#BOOSTNOTE PYTHON GENERATOR#

Don't expect to findĪnything you can use in these directories. Most libraries place private headers in a subdirectory calledĭetail /, or aux_ /. ForĮxample, Boost.Python's aggregate header is boost / python.hpp. #includes all of the library's other headers. Some libraries have an “aggregate header” in boost / that The Python library's def.hpp header in boost / python / def.hpp. Most libraries' public headers live in a subdirectory ofīoost /, named after the library. Some older libraries and most very small libraries place all er_perf.The organization of Boost library headers isn't entirely uniform,īut most libraries follow a few patterns: I only have one example (coded by somebody else. I know from being subscribed to the github list, that you can't quite do everything using micropython.native that you can do using regular python, but there is still quite a bit you can do. I couldn't find any docs on micropython.native/viper. There was also a bit of a comparison done as part of the kickstarter campaign. You can search the forum for the word viper.

#BOOSTNOTE PYTHON GENERATOR#

mbler.htmlġ - the defailt - which produces bytecodeĢ - Use This generates assembler output instead of bytecodeģ - Use - This requires aseembler-ish source done in a python-esk syntax, but will be the fastest possible way of doing things.Ĥ - Use - I think this code generator is a bit more experimental and not fully supported yet. t=10#p4780 There is also some information here. There was a thread here about writing some code using some assembler (it has some assembler examples): Thanks for all the help, I really appreciate it! This has come as a surprise to me since it is the first time that I have found something that I could do on Arduino but cannot do on the pyboard. Or maybe use of interrupts would be quicker?

Is there any way of byte compiling the code to get it to execute faster? That only leaves a very short time to execute python function calls. In any event, I am wondering how I can set/un set a pin with enough precision to communicate on the ps/2 protocol: The clock period is less than 100 µs, the pin is read while the clock is low, and can have its value changed while the clock is high. It may be the cost of the call to pyb.micros() that is expensive? Just calling pyb.micros() twice costs +/- 10µs! So, if I subtract that from the times in the 1st experiment, I get switching times of around 40-50µs which is still a lot. I have done further experiments and have found quite some long times for code execution.

The times that are on the order of 50ms are way too high for the ps/2 protocol, which is really slow, so I must be doing something wrong.

time for a LOW output to go to HIGH output: 20.

time for a HIGH input to go to LOW output: 63.

time for LOW output to get PULLED HIGH as input: 51.

time for a LOW output to go to HIGH OUTPUT16.

time for a High input to go to LOW output: 16.

time for a LOW OUTPUT to get pulled HIGH as INPUT: 16.

ino suffix) you'll see the code.Īfter long hours, days, even weeks, I have discovered that the switching time for the pyboard seems to be much longer than that of the Arduino UNO.

If you look in the examples folder at the ps_mouse.pde (which should have a. I found a working example for Arduino here. Hello out there in the micropython world!Īs some of you know, I have been trying to implement the host side of the ps/2 mouse protocol on the pyboard.

Boostnote python