Arduino question?
Bruce Labitt
bruce.labitt at myfairpoint.net
Sat May 4 18:08:27 EDT 2019
Hi Paul,
Based on both your suggestion and a few of Bill Freeman's, I found the
problem. It basically was a problem with the scope of the #ifndef
statement. The problem was that if moreutils was compiled prior to
RunningMedian, then RunningMedian_h was defined. Due to an error in
placement of the #endif statement in RunningMedian.h, the whole body of
the header file was not read in. I changed the location of the #endif
statement in RunningMedian.h so that the body of the header would be
read. Problem solved.
So, it was simple... once it was pointed out. Thanks to you both.
I'll still be coming up to the MakerSpace on Monday. I'd like to check
out the setup. There is one in Nashua, just wondering how the
Manchester one compares.
Regards,
Bruce
On 5/3/19 7:34 PM, Bruce Labitt wrote:
> Hi Paul,
>
> It seems this is a variant of header hell that I am stewing in... At
> the moment the project is not up on github, as I was under some sort
> of delusion that this might eventually have some commercial value.
> I'm not so sure about that, but I have put a lot of effort into it so
> far...
>
> Arduino, does indeed do some funky stuff. If I could figure it out,
> I'd probably just write my own scripts to have things compile in the
> order I want. A makefile isn't hard. I just don't know all the crazy
> locations where Arduino stuff is stashed. Suppose I could poke about.
>
> Thanks for the tip on going through the #ifndef's. It's got to be in
> that neighborhood. I put in the #ifndef stuff to prevent multiple
> inclusions, which Arduino 1.8.9 IDE does not seem to like, the
> compiler squawked at me when I tried to do that. What is puzzling me
> is that even if I put the include #include "RunningMedian.h" in the
> file, I get "nearly" the same error, if I comment out the include.
>
> WITH include:
> #include "moreutils.h"
> *#include "RunningMedian.h"*
> home/bruce/Arduino/adcdmafftM4bruce/adcdmafft/adcdmafftm4/moreutils.ino:
> In function 'void doMedian(float*, float*, int)':
> moreutils:14:3: error: 'RunningMedian' was not declared in this scope
> RunningMedian samples = RunningMedian(medianlength);
> ^
> moreutils:14:17: error: expected ';' before 'samples'
> RunningMedian samples = RunningMedian(medianlength);
> ^
> moreutils:16:5: error: 'samples' was not declared in this scope
> samples.add(abuf[i]);
> ^
> exit status 1
> 'RunningMedian' was not declared in this scope
>
> WITHOUT include:
> #include "moreutils.h"
> /home/bruce/Arduino/adcdmafftM4bruce/adcdmafft/adcdmafftm4/moreutils.ino:
> In function 'void doMedian(float*, float*, int)':
> moreutils:14:3: error: 'RunningMedian' was not declared in this scope
> RunningMedian samples = RunningMedian(medianlength);
> ^
> moreutils:14:17: error: expected ';' before 'samples'
> RunningMedian samples = RunningMedian(medianlength);
> ^
> moreutils:16:5: error: 'samples' was not declared in this scope
> samples.add(abuf[i]);
> ^
> exit status 1
> 'RunningMedian' was not declared in this scope
>
> In both cases, the top of the file states #include moreutils.h. And
> inside of moreutils.h there is:
> #ifndef RunningMedian_h
> #define RunningMedian_h
> #include "RunningMedian.h"
> ...
> #endif
>
> It's got to be simple, it's got to be simple,... Wish I could see it
> what it was...
>
> I think, I'll come up on Monday. I haven't been to the Manchester
> Maker Space before.
>
> -Bruce
>
> On 5/3/19 6:50 PM, Paul Beaudet wrote:
>> The compiler is basically telling you the library is not imported or
>> has yet to be imported for whatever reason.
>>
>> The Arduinoy parts of the compilation process do some weird things
>> behind the scenes to reorder the code before compiling so that it
>> actually makes sense. Done with extra .ino files not named the same
>> as the project and functions after the main loop in said
>> project_name.ino file. Its possible for an Arduino developer/tinkerer
>> to declare a function after the two primary loops that use those
>> exact functions before actual deceleration because of this behavior.
>> Which breaks some expectations of hardened engineers I think. All in
>> the interest of catering to new people that would probably be
>> frustrated by a strict order of operation. If this behaviour
>> rearranges library functions before the "#include <mylibrary.h>",
>> that might give an unexpected result, but that doesn't seem to be
>> what's going on here. Under the hood strict order of operations still
>> exist. Arduino does ultimately use the gcc compiler, though maybe its
>> configured for that switcharoo magic, I didn't write the code, I've
>> just fallen into the traps.
>>
>> Extra .ino files in the same folder will be arranged before your main
>> .ino file, but it could be right before setup. Not sure, its been a
>> while since I've been regularly working with the Arduino, I could be
>> wrong. The .h file should be a bit more obvious as to where its
>> included. Bit more explicit
>>
>> Do you have a github link to your current project state Bruce?
>>
>> I'm looking back at an older project of mine, it looks like I was
>> doing something similar in this one
>> https://github.com/PaulBeaudet/SimpleBotSocketIo/blob/master/SimpleBotSocketIo.ino
>> I declared a timer library in a .h file then declared and instance of
>> it in the .ino fille. We demoed that bot at the mini makerfaire in
>> dover, I'm assuming it compiled and that should be something that's
>> doable
>>
>> I think you may want to inspect your #ifndef statement a bit more
>> closely to be sure the RunningMedian data type (library import) is
>> actually being instantiated after that pre-compilation logic runs its
>> course. Why do you need to prevent multiple instances from being
>> called? #import should only call one instance, right? Also, spelling,
>> that's normally my issue.. haha
>>
>> If you are still having trouble after spending more time with it,
>> feel free to stop by the Manchester Makerspace Monday during open
>> house (6pm-8pm). I'll likely be in giving tours, I can take a look
>> with you when tours simmer down.
>>
>> Cheers,
>> Paul Beaudet
>>
>> On Fri, May 3, 2019 at 4:57 PM Bruce Labitt
>> <bruce.labitt at myfairpoint.net <mailto:bruce.labitt at myfairpoint.net>>
>> wrote:
>>
>> Can I ask an Arduino/C/C++ question here? If not, where is a decent
>> place to ask? Full code is just under 50KB (unzipped).
>>
>> It's a "Variable was not declared in this scope" problem.
>> Basically, I'm
>> in over my head at the moment. I'm not a good structured
>> programmer -
>> so let's get that out of the way. I'm a hack, in the worst sense...
>>
>> Everything was working... when I had a huge file. I then
>> decided, wow,
>> this is a mess, lets break this up a bit into modules, so that it is
>> more supportable and debug-able (for myself). If anyone is remotely
>> interested, it is a homebrew radar based chronograph. I've got
>> most of
>> the pieces working (or at least it worked before I recently busted
>> things). The 100KHz sampling using DMA, the ping pong floating
>> point 1K
>> FFT's running in 'real' time, and some display stuff. Separately, I
>> have a live update of a tft screen (320x240) running with the FFT
>> output. I'm running on an ARM M4F processor, but using the Arduino
>> IDE. The Arduino way of doing things is a little confusing to
>> me, to be
>> honest. It hides a lot of things.
>>
>> Ok, here is the error.
>>
>> /home/bruce/Arduino/adcdmafftM4bruce/adcdmafft/adcdmafftm4/moreutils.ino:
>>
>> In function 'void doMedian(float*, float*, int)':
>> moreutils:14:3: error: 'RunningMedian' was not declared in this scope
>> RunningMedian samples = RunningMedian(medianlength);
>> ^
>> moreutils:14:17: error: expected ';' before 'samples'
>> RunningMedian samples = RunningMedian(medianlength);
>> ^
>> moreutils:16:5: error: 'samples' was not declared in this scope
>> samples.add(abuf[i]);
>> ^
>> exit status 1
>> 'RunningMedian' was not declared in this scope
>>
>> The code in moreutils.ino is:
>>
>> // additional processing
>> #include "moreutils.h"
>>
>> void doMedian( float abuf[], float runmed[], int medianlength) { //
>> needs work!
>>
>> RunningMedian samples = RunningMedian(medianlength);
>> for (int i=0; i< FFT_SIZE/2; i++) {
>> samples.add(abuf[i]);
>> if (i>medianlength-1) {
>> runmed[i-medianlength] = samples.getMedian();
>> // don't put value until the circ buffer is filled
>> }
>> }
>> for (int i= (FFT_SIZE/2 -medianlength-7); i< (FFT_SIZE/2); i++) {
>> runmed[i] = runmed[FFT_SIZE/4]; // hack for now
>> // at tail end of median there are some bizarre numbers. root
>> cause has not been
>> // determined, so we just fill the last samples from
>> 'something close'
>> }
>> }
>>
>> Inside of moreutils.h, is #include RunningMedian.h with an
>> #ifndef/#define/#include statement, to prevent multiple includes
>> of the
>> same file (RunningMedian.h).
>>
>> I'm really kind of confused as to where I need to do the declaration.
>>
>> In
>> https://github.com/RobTillaart/Arduino/blob/master/libraries/RunningMedian/examples/RunningMedian/RunningMedian.ino
>>
>> the declaration is simply done prior to setup. Snippet below
>>
>> #include <RunningMedian.h>
>>
>> RunningMedian samples = RunningMedian(5);
>>
>> RunningMedian samples2 = RunningMedian(9);
>>
>> void setup() {
>>
>> ...
>>
>> }
>>
>> void loop() {
>>
>> use samples here...
>>
>> }
>>
>> I'm sure this is trivial for most of you - but I'm both perplexed
>> and
>> stuck. If one of you kind souls could help me, I'd greatly
>> appreciate
>> it. I'd even travel to see someone if that would work out better.
>>
>> TIA, Bruce
>>
>> _______________________________________________
>> gnhlug-discuss mailing list
>> gnhlug-discuss at mail.gnhlug.org
>> <mailto:gnhlug-discuss at mail.gnhlug.org>
>> http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/
>>
>
>
> _______________________________________________
> gnhlug-discuss mailing list
> gnhlug-discuss at mail.gnhlug.org
> http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.gnhlug.org/pipermail/gnhlug-discuss/attachments/20190504/c2aa8f28/attachment-0001.html
More information about the gnhlug-discuss
mailing list