MINIFI-218: Support for GPSD integration #52
Conversation
libminifi/include/GetGPS.h
Outdated
|
|
||
| private: | ||
| //! Logger | ||
| Logger *_logger; |
There was a problem hiding this comment.
logger_ is the google cpp style guide approach. Same with all of these.
There was a problem hiding this comment.
No, I agree with you about this. I'll change them to make them follow the google guidelines. Best to start those new habits now =)
libminifi/src/gps/GetGPS.cpp
Outdated
| return; | ||
| } | ||
|
|
||
| for (;;) { |
There was a problem hiding this comment.
Could you wait on a condition in case this processor is forced shutdown.
| #include "../FlowFileRecord.h" | ||
|
|
||
|
|
||
| TEST_CASE("Test GPSD Create", "[TestGPSEventRecord]"){ |
There was a problem hiding this comment.
Would we be able to create some processor specific tests?
There was a problem hiding this comment.
certainly. I'm wondering the best way to do this at the moment. I want to use gpsfake for this but thinking through the best way to do so. Using gpsfake would require another system dependency since it is an extra GPSD service itself. That would be the most robust way for testing this however
There was a problem hiding this comment.
Might make sense to roll in a mocking framework? https://github.com/eranpeer/FakeIt seems similarly lightweight to catch. @phrocker any recommendations or preferences?
There was a problem hiding this comment.
@apiri and @phrocker I'm leaning on you guys for recommendations around the tests here. I was planning on using http://www.catb.org/gpsd/gpsfake.html
There was a problem hiding this comment.
yeah I initially approved because I wasn't sure there was a good way to test this. I didn't know about gpsfake, and would love if there was a way to roll that in. Woe is me thinking there isn't a good way to test something.
|
Could we also get travis updated to include this in the build by default? |
README.md
Outdated
| @@ -110,6 +118,9 @@ $ brew install cmake \ | |||
| leveldb \ | |||
| ossp-uuid \ | |||
| boost \ openssl | |||
|
|
|||
| # If building with GPS support | |||
| $ brew install gps | |||
There was a problem hiding this comment.
Its actually is just gps for OS X. Its the gps.h headers and such that is needed. However it might make sense to also install gpsd so I'll add that
There was a problem hiding this comment.
Hmm... gps doesn't show up as a package for me in brew.
brew search gps
bgpstream gpsbabel gpsd ✔ gpsim osm-gps-map
Caskroom/cask/gpsdump
Am I overlooking something?
libminifi/src/gps/GetGPS.cpp
Outdated
| #include "ProcessContext.h" | ||
| #include "ProcessSession.h" | ||
|
|
||
| #include <libgpsmm.h> |
There was a problem hiding this comment.
Having issues on OS X 10.11 with the following:
In file included from /Users/apiri/Development/code/nifi-minifi-cpp/libminifi/src/gps/GetGPS.cpp:43:
In file included from /usr/local/include/libgpsmm.h:12:
/usr/local/include/gps.h:1887:8: error: definition of type 'policy_t' conflicts with
typedef of the same name
struct policy_t {
^
/usr/include/mach/policy.h:79:17: note: 'policy_t' declared here
typedef int policy_t;
^
In file included from /Users/apiri/Development/code/nifi-minifi-cpp/libminifi/src/gps/GetGPS.cpp:43:
In file included from /usr/local/include/libgpsmm.h:12:
/usr/local/include/gps.h:2014:12: error: elaborated type refers to a typedef
struct policy_t policy; /* our listening policy */
^
/usr/include/mach/policy.h:79:17: note: declared here
typedef int policy_t;
^
2 errors generated.
make[2]: *** [libminifi/CMakeFiles/minifi_gps.dir/src/gps/GetGPS.cpp.o] Error 1
README.md
Outdated
| @@ -91,6 +93,9 @@ $ yum install cmake \ | |||
| leveldb-devel leveldb \ | |||
| libuuid libuuid-devel \ | |||
| boost-devel \ libssl-dev | |||
|
|
|||
| # If building with GPS support | |||
| $ yum install libgps-devel | |||
There was a problem hiding this comment.
Believe this should gpsd-devel for yum based distros
libminifi/src/FlowController.cpp
Outdated
| @@ -289,6 +289,12 @@ Processor *FlowControllerImpl::createProcessor(std::string name, uuid_t uuid) { | |||
| return NULL; | |||
| } | |||
|
|
|||
| #ifdef BUILD_GPS | |||
There was a problem hiding this comment.
This was not very smart of me ... this will cause an error to be thrown that the processor of type GetGPS does not exist at runtime because the last else condition will be triggered before this. I need to move this higher up the condition before the last else.
jdye64
changed the title
|
This poor PR has been neglected. I think that now @phrocker PR for extensions has been committed its a good time to make another pass at this and place it in that extensions format |
|
Should this PR be closed? I believe we got gpsd in with another recent commit. |
|
f09ad61 is in master |
|
@achristianson yes per the comment nine days ago. I've spoken to Jeremy offline. |
|
Sorry for the late response here folks. I'm going to close this issue because as stated it has already been taken care of elsewhere. |
No description provided.