_timeout(60);
class PlgWithSlave::Internal {
public:
Internal(PlgWithSlave *_plg, const string& hst,
int prt, const string& pp)
: plg(_plg), cmd(read_timeout), upnphost(hst),
upnpport(prt), pathprefix(pp), laststream(this) {
string val;
if (g_config->get("plgproxymethod", val) && !val.compare("proxy")) {
doingproxy = true;
}
#ifdef ENABLE_SPOTIFY
if (!plg->getname().compare("spotify")) {
g_config->get("spotifyuser", user);
g_config->get("spotifypass", password);
string cachedir = path_cat(g_cachedir, "spotify");
// Make sure doingproxy is set, independantly of the
// config variable, which is only useful for Qobuz et al
doingproxy = true;
SpotiProxy::setParams(user, password, cachedir, cachedir);
}
#endif
}
bool doproxy() {
return doingproxy;
}
bool maybeStartCmd();
PlgWithSlave *plg;
CmdTalk cmd;
// Upnp Host and port. This would only be used to generate URLs
// *if* we were using the libupnp miniserver. We currently use
// microhttp because it can do redirects
string upnphost;
int upnpport;
// path prefix (this is used by upmpdcli that gets it for us).
string pathprefix;
bool doingproxy{false};
// This is only used by spotify (also needs login in the c++
// streamer in addition to python). We could create a derived
// class, but seems simpler this way.
string user;
string password;
// Cached uri translation
StreamHandle laststream;
};
// HTTP Proxy/Redirect handler
static StreamProxy *o_proxy;
StreamProxy::UrlTransReturn translateurl(
CDPluginServices *cdsrv,
std::string& url,
const std::unordered_map<std::string, std::string>& querymap,
std::unique_ptr<NetFetch>& fetcher
)
{
LOGDEB("PlgWithSlave::translateurl: url " << url << endl);
PlgWithSlave *realplg =
dynamic_cast<PlgWithSlave*>(cdsrv->getpluginforpath(url));
if (nullptr == realplg) {
LOGERR("PlgWithSlave::translateurl: no plugin for path ["<<url<< endl);
return StreamProxy::Error;
}
string path(url);
// The streaming services plugins set a trackId parameter in the
// URIs. This gets parsed out by mhttpd. We rebuild a full url
// which we pass to them for translation (they will extract the
// trackid and use it, the rest of the path is bogus).
// The uprcl module has a real path and no trackid. Handle both cases
const auto it = querymap.find("trackId");
if (it != querymap.end() && !it->second.empty()) {
path += string("?version=1&trackId=") + it->second;
}
// Translate to Tidal/Qobuz etc real temporary URL
url = realplg->get_media_url(path);
if (url.empty()) {
LOGERR("answer_to_connection: no media_uri for: " << url << endl);
return StreamProxy::Error;
}
StreamProxy::UrlTransReturn method = realplg->doproxy() ?
StreamProxy::Proxy : StreamProxy::Redirect;
if (method == StreamProxy::Proxy) {
if (!realplg->getname().compare("spotify")) {
#ifdef ENABLE_SPOTIFY
fetcher = std::move(std::unique_ptr<NetFetch>(new SpotiFetch(url)));
#else
LOGERR("Spotify URL but Spotify not supported by build\n");
return StreamProxy::Error;
#endif
} else {
fetcher = std::move(std::unique_ptr<NetFetch>(new CurlFetch(url)));
}
}
return method;
}
// Static
bool PlgWithSlave::startPluginCmd(CmdTalk& cmd, const string& appname,
const string& host, unsigned int port,
const string& pathpref)
{
string pythonpath = string("PYTHONPATH=") +
path_cat(g_datadir, "cdplugins") + ":" +
path_cat(g_datadir, "cdplugins/pycommon") + ":" +
path_cat(g_datadir, "cdplugins/" + appname);
string configname = string("UPMPD_CONFIG=") + g_configfilename;
stringstream ss;
ss << host << ":" << port;
string hostport = string("UPMPD_HTTPHOSTPORT=") + ss.str();
string pp = string("UPMPD_PATHPREFIX=") + pathpref;
string exepath = path_cat(g_datadir, "cdplugins");
exepath = path_cat(exepath, appname);
exepath = path_cat(exepath, appname + "-app" + ".py");
if (!cmd.startCmd(exepath, {/*args*/},
/* env */ {pythonpath, configname, hostport, pp})) {
LOGERR("PlgWithSlave::maybeStartCmd: startCmd failed\n");
return false;
}
return true;
}
// Static
bool PlgWithSlave::maybeStartProxy(CDPluginServices *cdsrv)
{
if (nullptr == o_proxy) {
int port = CDPluginServices::microhttpport();
o_proxy = new StreamProxy(
port,
std::bind(&translateurl, cdsrv, _1, _2, _3));
if (nullptr == o_proxy) {
LOGERR("PlgWithSlave: Proxy creation failed\n");
return false;
}
}
return true;
}
// Called once for starting the Python program and do other initialization.
bool PlgWithSlave::Internal::maybeStartCmd()
{
if (cmd.running()) {
LOGDEB1("PlgWithSlave::maybeStartCmd: already running\n");
return true;
}
if (!maybeStartProxy(this->plg->m_services)) {
LOGDEB1("PlgWithSlave::maybeStartCmd: maybeStartMHD failed\n");
return false;
}
int port = CDPluginServices::microhttpport();
if (!startPluginCmd(cmd, plg->m_name, upnphost, port, pathprefix)) {
LOGDEB1("PlgWithSlave::maybeStartCmd: startPluginCmd failed\n");
return false;
}
// If the creds have been set in shared mem, login at once, else
// the plugin will try later from file config data
LockableShmSeg seg(ohcreds_segpath, ohcreds_segid, ohcreds_segsize);
if (seg.ok()) {
LockableShmSeg::Accessor access(seg);
char *cp = (char *)(access.getseg());
string data(cp);
LOGDEB1("PlgWithSlave::maybeStartCmd: segment content [" << data << "]\n");
ConfSimple credsconf(data, true);
string user, password;
if (credsconf.get(plg->m_name + "user", user) &&
credsconf.get(plg->m_name + "pass", password)) {
unordered_map<string,string> res;
if (!cmd.callproc("login", {{"user", user}, {"password", password}}, res)) {
LOGINF("PlgWithSlave::maybeStartCmd: tried login but failed for " <<
plg->m_name);
}
}
} else {
LOGDEB0("PlgWithSlave::maybeStartCmd: shm attach failed (probably ok)\n");
}
return true;
}
bool PlgWithSlave::startInit()
{
return m && m->maybeStartCmd();
}
// Translate the slave-generated HTTP URL (based on the trackid), to
// an actual temporary service (e.g. tidal one), which will be an HTTP
// URL pointing to either an AAC or a FLAC stream.
// Older versions of this module handled AAC FLV transported over
// RTMP, apparently because of the use of a different API key. Look up
// the git history if you need this again.
// The Python code calls the service to translate the trackid to a temp
// URL. We cache the result for a few seconds to avoid multiple calls
// to tidal.
string PlgWithSlave::get_media_url(const string& path)
{
LOGDEB0("PlgWithSlave::get_media_url: " << path << endl);
if (!m->maybeStartCmd()) {
return string();
}
time_t now = time(0);
if (m->laststream.path.compare(path) ||
(now - m->laststream.opentime > 10)) {
unordered_map<string, string> res;
if (!m->cmd.callproc("trackuri", {{"path", path}}, res)) {
LOGERR("PlgWithSlave::get_media_url: slave failure\n");
return string();
}
auto it = res.find("media_url");
if (it == res.end()) {
LOGERR("PlgWithSlave::get_media_url: no media url in result\n");
return string();
}
m->laststream.clear();
m->laststream.path = path;
m->laststream.media_url = it->second;
m->laststream.opentime = now;
}
LOGDEB("PlgWithSlave: media url [" << m->laststream.media_url << "]\n");
return m->laststream.media_url;
}
PlgWithSlave::PlgWithSlave(const string& name, CDPluginServices *services)
: CDPlugin(name, services)
{
m = new Internal(this,
services->getupnpaddr(this),
services->getupnpport(this),
services->getpathprefix(this));
}
PlgWithSlave::~PlgWithSlave()
{
delete m;
}
bool PlgWithSlave::doproxy()
{
return m->doproxy();
}
static void catstring(string& dest, const string& s2)
{
if (s2.empty()) {
return;
}
if (dest.empty()) {
dest = s2;
} else {
dest += string(" ") + s2;
}
}
static int resultToEntries(const string& encoded, int stidx, int cnt,
vector<UpSong>& entries)
{
Json::Value decoded;
istringstream input(encoded);
input >> decoded;
LOGDEB0("PlgWithSlave::results: got " << decoded.size() << " entries \n");
LOGDEB1("PlgWithSlave::results: undecoded: " << decoded.dump() << endl);
bool dolimit = cnt > 0;
for (unsigned int i = stidx; i < decoded.size(); i++) {
#define JSONTOUPS(fld, nm) {catstring(song.fld, \
decoded[i].get(#nm, "").asString());}
if (dolimit && --cnt < 0) {
break;
}
UpSong song;
JSONTOUPS(id, id);
JSONTOUPS(parentid, pid);
JSONTOUPS(title, tt);
JSONTOUPS(artUri, upnp:albumArtURI);
JSONTOUPS(artist, upnp:artist);
JSONTOUPS(upnpClass, upnp:class);
JSONTOUPS(date, dc:date)
JSONTOUPS(date, releasedate)
// tp is container ("ct") or item ("it")
string stp = decoded[i].get("tp", "").asString();
if (!stp.compare("ct")) {
song.iscontainer = true;
string ss = decoded[i].get("searchable", "").asString();
if (!ss.empty()) {
song.searchable = stringToBool(ss);
}
} else if (!stp.compare("it")) {
song.iscontainer = false;
JSONTOUPS(uri, uri);
JSONTOUPS(artist, dc:creator);
JSONTOUPS(genre, upnp:genre);
JSONTOUPS(album, upnp:album);
JSONTOUPS(tracknum, upnp:originalTrackNumber);
JSONTOUPS(mime, res:mime);
string ss = decoded[i].get("duration", "").asString();
if (!ss.empty()) {
song.duration_secs = atoi(ss.c_str());
}
ss = decoded[i].get("res:size", "").asString();
if (!ss.empty()) {
song.size = atoll(ss.c_str());
}
ss = decoded[i].get("res:bitrate", "").asString();
if (!ss.empty()) {
song.bitrate = atoi(ss.c_str());
}
ss = decoded[i].get("res:samplefreq", "").asString();
if (!ss.empty()) {
song.samplefreq = atoi(ss.c_str());
}
ss = decoded[i].get("res:bitsPerSample", "").asString();
if (!ss.empty()) {
song.bitsPerSample = atoi(ss.c_str());
}
ss = decoded[i].get("res:channels", "").asString();
if (!ss.empty()) {
song.channels = atoi(ss.c_str());
}
} else {
LOGERR("PlgWithSlave::result: bad type in entry: " << stp <<
"(title: " << song.title << ")\n");
continue;
}
LOGDEB1("PlgWitSlave::result: pushing: " << song.dump() << endl);
entries.push_back(song);
}
// We return the total match size, the count of actually returned
// entries can be obtained from the vector
return decoded.size();
}
class ContentCacheEntry {
public:
ContentCacheEntry()
: m_time(time(0)) {
}
int toResult(const string& classfilter, int stidx, int cnt,
vector<UpSong>& entries) const;
time_t m_time;
vector<UpSong> m_results;
};
int ContentCacheEntry::toResult(const string& classfilter, int stidx, int cnt,
vector<UpSong>& entries) const
{
const vector<UpSong>& res = m_results;
LOGDEB0("searchCacheEntryToResult: filter " << classfilter << " start " <<
stidx << " cnt " << cnt << " res.size " << res.size() << endl);
entries.reserve(cnt);
int total = 0;
for (unsigned int i = 0; i < res.size(); i++) {
if (!classfilter.empty() && res[i].upnpClass.find(classfilter) != 0) {
continue;
}
total++;
if (stidx > int(i)) {
continue;
}
if (cnt && int(entries.size()) >= cnt) {
break;
}
LOGDEB1("ContentCacheEntry::toResult: pushing class " <<
res[i].upnpClass << " tt " << res[i].title << endl);
entries.push_back(res[i]);
}
return res.size();
}
class ContentCache {
public:
ContentCache(int retention_secs = 300);
ContentCacheEntry *get(const string& query);
void set(const string& query, ContentCacheEntry &entry);
void purge();
private:
time_t m_lastpurge;
int m_retention_secs;
unordered_map<string, ContentCacheEntry> m_cache;
};
ContentCache::ContentCache(int retention_secs)
: m_lastpurge(time(0)), m_retention_secs(retention_secs)
{
}
void ContentCache::purge()
{
time_t now(time(0));
if (now - m_lastpurge < 5) {
return;
}
for (auto it = m_cache.begin(); it != m_cache.end(); ) {
if (now - it->second.m_time > m_retention_secs) {
LOGDEB0("ContentCache::purge: erasing " << it->first << endl);
it = m_cache.erase(it);
} else {
it++;
}
}
m_lastpurge = now;
}
ContentCacheEntry *ContentCache::get(const string& key)
{
purge();
auto it = m_cache.find(key);
if (it != m_cache.end()) {
LOGDEB0("ContentCache::get: found " << key << endl);
// we return a copy of the vector. Make our multi-access life simpler...
return new ContentCacheEntry(it->second);
}
LOGDEB0("ContentCache::get: not found " << key << endl);
return nullptr;
}
void ContentCache::set(const string& key, ContentCacheEntry &entry)
{
LOGDEB0("ContentCache::set: " << key << endl);
m_cache[key] = entry;
}
// Cache for searches
static ContentCache o_scache(300);
// Cache for browsing
static ContentCache o_bcache(180);
// Better return a bogus informative entry than an outright error:
static int errorEntries(const string& pid, vector<UpSong>& entries)
{
entries.push_back(
UpSong::item(pid + "$bogus", pid,
"Service login or communication failure"));
return 1;
}
// Note that the offset and count don't get to the plugin for
// now. Plugins just return a (plugin-dependant) fixed number of
// entries from offset 0, which we cache. There is no good reason for
// this, beyond the fact that we have to cap the entry count anyway,
// else the CP is going to read to the end which might be
// reaaaaalllllyyyy long.
int PlgWithSlave::browse(const string& objid, int stidx, int cnt,
vector<UpSong>& entries,
const vector<string>& sortcrits,
BrowseFlag flg)
{
LOGDEB1("PlgWithSlave::browse\n");
entries.clear();
if (!m->maybeStartCmd()) {
return errorEntries(objid, entries);
}
string sbflg;
switch (flg) {
case CDPlugin::BFMeta:
sbflg = "meta";
break;
case CDPlugin::BFChildren:
default:
sbflg = "children";
break;
}
string cachekey(m_name + ":" + objid);
if (flg == CDPlugin::BFChildren) {
// Check cache
ContentCacheEntry *cep;
if ((cep = o_bcache.get(cachekey)) != nullptr) {
int total = cep->toResult("", stidx, cnt, entries);
delete cep;
return total;
}
}
unordered_map<string, string> res;
if (!m->cmd.callproc("browse", {{"objid", objid}, {"flag", sbflg}}, res)) {
LOGERR("PlgWithSlave::browse: slave failure\n");
return errorEntries(objid, entries);
}
auto ite = res.find("entries");
if (ite == res.end()) {
LOGERR("PlgWithSlave::browse: no entries returned\n");
return errorEntries(objid, entries);
}
bool nocache = false;
auto itc = res.find("nocache");
if (itc != res.end()) {
nocache = stringToBool(itc->second);
}
if (flg == CDPlugin::BFChildren) {
ContentCacheEntry e;
resultToEntries(ite->second, 0, 0, e.m_results);
if (!nocache) {
o_bcache.set(cachekey, e);
}
return e.toResult("", stidx, cnt, entries);
} else {
return resultToEntries(ite->second, stidx, cnt, entries);
}
}
// Note that the offset and count don't get to the plugin for
// now. Plugins just return a (plugin-dependant) fixed number of
// entries from offset 0, which we cache. There is no good reason for
// this, beyond the fact that we have to cap the entry count anyway,
// else the CP is going to read to the end which might be
// reaaaaalllllyyyy long.
int PlgWithSlave::search(const string& ctid, int stidx, int cnt,
const string& searchstr,
vector<UpSong>& entries,
const vector<string>& sortcrits)
{
LOGDEB("PlgWithSlave::search: [" << searchstr << "]\n");
entries.clear();
if (!m->maybeStartCmd()) {
return errorEntries(ctid, entries);
}
// Computing a pre-cooked query. For simple-minded plugins.
// Note that none of the qobuz/gmusic/tidal plugins actually use
// the slavefield part (defining in what field the term should
// match).
//
// Ok, so the upnp query language is quite powerful, but us, not
// so much. We get rid of parenthesis and then try to find the
// first searchExp on a field we can handle, pretend the operator
// is "contains" and just do it. I so don't want to implement a
// parser for the query language when the services don't support
// anything complicated anyway, and the users don't even want it...
string ss;
neutchars(searchstr, ss, "()");
// The search had better be space-separated. no
// upnp:artist="beatles" for you
vector<string> vs;
stringToStrings(ss, vs);
// The sequence can now be either [field, op, value], or
// [field, op, value, and/or, field, op, value,...]
if ((vs.size() + 1) % 4 != 0) {
LOGERR("PlgWithSlave::search: bad search string: [" << searchstr <<
"]\n");
return errorEntries(ctid, entries);
}
string slavefield;
string value;
string classfilter;
string objkind;
for (unsigned int i = 0; i < vs.size()-2; i += 4) {
const string& upnpproperty = vs[i];
LOGDEB("PlgWithSlave::search:clause: " << vs[i] << " " << vs[i+1] <<
" " << vs[i+2] << endl);
if (!upnpproperty.compare("upnp:class")) {
// This defines -what- we are looking for (track/album/artist)
const string& what(vs[i+2]);
if (beginswith(what, "object.item")) {
objkind = "track";
} else if (beginswith(what, "object.container.person")) {
objkind = "artist";
} else if (beginswith(what, "object.container.musicAlbum") ||
beginswith(what, "object.container.album")) {
objkind = "album";
} else if (beginswith(what, "object.container.playlistContainer")
|| beginswith(what, "object.container.playlist")) {
objkind = "playlist";
}
classfilter = what;
} else if (!upnpproperty.compare("upnp:artist") ||
!upnpproperty.compare("dc:author")) {
slavefield = "artist";
value = vs[i+2];
break;
} else if (!upnpproperty.compare("upnp:album")) {
slavefield = "album";
value = vs[i+2];
break;
} else if (!upnpproperty.compare("dc:title")) {
slavefield = "track";
value = vs[i+2];
break;
}
}
// In cache ?
ContentCacheEntry *cep;
string cachekey(m_name + ":" + ctid + ":" + searchstr);
if ((cep = o_scache.get(cachekey)) != nullptr) {
int total = cep->toResult(classfilter, stidx, cnt, entries);
delete cep;
return total;
}
// Run query
unordered_map<string, string> res;
if (!m->cmd.callproc("search", {
{"objid", ctid},
{"objkind", objkind},
{"origsearch", searchstr},
{"field", slavefield},
{"value", value} }, res)) {
LOGERR("PlgWithSlave::search: slave failure\n");
return errorEntries(ctid, entries);
}
auto ite = res.find("entries");
if (ite == res.end()) {
LOGERR("PlgWithSlave::search: no entries returned\n");
return errorEntries(ctid, entries);
}
bool nocache = false;
auto itc = res.find("nocache");
if (itc != res.end()) {
nocache = stringToBool(itc->second);
}
// Convert the whole set and store in cache
ContentCacheEntry e;
resultToEntries(ite->second, 0, 0, e.m_results);
if (!nocache) {
o_scache.set(cachekey, e);
}
return e.toResult(classfilter, stidx, cnt, entries);
}