Commit Graph

95 Commits

Author SHA1 Message Date
Damien George 6a445b60fa py/lexer: Add support for underscores in numeric literals.
This is a very convenient feature introduced in Python 3.6 by PEP 515.
2018-06-12 12:17:43 +10:00
Damien George 1ad0013dec tests: Add some tests for bigint hash, float hash and float parsing.
Following outcome of recent fuzz testing and sanitizing by @jepler.
2018-05-21 13:05:40 +10:00
Damien George d2c1db1e5c tests/float/float_parse: Allow test to run on 32-bit archs.
Printing of uPy floats can differ by the floating-point precision on
different architectures (eg 64-bit vs 32-bit x86), so it's not possible to
using printing of floats in some parts of this test.  Instead we can just
check for equivalence with what is known to be the correct answer.
2018-05-11 13:51:18 +10:00
Damien George 955ee6477f py/formatfloat: Fix case where floats could render with negative digits.
Prior to this patch, some architectures (eg unix x86) could render floats
with "negative" digits, like ")".  For example, '%.23e' % 1e-80 would come
out as "1.0000000000000000/)/(,*0e-80".  This patch fixes the known cases.
2018-03-01 17:00:02 +11:00
Damien George 7b050fa76c py/formatfloat: Fix case where floats could render with a ":" character.
Prior to this patch, some architectures (eg unix x86) could render floats
with a ":" character in them, eg 1e+39 would come out as ":e+38" (":" is
just after "9" in ASCII so this is like 10e+38).  This patch fixes some of
these cases.
2018-03-01 16:02:59 +11:00
Damien George bc12eca461 py/formatfloat: Fix rounding of %f format with edge-case FP values.
Prior to this patch the %f formatting of some FP values could be off by up
to 1, eg '%.0f' % 123 would return "122" (unix x64).  Depending on the FP
precision (single vs double) certain numbers would format correctly, but
others wolud not.  This patch should fix all cases of rounding for %f.
2018-03-01 15:51:03 +11:00
Damien George 6dad088569 tests/float: Adjust float-parsing tests to pass with only a small error.
Float parsing (both single and double precision) may have a relative error
of order the floating point precision, so adjust tests to take this into
account by not printing all of the digits of the answer.
2018-02-26 15:54:03 +11:00
Damien George bbb08431f3 py/objfloat: Fix case of raising 0 to -infinity.
It was raising an exception but it should return infinity.
2018-02-08 14:35:43 +11:00
Damien George b75cb8392b py/parsenum: Fix parsing of floats that are close to subnormal.
Prior to this patch, a float literal that was close to subnormal would
have a loss of precision when parsed.  The worst case was something like
float('10000000000000000000e-326') which returned 0.0.
2018-02-08 14:02:50 +11:00
Damien George 7cae17fac7 tests/float/builtin_float_hash: Add test to improve objfloat.c coverage. 2017-12-19 14:50:33 +11:00
Damien George 84895f1a21 py/parsenum: Improve parsing of floating point numbers.
This patch improves parsing of floating point numbers by converting all the
digits (integer and fractional) together into a number 1 or greater, and
then applying the correct power of 10 at the very end.  In particular the
multiple "multiply by 0.1" operations to build a fraction are now combined
together and applied at the same time as the exponent, at the very end.

This helps to retain precision during parsing of floats, and also includes
a check that the number doesn't overflow during the parsing.  One benefit
is that a float will have the same value no matter where the decimal point
is located, eg 1.23 == 123e-2.
2017-11-27 12:51:52 +11:00
Damien George a07fc5b640 py/objfloat: Allow float() to parse anything with the buffer protocol.
This generalises and simplifies the code and follows CPython behaviour.
2017-11-21 15:01:38 +11:00
Damien George 08a196697c py/formatfloat: Don't print the negative sign of a NaN value.
NaN may have the sign bit set but it has no meaning, so don't print it out.
2017-10-10 16:01:13 +11:00
Damien George 25e140652b py/modmath: Add full checks for math domain errors.
This patch changes how most of the plain math functions are implemented:
there are now two generic math wrapper functions that take a pointer to a
math function (like sin, cos) and perform the necessary conversion to and
from MicroPython types.  This helps to reduce code size.  The generic
functions can also check for math domain errors in a generic way, by
testing if the result is NaN or infinity combined with finite inputs.

The result is that, with this patch, all math functions now have full
domain error checking (even gamma and lgamma) and code size has decreased
for most ports.  Code size changes in bytes for those with the math module
are:

   unix x64:  -432
unix nanbox:  -792
      stm32:   -88
    esp8266:   +12

Tests are also added to check domain errors are handled correctly.
2017-10-10 15:57:45 +11:00
Damien George bdc6e86e07 py/objfloat: Support raising a negative number to a fractional power.
This returns a complex number, following CPython behaviour.  For ports that
don't have complex numbers enabled this will raise a ValueError which gives
a fail-safe for scripts that were written assuming complex numbers exist.
2017-09-26 12:57:51 +10:00
Damien George d4b75f6b68 py/obj: Fix comparison of float/complex NaN with itself.
IEEE floating point is specified such that a comparison of NaN with itself
returns false, and Python respects these semantics.  This patch makes uPy
also have these semantics.  The fix has a minor impact on the speed of the
object-equality fast-path, but that seems to be unavoidable and it's much
more important to have correct behaviour (especially in this case where
the wrong answer for nan==nan is silently returned).
2017-09-04 14:16:27 +10:00
Paul Sokolovsky 9950865c39 py/objfloat: Fix binary ops with incompatible objects.
These are now returned as "operation not supported" instead of raising
TypeError. In particular, this fixes equality for float vs incompatible
types, which now properly results in False instead of exception. This
also paves the road to support reverse operation (e.g. __radd__) with
float objects.

This is achieved by introducing mp_obj_get_float_maybe(), similar to
existing mp_obj_get_int_maybe().
2017-09-02 23:05:24 +03:00
Damien George 04552ff71b py: Implement raising a big-int to a negative power.
Before this patch raising a big-int to a negative power would just return
0.  Now it returns a floating-point number with the correct value.
2017-07-25 11:49:22 +10:00
Damien George 6b8b56f859 py/modmath: Check for zero division in log with 2 args. 2017-07-04 02:15:11 +10:00
Damien George 6ed4581f54 py/formatfloat: Fix number of digits and exponent sign when rounding.
This patch fixes 2 things when printing a floating-point number that
requires rounding up of the mantissa:
- retain the correct precision; eg 0.99 becomes 1.0, not 1.00
- if the exponent goes from -1 to 0 then render it as +0, not -0
2017-06-13 13:36:56 +10:00
Paul Sokolovsky e094200750 tests/float/builtin_float_minmax: PEP8 fixes. 2017-06-08 17:23:22 +03:00
Damien George fde54350a8 tests/float: Convert "sys.exit()" to "raise SystemExit".
The latter is shorter and simpler because it doesn't require importing the
sys module.
2017-06-08 14:00:57 +10:00
Damien George 677fb31015 tests/float: Add tests for hashing float and complex numbers. 2017-04-04 12:14:34 +10:00
Paul Sokolovsky 28876d3902 tests/float/byte*_construct: Skip on missing array module. 2017-04-03 00:17:43 +03:00
Damien George bfb48c1620 tests/float: Add tests for round() of inf, nan and large number. 2017-03-24 11:00:45 +11:00
Damien George bacb52aa2d tests/float: Add tests for math funcs that return ints.
One should test bigint, inf and nan to make sure all cases are covered.
2017-03-23 23:54:10 +11:00
Paul Sokolovsky a0cbc108ba tests/float: Make various tests skippable. 2017-03-09 00:11:05 +01:00
Paul Sokolovsky 88ffe0d5cc tests/string_format_modulo2: Split off intbig test. 2017-03-07 00:13:36 +01:00
Paul Sokolovsky 4b03941f5e tests/float2int*: Suffix with _intbig, don't run on any other int type.
I.e. they don't run successfully with MICROPY_LONGINT_IMPL_NONE
and MICROPY_LONGINT_IMPL_LONGLONG (the problem is that they generate
different output than CPython, TODO to fix that).
2017-03-06 16:30:12 +01:00
Paul Sokolovsky 121fb88988 float/float2int*: Make actually be parsable for MICROPY_LONGINT_IMPL_NONE.
The use of large literal numbers is a big no-no when it comes to writing
programs which work with different int representations. Also, some checks
are pretty adhoc (e.g using struct module to check for 64-bitness). This
change bases entire detection on sys.maxsize and integer operarions, and
thus more correct, even if longer.

Note that this change doesn't mean that any of these tests can pass with
anything but MPZ - even despite checking for various int representations,
the tests aren't written to be portable among them.
2017-03-06 16:23:09 +01:00
Paul Sokolovsky 325c4473a5 tests/float/complex1: Split out intbig test. 2017-03-06 15:46:01 +01:00
Damien George 8a39e18f5f tests/float: Add tests for zero to a negative power. 2017-02-03 00:04:13 +11:00
Damien George 213a718953 tests/float: Add test for assigning to attribute of complex number. 2017-01-19 23:37:21 +11:00
Rami Ali 64dc925c4a tests/float: Improve formatfloat.c test coverage using Python. 2017-01-05 12:31:05 +11:00
Rami Ali 531c206e8b tests: Add tests to improve coverage of runtime.c. 2016-12-21 15:44:41 +11:00
Damien George 49bf7617d0 tests/float/builtin_float_round: Test round() with second arg. 2016-12-20 14:01:10 +11:00
Damien George 290daa15d9 tests/float: Add test for parsing a float from an empty string. 2016-09-27 15:44:56 +10:00
Damien George e1e7657277 py/formatfloat: Fix further cases of buffer overflow in formatting.
Includes extensive test cases to catch hopefully all cases where
buffer might overflow.
2016-03-29 22:07:15 +01:00
Damien George cea6cf8a5e py/formatfloat: Fix buffer overflow when formatting tiny numbers. 2016-03-15 12:21:56 +00:00
Paul Sokolovsky d973c1bc12 test/string_format_fp30: Variant of string_format for 30-bit stuffed float. 2016-03-06 06:10:40 +02:00
Paul Sokolovsky 50e0a7b9d4 test/float2int_fp30: Variant of float2int for 30-bit stuffed float. 2016-03-06 06:08:38 +02:00
Paul Sokolovsky d4315a6caf float/string_format: Split large test in 2. 2016-02-13 17:18:55 +02:00
Paul Sokolovsky dc587a3623 test/float2int: Make test output clearer. 2016-02-13 17:03:02 +02:00
Damien George 3b936a5f4c tests: Fix math_fun_special test so it passes with single prec float. 2016-01-08 17:58:02 +00:00
Damien George b1fa907d56 tests: Allow float tests to run when MATH_SPECIAL_FUNCTIONS is disabled. 2016-01-08 14:27:21 +00:00
Michael Buesch 17298af61e py/modmath: Add domain error checking to sqrt, log, log2, log10.
These functions will raise 'ValueError: math domain error' on invalid
input.
2015-12-12 15:12:54 +00:00
Paul Sokolovsky b64e0575fd tests/float/string_format: Add testcase for incorrect rounding for %f. 2015-11-22 18:08:49 +02:00
Damien George e5635f4ab3 py: Catch all cases of integer (big and small) division by zero. 2015-10-01 22:48:48 +01:00
Damien George a81539db25 tests: Add further tests for mpz code. 2015-10-01 18:49:37 +01:00
Dave Hylands 9d6128acdc stmhal: fix single precision float printing error
Fixes #1435.
2015-09-11 23:09:50 +03:00