Qore DataProvider Module Reference  1.0.2
DataProviderPipeline.qc.dox.h
1 // -*- mode: c++; indent-tabs-mode: nil -*-
3 
25 // assume local scope for variables, do not use "$" signs
26 // require type definitions everywhere
28 // enable all warnings
29 // allow weak references
30 
32 namespace DataProvider {
36 const PS_ABORTED = "ABORTED";
38 
40 const PS_RUNNING = "RUNNING";
41 
43 const PS_IDLE = "IDLE";
45 
49 
53 const DPE_SKIP_DATA = "DPE-SKIP-DATA";
55 
57 public struct PipelineInfo {
60 
63 
65 
67  string status;
68 
71 
74 
76  bool bulk;
77 
80 
83 
85  float recs_per_sec;
86 };
87 
89 public struct PipelineOptionInfo {
91 
93  *code debug_log;
94 
96 
98  *code error_log;
99 
101 
103  *code info_log;
104 
106  *string name;
107 };
108 
111 
112 public:
114  int id;
115 
117  Condition cond();
118 
120  int queue_waiting = 0;
121 
123  int data_waiting = 0;
124 
126  list<auto> queue;
127 
129  int size;
130 
132  int tid;
133 
135 
139  list<auto> elems();
140 
143 
146 
147 
149  int getId();
150 
151 
153 
157  submit(auto record);
158 
159 
161  run(Counter run_cnt);
162 
163 
165 
167  waitDone();
168 
169 };
170 
172 
176 
177 public:
179  Mutex lck();
180 
182  bool stop_flag;
183 
186 
188  Counter cnt();
189 
190 protected:
192  string name;
193 
195  hash<string, PipelineQueue> pmap;
196 
198  bool do_bulk = True;
199 
201  bool locked = False;
202 
204  Sequence seq(1);
205 
207  list<hash<ExceptionInfo>> error_list;
208 
211 
214 
216  int record_count = 0;
217 
219  *code info_log;
220 
222  *code error_log;
223 
225  *code debug_log;
226 
227 public:
228 
230 
234  constructor(*hash<PipelineOptionInfo> opts);
235 
236 
238 
244  destructor();
245 
246 
248  string getName();
249 
250 
252 
255  append(AbstractDataProcessor processor);
256 
257 
259 
266  append(int id, AbstractDataProcessor processor);
267 
268 
270 
280  int appendQueue(int id);
281 
282 
284  bool isProcessing();
285 
286 
288  submit(auto _data);
289 
290 
292  submitData(AbstractIterator i);
293 
294 
297 
298 
301 
302 
304 
307  reset();
308 
309 
311 
314  waitDone();
315 
316 
318 
326  abort(*bool ignore_exceptions);
327 
328 
330 
334  hash<PipelineInfo> getInfo();
335 
336 
338  reportError(PipelineQueue queue, hash<ExceptionInfo> ex);
339 
340 
342  logInfo(string fmt);
343 
344 
346  logError(string fmt);
347 
348 
350  logDebug(string fmt);
351 
352 
354 
356 protected:
357  submitIntern(auto _data);
358 public:
359 
360 
362 
364 protected:
365  submitDataIntern(auto _data);
366 public:
367 
368 
370 
372 protected:
374 public:
375 
376 
378 
380 protected:
381  resetIntern();
382 public:
383 
384 
386 
388 protected:
390 public:
391 
392 
394 
396 protected:
398 public:
399 
400 
402 
404 protected:
406 public:
407 
408 
410 protected:
412 public:
413 
414 
416 protected:
417  stopIntern();
418 public:
419 
420 
422 protected:
424 public:
425 
426 };
427 };
DataProvider::DPE_SKIP_DATA
const DPE_SKIP_DATA
Skips processing a record.
Definition: DataProviderPipeline.qc.dox.h:53
DataProvider::DataProviderPipeline::logError
logError(string fmt)
Logs to the error log, if set.
DataProvider::DataProviderPipeline::start_time
date start_time
run start time
Definition: DataProviderPipeline.qc.dox.h:210
DataProvider::PipelineOptionInfo::name
*string name
the name of the pipeline for logging purposes; if this key is not included, a generic name will be ge...
Definition: DataProviderPipeline.qc.dox.h:106
DataProvider::PipelineOptionInfo::debug_log
*code debug_log
a closure or call reference for debug logging
Definition: DataProviderPipeline.qc.dox.h:93
DataProvider::DataProviderPipeline::pmap
hash< string, PipelineQueue > pmap
Hash of queues keyed by queue ID.
Definition: DataProviderPipeline.qc.dox.h:195
DataProvider::DataProviderPipeline::locked
bool locked
Locked flag.
Definition: DataProviderPipeline.qc.dox.h:201
DataProvider::DataProviderPipeline::name
string name
A descriptive name for logging purposes.
Definition: DataProviderPipeline.qc.dox.h:192
DataProvider::DataProviderPipeline::submit
submit(auto _data)
Submits data for processing.
DataProvider::DataProviderPipeline::stop_flag
bool stop_flag
Stop flag.
Definition: DataProviderPipeline.qc.dox.h:182
DataProvider::PipelineQueue::submit
submit(auto record)
Submits data for processing.
DataProvider::DataProviderPipeline::reset
reset()
Resets the pipeline.
DataProvider::PipelineQueue::getId
int getId()
Returns the pipeline ID.
DataProvider::PipelineInfo
Pipeline info.
Definition: DataProviderPipeline.qc.dox.h:57
DataProvider::PipelineQueue::run
run(Counter run_cnt)
Processing thread.
DataProvider::PipelineQueue::data_waiting
int data_waiting
Number of threads waiting on data.
Definition: DataProviderPipeline.qc.dox.h:123
DataProvider::PipelineInfo::start_time
*date start_time
Start of processing.
Definition: DataProviderPipeline.qc.dox.h:59
DataProvider::DataProviderPipeline::error_log
*code error_log
Error log closure; takes a single format string and then arguments for format placeholders.
Definition: DataProviderPipeline.qc.dox.h:222
DataProvider::DataProviderPipeline::debug_log
*code debug_log
Debug log closure; takes a single format string and then arguments for format placeholders.
Definition: DataProviderPipeline.qc.dox.h:225
DataProvider::PipelineInfo::status
string status
Pipeline status.
Definition: DataProviderPipeline.qc.dox.h:67
True
const True
DataProvider::DataProviderPipeline::append
append(AbstractDataProcessor processor)
Appends a data processor to the default pipeline.
DataProvider::DataProviderPipeline::getName
string getName()
Returns the pipeline name.
DataProvider::PipelineOptionInfo::error_log
*code error_log
error_log: a closure or call reference for error logging
Definition: DataProviderPipeline.qc.dox.h:98
DataProvider::PipelineQueue::cond
Condition cond()
Queue condition variable.
DataProvider::DataProviderPipeline::appendQueue
int appendQueue(int id)
Appends a new queue to an existing pipeline and returns the new queue ID.
DataProvider
Qore AbstractDataField class definition.
Definition: AbstractDataField.qc.dox.h:32
DataProvider::AbstractDataProcessor
Defines an abstract class for accepting data and outputting optionally transformed or filtered data.
Definition: AbstractDataProcessor.qc.dox.h:33
DataProvider::PS_RUNNING
const PS_RUNNING
Pipeline status: RUNNING.
Definition: DataProviderPipeline.qc.dox.h:40
DataProvider::DataProviderPipeline::isProcessing
bool isProcessing()
Returns True if the pipeline is processing data.
DataProvider::DataProviderPipeline::do_bulk
bool do_bulk
Bulk flag.
Definition: DataProviderPipeline.qc.dox.h:198
DataProvider::PipelineInfo::record_count
int record_count
Number of input records submitted.
Definition: DataProviderPipeline.qc.dox.h:70
DataProvider::PipelineOptionInfo::info_log
*code info_log
info_log: a closure or call reference for info logging
Definition: DataProviderPipeline.qc.dox.h:103
DataProvider::PipelineInfo::num_queues
int num_queues
Number of pipeline queues.
Definition: DataProviderPipeline.qc.dox.h:73
DataProvider::PipelineQueue::parent
DataProviderPipeline parent
Parent object.
Definition: DataProviderPipeline.qc.dox.h:142
DataProvider::DataProviderPipeline::stop_time
date stop_time
run stop time (set in waitDone())
Definition: DataProviderPipeline.qc.dox.h:213
DataProvider::DataProviderPipeline::abort
abort(*bool ignore_exceptions)
Aborts execution of a pipeline in progress.
DataProvider::PipelineQueue::elems
list< auto > elems()
Pipeline elements.
DataProvider::AbstractDataProviderRecordIterator
Defines the abstract class for data provider iterators; the destructor releases the iterator.
Definition: AbstractDataProviderRecordIterator.qc.dox.h:33
DataProvider::DataProviderPipeline::abort_flag
bool abort_flag
Abort flag.
Definition: DataProviderPipeline.qc.dox.h:185
DataProvider::PipelineOptionInfo
Pipeline option info.
Definition: DataProviderPipeline.qc.dox.h:89
DataProvider::PipelineQueue
Pipeline element.
Definition: DataProviderPipeline.qc.dox.h:110
DataProvider::DataProviderPipeline::error_list
list< hash< ExceptionInfo > > error_list
list of exceptions in pipelines
Definition: DataProviderPipeline.qc.dox.h:207
DataProvider::DataProviderPipeline::destructor
destructor()
Destroys the object.
DataProvider::PipelineQueue::queue_waiting
int queue_waiting
Number of threads waiting data to be removed from the queue.
Definition: DataProviderPipeline.qc.dox.h:120
DataProvider::DataProviderPipeline::waitDone
waitDone()
Waits for all queues in all pipelines to have processed remaining data.
DataProvider::AbstractDataProviderBulkRecordInterface
Defines the abstract class for data provider iterators; the destructor releases the iterator.
Definition: AbstractDataProviderBulkRecordInterface.qc.dox.h:33
DataProvider::DataProviderPipeline::submitData
submitData(AbstractIterator i)
Submits data for processing.
DataProvider::DataProviderPipeline::submitIntern
submitIntern(auto _data)
Submits data for processing.
DataProvider::PipelineInfo::stop_time
*date stop_time
Stop time for processing.
Definition: DataProviderPipeline.qc.dox.h:62
DataProvider::PipelineQueue::waitDone
waitDone()
Wait for the queue to be empty, then wait for all terminating pipelines to be empty.
DataProvider::PS_IDLE
const PS_IDLE
Pipeline status: IDLE.
Definition: DataProviderPipeline.qc.dox.h:43
DataProvider::PipelineQueue::size
int size
Maximum queue size.
Definition: DataProviderPipeline.qc.dox.h:129
DataProvider::PipelineInfo::duration_secs
float duration_secs
Total time processing end to end as a floating-point value in durationSecondsFloat.
Definition: DataProviderPipeline.qc.dox.h:82
DataProvider::DataProviderPipeline::reportError
reportError(PipelineQueue queue, hash< ExceptionInfo > ex)
Called from a pipeline queue object to report a fatal error durring processing.
DataProvider::DataProviderPipeline::checkLockedIntern
checkLockedIntern()
Throws an exception if the pipeline is locked.
DataProvider::PipelineQueue::id
int id
Pipeline ID.
Definition: DataProviderPipeline.qc.dox.h:114
DataProvider::DataProviderPipeline::info_log
*code info_log
Info log closure; takes a single format string and then arguments for format placeholders.
Definition: DataProviderPipeline.qc.dox.h:219
DataProvider::PipelineInfo::recs_per_sec
float recs_per_sec
Records processed per second end to end.
Definition: DataProviderPipeline.qc.dox.h:85
DataProvider::DataProviderPipeline::throwPipelineException
throwPipelineException()
Throws an exception if errors occured in background pipeline processing.
DataProvider::DataProviderPipeline::checkUpdatePipelineIntern
checkUpdatePipelineIntern(int id)
Check if the given queue exists.
False
const False
DataProvider::DataProviderPipeline::stopInternUnlocked
stopInternUnlocked()
Stops all background pipeline queues; lock must be held.
DataProvider::PipelineQueue::tid
int tid
TID of the background thread.
Definition: DataProviderPipeline.qc.dox.h:132
DataProvider::DataProviderPipeline::logDebug
logDebug(string fmt)
Logs to the debug log, if set.
DataProvider::DataProviderPipeline::checkSubmitIntern
checkSubmitIntern()
Throws an exception if the pipeline cannot be used; locks the pipeline for changes otherwise.
DataProvider::DataProviderPipeline::record_count
int record_count
Record count.
Definition: DataProviderPipeline.qc.dox.h:216
DataProvider::PipelineQueue::constructor
constructor(DataProviderPipeline parent, int id, int size)
Creates the object.
DataProvider::DataProviderPipeline::logInfo
logInfo(string fmt)
Logs to the info log, if set.
DataProvider::DataProviderPipeline::constructor
constructor(*hash< PipelineOptionInfo > opts)
Creates the object with the given input iterator.
DataProvider::DataProviderPipeline::getInfo
hash< PipelineInfo > getInfo()
Returns pipeline info.
DataProvider::DataProviderPipeline::seq
Sequence seq(1)
Pipeline ID sequence generator.
DataProvider::PipelineQueue::queue
list< auto > queue
Data queue.
Definition: DataProviderPipeline.qc.dox.h:126
DataProvider::DataProviderPipeline::cnt
Counter cnt()
Thread counter.
DataProvider::PipelineInfo::duration
date duration
Total time processing end to end.
Definition: DataProviderPipeline.qc.dox.h:79
date
date date(date dt)
DataProvider::DataProviderPipeline::submitDataIntern
submitDataIntern(auto _data)
Submits a single record for processing.
DataProvider::DataProviderPipeline
Defines a class for passing data through record processors.
Definition: DataProviderPipeline.qc.dox.h:175
DataProvider::DataProviderPipeline::submitBulkIntern
submitBulkIntern(AbstractDataProviderBulkRecordInterface i)
Submits bulk data for processing.
DataProvider::PS_ABORTED
const PS_ABORTED
Pipeline status: ABORTED.
Definition: DataProviderPipeline.qc.dox.h:37
DataProvider::DataProviderPipeline::lck
Mutex lck()
Atomic lock.
DataProvider::DataProviderPipeline::stopIntern
stopIntern()
Stops all background pipeline queues.
DataProvider::PipelineInfo::bulk
bool bulk
Flag that indicates if the pipeline is capable of bulk record processing.
Definition: DataProviderPipeline.qc.dox.h:76
DataProvider::DataProviderPipeline::resetIntern
resetIntern()
Resets the pipeline.