This website requires JavaScript.
About
Status
₊˚⊹Discord
Explore
Register
Sign in
Reference in a new issue
Blame
Copy permalink
Reference in a new issue
Copy permalink
DKB
/
zig
Star
0
Fork
You've already forked zig
0
mirror of
https://codeberg.org/ziglang/zig
synced
2026-11-18 07:05:05 +00:00
Code
Issues
Releases
Packages
Activity
e8bcdca044
zig
/
lib
/
std
/
fmt
/
parse_float.zig
171 lines
9.5 KiB
Raw
Normal View
History
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
Add parseFloat to std.fmt This is not intended to be the long-term implementation as it doesn't provide various properties that we eventually will want (e.g. round-tripping, denormal support). It also uses f64 internally so the wider f128 will be inaccurate.
2019-02-14 00:06:32 +13:00
std.fmt.parse_float: disable failing aarch64 test from LLVM 14 See #12027
2022-07-06 16:23:25 -07:00
add ast-check flag to zig fmt, fix found bugs
2021-06-13 22:27:27 +03:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
Add parseFloat to std.fmt This is not intended to be the long-term implementation as it doesn't provide various properties that we eventually will want (e.g. round-tripping, denormal support). It also uses f64 internally so the wider f128 will be inaccurate.
2019-02-14 00:06:32 +13:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
Add parseFloat to std.fmt This is not intended to be the long-term implementation as it doesn't provide various properties that we eventually will want (e.g. round-tripping, denormal support). It also uses f64 internally so the wider f128 will be inaccurate.
2019-02-14 00:06:32 +13:00
different array literal syntax when inferring the size old syntax: []i32{1, 2, 3} new syntax: [_]i32{1, 2, 3} closes #1797
2019-06-09 19:24:24 -04:00
std: update usage of std.testing
2021-05-04 20:47:26 +03:00
Add parseFloat to std.fmt This is not intended to be the long-term implementation as it doesn't provide various properties that we eventually will want (e.g. round-tripping, denormal support). It also uses f64 internally so the wider f128 will be inaccurate.
2019-02-14 00:06:32 +13:00
std: update usage of std.testing
2021-05-04 20:47:26 +03:00
Add parseFloat to std.fmt This is not intended to be the long-term implementation as it doesn't provide various properties that we eventually will want (e.g. round-tripping, denormal support). It also uses f64 internally so the wider f128 will be inaccurate.
2019-02-14 00:06:32 +13:00
std: update usage of std.testing
2021-05-04 20:47:26 +03:00
Fix addition direction, remove superfluous loop counter, add tests
2019-09-01 13:27:26 +03:00
std: update usage of std.testing
2021-05-04 20:47:26 +03:00
Add parseFloat to std.fmt This is not intended to be the long-term implementation as it doesn't provide various properties that we eventually will want (e.g. round-tripping, denormal support). It also uses f64 internally so the wider f128 will be inaccurate.
2019-02-14 00:06:32 +13:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
Make parseFloat stricter in what it accepts as input
2019-02-15 17:32:13 +13:00
std: update usage of std.testing
2021-05-04 20:47:26 +03:00
add ast-check flag to zig fmt, fix found bugs
2021-06-13 22:27:27 +03:00
fix overflow in parseFloat
2020-03-29 06:54:23 +02:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
Fix addition direction, remove superfluous loop counter, add tests
2019-09-01 13:27:26 +03:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
add more corner case tests to float_parse (#11727) also drop some unused constants Co-authored-by: Andrew Kelley <andrew@ziglang.org>
2022-05-27 04:05:53 +02:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
Add parseFloat to std.fmt This is not intended to be the long-term implementation as it doesn't provide various properties that we eventually will want (e.g. round-tripping, denormal support). It also uses f64 internally so the wider f128 will be inaccurate.
2019-02-14 00:06:32 +13:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
std.fmt.parse_float: disable failing aarch64 test from LLVM 14 See #12027
2022-07-06 16:23:25 -07:00
Revert "aarch64: reenable tests that are no longer regressed" This reverts commit 3370d58956ecc744a004dff47b0437473f0ef7da. This commit was done with an LLVM build that did not have assertions enabled. There are LLVM assertions being triggered due to this commit. Reopens #10627 Reopens #12013 Reopens #12027
2022-12-27 20:47:51 -07:00
std.fmt.parse_float: disable failing aarch64 test from LLVM 14 See #12027
2022-07-06 16:23:25 -07:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
Fix #14901: parseFloat parsing `0x` successfully
2023-03-15 12:53:53 -04:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
enable more float-parsing tests Since removing the stage1 backend we no longer have a disagreement here.
2023-03-25 22:49:40 +13:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00
enable more float-parsing tests Since removing the stage1 backend we no longer have a disagreement here.
2023-03-25 22:49:40 +13:00
add new float-parser based on eisel-lemire algorithm The previous float-parsing method was lacking in a lot of areas. This commit introduces a state-of-the art implementation that is both accurate and fast to std. Code is derived from working repo https://github.com/tiehuis/zig-parsefloat. This includes more test-cases and performance numbers that are present in this commit. * Accuracy The primary testing regime has been using test-data found at https://github.com/tiehuis/parse-number-fxx-test-data. This is a fork of upstream with support for f128 test-cases added. This data has been verified against other independent implementations and represents accurate round-to-even IEEE-754 floating point semantics. * Performance Compared to the existing parseFloat implementation there is ~5-10x performance improvement using the above corpus. (f128 parsing excluded in below measurements). ** Old $ time ./test_all_fxx_data 3520298/5296694 succeeded (1776396 fail) ________________________________________________________ Executed in 28.68 secs fish external usr time 28.48 secs 0.00 micros 28.48 secs sys time 0.08 secs 694.00 micros 0.08 secs ** This Implementation $ time ./test_all_fxx_data 5296693/5296694 succeeded (1 fail) ________________________________________________________ Executed in 4.54 secs fish external usr time 4.37 secs 515.00 micros 4.37 secs sys time 0.10 secs 171.00 micros 0.10 secs Further performance numbers can be seen using the https://github.com/tiehuis/simple_fastfloat_benchmark/ repository, which compares against some other well-known string-to-float conversion functions. A breakdown can be found here: https://github.com/tiehuis/zig-parsefloat/blob/0d9f020f1a37ca88bf889703b397c1c41779f090/PERFORMANCE.md#commit-b15406a0d2e18b50a4b62fceb5a6a3bb60ca5706 In summary, we are within 20% of the C++ reference implementation and have about ~600-700MB/s throughput on a Intel I5-6500 3.5Ghz. * F128 Support Finally, f128 is now completely supported with full accuracy. This does use a slower path which is possible to improve in future. * Behavioural Changes There are a few behavioural changes to note. - `parseHexFloat` is now redundant and these are now supported directly in `parseFloat`. - We implement round-to-even in all parsing routines. This is as specified by IEEE-754. Previous code used different rounding mechanisms (standard was round-to-zero, hex-parsing looked to use round-up) so there may be subtle differences. Closes #2207. Fixes #11169.
2022-05-02 22:45:06 +12:00