0.8.6 more precise (and faster) float parsing

Author: maciejhirsz

Cargo.toml

@@ -9,4 +9,3 @@ license = "MIT"
 
 [dependencies]
 itoa = "0.1"
-dtoa = { git = "https://github.com/maciejhirsz/dtoa", branch = "master" }

src/codegen.rs

@@ -3,7 +3,7 @@ use std::num::FpCategory;
 use JsonValue;
 
 extern crate itoa;
-extern crate dtoa;
+use dtoa;
 
 const QU: u8 = b'"';
 const BS: u8 = b'\\';

src/dtoa/diyfp.rs

@@ -0,0 +1,251 @@
+// Copyright 2016 Dtoa Developers
+//
+// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+// option. This file may not be copied, modified, or distributed
+// except according to those terms.
+
+use std::{mem, ops};
+
+const DIY_SIGNIFICAND_SIZE: isize = 64;
+const DP_SIGNIFICAND_SIZE: isize = 52;
+const DP_EXPONENT_BIAS: isize = 0x3FF + DP_SIGNIFICAND_SIZE;
+const DP_MIN_EXPONENT: isize = -DP_EXPONENT_BIAS;
+const DP_EXPONENT_MASK: u64 = 0x7FF0000000000000;
+const DP_SIGNIFICAND_MASK: u64 = 0x000FFFFFFFFFFFFF;
+const DP_HIDDEN_BIT: u64 = 0x0010000000000000;
+
+#[derive(Copy, Clone)]
+pub struct DiyFp {
+    pub f: u64,
+    pub e: isize,
+}
+
+impl DiyFp {
+    pub fn new(f: u64, e: isize) -> Self {
+        DiyFp { f: f, e: e }
+    }
+
+    /*
+    explicit DiyFp(double d) {
+        union {
+            double d;
+            uint64_t u64;
+        } u = { d };
+
+        int biased_e = static_cast<int>((u.u64 & kDpExponentMask) >> kDpSignificandSize);
+        uint64_t significand = (u.u64 & kDpSignificandMask);
+        if (biased_e != 0) {
+            f = significand + kDpHiddenBit;
+            e = biased_e - kDpExponentBias;
+        }
+        else {
+            f = significand;
+            e = kDpMinExponent + 1;
+        }
+    }
+    */
+    pub unsafe fn from_f64(d: f64) -> Self {
+        let u: u64 = mem::transmute(d);
+
+        let biased_e = ((u & DP_EXPONENT_MASK) >> DP_SIGNIFICAND_SIZE) as isize;
+        let significand = u & DP_SIGNIFICAND_MASK;
+        if biased_e != 0 {
+            DiyFp {
+                f: significand + DP_HIDDEN_BIT,
+                e: biased_e - DP_EXPONENT_BIAS,
+            }
+        } else {
+            DiyFp {
+                f: significand,
+                e: DP_MIN_EXPONENT + 1,
+            }
+        }
+    }
+
+    /*
+    DiyFp Normalize() const {
+        DiyFp res = *this;
+        while (!(res.f & (static_cast<uint64_t>(1) << 63))) {
+            res.f <<= 1;
+            res.e--;
+        }
+        return res;
+    }
+    */
+    pub fn normalize(self) -> DiyFp {
+        let mut res = self;
+        while (res.f & (1u64 << 63)) == 0 {
+            res.f <<= 1;
+            res.e -= 1;
+        }
+        res
+    }
+
+    /*
+    DiyFp NormalizeBoundary() const {
+        DiyFp res = *this;
+        while (!(res.f & (kDpHiddenBit << 1))) {
+            res.f <<= 1;
+            res.e--;
+        }
+        res.f <<= (kDiySignificandSize - kDpSignificandSize - 2);
+        res.e = res.e - (kDiySignificandSize - kDpSignificandSize - 2);
+        return res;
+    }
+    */
+    pub fn normalize_boundary(self) -> DiyFp {
+        let mut res = self;
+        while (res.f & DP_HIDDEN_BIT << 1) == 0 {
+            res.f <<= 1;
+            res.e -= 1;
+        }
+        res.f <<= DIY_SIGNIFICAND_SIZE - DP_SIGNIFICAND_SIZE - 2;
+        res.e -= DIY_SIGNIFICAND_SIZE - DP_SIGNIFICAND_SIZE - 2;
+        res
+    }
+
+    /*
+    void NormalizedBoundaries(DiyFp* minus, DiyFp* plus) const {
+        DiyFp pl = DiyFp((f << 1) + 1, e - 1).NormalizeBoundary();
+        DiyFp mi = (f == kDpHiddenBit) ? DiyFp((f << 2) - 1, e - 2) : DiyFp((f << 1) - 1, e - 1);
+        mi.f <<= mi.e - pl.e;
+        mi.e = pl.e;
+        *plus = pl;
+        *minus = mi;
+    }
+    */
+    pub fn normalized_boundaries(self) -> (DiyFp, DiyFp) {
+        let pl = DiyFp::new((self.f << 1) + 1, self.e - 1).normalize_boundary();
+        let mut mi = if self.f == DP_HIDDEN_BIT {
+            DiyFp::new((self.f << 2) - 1, self.e - 2)
+        } else {
+            DiyFp::new((self.f << 1) - 1, self.e - 1)
+        };
+        mi.f <<= mi.e - pl.e;
+        mi.e = pl.e;
+        (mi, pl)
+    }
+}
+
+impl ops::Sub for DiyFp {
+    type Output = DiyFp;
+    fn sub(self, rhs: DiyFp) -> DiyFp {
+        DiyFp {
+            f: self.f - rhs.f,
+            e: self.e,
+        }
+    }
+}
+
+impl ops::Mul for DiyFp {
+    type Output = DiyFp;
+    fn mul(self, rhs: DiyFp) -> DiyFp {
+        let m32 = 0xFFFFFFFFu64;
+        let a = self.f >> 32;
+        let b = self.f & m32;
+        let c = rhs.f >> 32;
+        let d = rhs.f & m32;
+        let ac = a * c;
+        let bc = b * c;
+        let ad = a * d;
+        let bd = b * d;
+        let mut tmp = (bd >> 32) + (ad & m32) + (bc & m32);
+        tmp += 1u64 << 31; // mult_round
+        DiyFp {
+            f: ac + (ad >> 32) + (bc >> 32) + (tmp >> 32),
+            e: self.e + rhs.e + 64,
+        }
+    }
+}
+
+fn get_cached_power_by_index(index: usize) -> DiyFp {
+    // 10^-348, 10^-340, ..., 10^340
+    static CACHED_POWERS_F: [u64; 87] = [
+        0xfa8fd5a0081c0288, 0xbaaee17fa23ebf76,
+        0x8b16fb203055ac76, 0xcf42894a5dce35ea,
+        0x9a6bb0aa55653b2d, 0xe61acf033d1a45df,
+        0xab70fe17c79ac6ca, 0xff77b1fcbebcdc4f,
+        0xbe5691ef416bd60c, 0x8dd01fad907ffc3c,
+        0xd3515c2831559a83, 0x9d71ac8fada6c9b5,
+        0xea9c227723ee8bcb, 0xaecc49914078536d,
+        0x823c12795db6ce57, 0xc21094364dfb5637,
+        0x9096ea6f3848984f, 0xd77485cb25823ac7,
+        0xa086cfcd97bf97f4, 0xef340a98172aace5,
+        0xb23867fb2a35b28e, 0x84c8d4dfd2c63f3b,
+        0xc5dd44271ad3cdba, 0x936b9fcebb25c996,
+        0xdbac6c247d62a584, 0xa3ab66580d5fdaf6,
+        0xf3e2f893dec3f126, 0xb5b5ada8aaff80b8,
+        0x87625f056c7c4a8b, 0xc9bcff6034c13053,
+        0x964e858c91ba2655, 0xdff9772470297ebd,
+        0xa6dfbd9fb8e5b88f, 0xf8a95fcf88747d94,
+        0xb94470938fa89bcf, 0x8a08f0f8bf0f156b,
+        0xcdb02555653131b6, 0x993fe2c6d07b7fac,
+        0xe45c10c42a2b3b06, 0xaa242499697392d3,
+        0xfd87b5f28300ca0e, 0xbce5086492111aeb,
+        0x8cbccc096f5088cc, 0xd1b71758e219652c,
+        0x9c40000000000000, 0xe8d4a51000000000,
+        0xad78ebc5ac620000, 0x813f3978f8940984,
+        0xc097ce7bc90715b3, 0x8f7e32ce7bea5c70,
+        0xd5d238a4abe98068, 0x9f4f2726179a2245,
+        0xed63a231d4c4fb27, 0xb0de65388cc8ada8,
+        0x83c7088e1aab65db, 0xc45d1df942711d9a,
+        0x924d692ca61be758, 0xda01ee641a708dea,
+        0xa26da3999aef774a, 0xf209787bb47d6b85,
+        0xb454e4a179dd1877, 0x865b86925b9bc5c2,
+        0xc83553c5c8965d3d, 0x952ab45cfa97a0b3,
+        0xde469fbd99a05fe3, 0xa59bc234db398c25,
+        0xf6c69a72a3989f5c, 0xb7dcbf5354e9bece,
+        0x88fcf317f22241e2, 0xcc20ce9bd35c78a5,
+        0x98165af37b2153df, 0xe2a0b5dc971f303a,
+        0xa8d9d1535ce3b396, 0xfb9b7cd9a4a7443c,
+        0xbb764c4ca7a44410, 0x8bab8eefb6409c1a,
+        0xd01fef10a657842c, 0x9b10a4e5e9913129,
+        0xe7109bfba19c0c9d, 0xac2820d9623bf429,
+        0x80444b5e7aa7cf85, 0xbf21e44003acdd2d,
+        0x8e679c2f5e44ff8f, 0xd433179d9c8cb841,
+        0x9e19db92b4e31ba9, 0xeb96bf6ebadf77d9,
+        0xaf87023b9bf0ee6b,
+    ];
+    static CACHED_POWERS_E: [i16; 87] = [
+        -1220, -1193, -1166, -1140, -1113, -1087, -1060, -1034, -1007,  -980,
+        -954,  -927,  -901,  -874,  -847,  -821,  -794,  -768,  -741,  -715,
+        -688,  -661,  -635,  -608,  -582,  -555,  -529,  -502,  -475,  -449,
+        -422,  -396,  -369,  -343,  -316,  -289,  -263,  -236,  -210,  -183,
+        -157,  -130,  -103,   -77,   -50,   -24,     3,    30,    56,    83,
+        109,   136,   162,   189,   216,   242,   269,   295,   322,   348,
+        375,   402,   428,   455,   481,   508,   534,   561,   588,   614,
+        641,   667,   694,   720,   747,   774,   800,   827,   853,   880,
+        907,   933,   960,   986,  1013,  1039,  1066,
+    ];
+    DiyFp::new(CACHED_POWERS_F[index], CACHED_POWERS_E[index] as isize)
+}
+
+/*
+inline DiyFp GetCachedPower(int e, int* K) {
+    //int k = static_cast<int>(ceil((-61 - e) * 0.30102999566398114)) + 374;
+    double dk = (-61 - e) * 0.30102999566398114 + 347;  // dk must be positive, so can do ceiling in positive
+    int k = static_cast<int>(dk);
+    if (dk - k > 0.0)
+        k++;
+
+    unsigned index = static_cast<unsigned>((k >> 3) + 1);
+    *K = -(-348 + static_cast<int>(index << 3));    // decimal exponent no need lookup table
+
+    return GetCachedPowerByIndex(index);
+}
+*/
+#[inline]
+pub fn get_cached_power(e: isize) -> (DiyFp, isize) {
+    let dk = (-61 - e) as f64 * 0.30102999566398114f64 + 347f64; // dk must be positive, so can do ceiling in positive
+    let mut k = dk as isize;
+    if dk - k as f64 > 0.0 {
+        k += 1;
+    }
+
+    let index = ((k >> 3) + 1) as usize;
+    let k = -(-348 + (index << 3) as isize); // decimal exponent no need lookup table
+
+    (get_cached_power_by_index(index), k)
+}