"FP2" is a 16-bit data type that is defined to store values from +/-0.001 to +/-7999 (up to 4 significant digits).
NOTE: This data type is not my invention; it is already used on numerous embedded devices.
This library provides functions to convert these data types to and
from strings, see fp2conv.h for details. The data type is defined
as follows:
| bit 15 (MSB) | bits 14 - 13 | bits 12 - 0 (LSB) |
|---|---|---|
| sign (0=pos, 1=neg) | decimal position | significand |
The meanings of bits 14-13 (decimal position) are defined as follows:
| bit 14 | bit 13 | dec. pos. | value range |
|---|---|---|---|
| 0 | 0 | XXXX. |
+/-800 to +/-7999. |
| 0 | 1 | XXX.X |
+/- 80 to +/-799.9 |
| 1 | 0 | XX.XX |
+/- 8 to +/-79.99 |
| 1 | 1 | X.XXX |
0 to +/-7.999 |
Note that identical values can be represented in different ways. This
library uses the above value range to decide the storage format, thus
for example the integer 1 is stored as 0x63E8, even though it could
also be stored as 0x4064, 0x200A, or 0x0001. As one exception,
zero (0) is stored as 0x0000. Note the conversion from FP2 to string
can handle all of the aforementioned values.
Though theoretically 13 bits can store values from 0 to 8191, this data type defines the maximum value at 7999; in addition, the following special values are defined:
| Value | FP2 bytes |
|---|---|
| +Inf | 0x1FFF |
| -Inf | 0x9FFF |
| NaN | 0x9FFE |
| Fail* | 0xFFFF |
* This is not part of the original specification, but as far as I can tell this value should never occur otherwise, so this code uses it to indicate that the conversion could not be performed.
Using the provided Makefile, you can simply run make to build
libfp2conv.so, and make test to build and run test_fp2conv.c.
The library and headers can be installed via sudo make install.
Tested on Linux with gcc 7.5.0.
The fp2conv.c / fp2conv.h files were tested to be working on an
Arduino MKR series board with a SAMD21 processor
using the Arduino IDE 2.0.0.
The source files can be installed to your local Arduino library path
with make ardulibinst, assuming the default environment on Linux.
Simple Python bindings are provided in pyfp2conv.py with tests in tests/.
Tested on Python 3.10 on Linux.
Simple Perl FFI bindings are provided in lib/Data/FP2.pm with tests in t/.
The CPAN module FFI::Platypus is required.
Tested on Perl 5.34 and 5.28 on Linux.
Note: If FFI::Platypus fails to build due to a missing xlocale.h,
a workaround is: cd /usr/include && sudo ln -si locale.h xlocale.h
All tests can be run via the included shell script prove.sh.
Copyright (c) 2022 Hauke Daempfling [email protected] at the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB), Berlin, Germany, https://www.igb-berlin.de/
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 3 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, see https://www.gnu.org/licenses/.