drop regex and handroll parser

the rustflags change, leaner parser, and dropping the regex crate
yields a 4x binary size decrease, i can also see it runs around 2.5x
faster

1 files changed, 81 insertions, 0 deletions

diff --git a/src/dateparse.rs b/src/dateparse.rs
new file mode 100644
index 0000000..ac5dab1
--- /dev/null
+++ b/src/dateparse.rs
@@ -0,0 +1,81 @@
+use std::simd::{
+    cmp::{SimdPartialEq, SimdPartialOrd},
+    simd_swizzle, u8x16,
+};
+
+#[must_use]
+pub fn dateparse(s: &str) -> Option<(usize, usize, usize)> {
+    // inf and sup for indiv. values
+    const LO: u8x16 = u8x16::from_array([
+        b'0', b'0', b'/', b'0', b'0', b'/', b'0', b'0', //
+        0, 0, 0, 0, 0, 0, 0, 0, // load_or_default isnt const
+    ]);
+    const HI: u8x16 = u8x16::from_array([
+        b'1', b'9', b'/', b'3', b'9', b'/', b'9', b'9', //
+        0xFF, 0xFF, 0xFF, 0xFF, //
+        0xFF, 0xFF, 0xFF, 0xFF, //
+    ]);
+    const INTMSK: u8x16 = u8x16::from_array([b'0'; 16]);
+    const SLHMSK: u8x16 = u8x16::from_array([b'/'; 16]);
+    const NUMMSK: u8x16 = u8x16::from_array([
+        10, 1, 0, // [m_10, m_1] -> m_10 * 10 + m_1 (0 is slash)
+        10, 1, 0, // dd ..
+        10, 1, 0, // yy ..
+        0, 0, 0, 0, 0, 0, 0,
+    ]);
+    let b = s.as_bytes();
+    let v = u8x16::load_or_default(b);
+    let mut valid = true;
+    // index 16 for swizzles with INTMSK is b'0'
+    // index 15 is 0
+    // this will compile to a shuffle, but a relaxed_laneselect would be cool
+    let mut v = match (b.len(), v.simd_eq(SLHMSK).to_bitmask()) {
+        // m/d/yy
+        (6, 0b1010) => simd_swizzle!(
+            v,
+            INTMSK,
+            [16, 0, 1, 16, 2, 3, 4, 5, 15, 15, 15, 15, 15, 15, 15, 15]
+        ),
+        // m/dd/yy
+        (7, 0b10010) => simd_swizzle!(
+            v,
+            INTMSK,
+            [16, 0, 1, 2, 3, 4, 5, 6, 15, 15, 15, 15, 15, 15, 15, 15]
+        ),
+        // mm/d/yy
+        (7, 0b10100) => simd_swizzle!(
+            v,
+            INTMSK,
+            [0, 1, 2, 16, 3, 4, 5, 6, 15, 15, 15, 15, 15, 15, 15, 15]
+        ),
+        // mm/dd/yy
+        (8, 0b100100) => v,
+        _ => {
+            valid = false;
+            v
+        }
+    };
+
+    // est-ce numerique?
+    valid &= (v.simd_ge(LO) & v.simd_le(HI)).all();
+
+    // parse
+    //   '0' '3' '/' '3' '1' '/' '2' '6'
+    // - '0' '0' '0' '0' '0' '0' '0' '0'
+    // ---------------------------------
+    //    0   3   .   3   1   .   2   6
+    // * 10   1   0  10   1   0  10   1
+    // ---------------------------------
+    //    0   3   0  30   1   0  20   6
+    // +  3   0  30   1   0  20   6<<<<
+    // ---------------------------------
+    //    3   .   .  31   .   .  26
+    v -= INTMSK;
+    v *= NUMMSK;
+    v += v.rotate_elements_left::<1>();
+    let n = v.to_array();
+    let (m, d, y) = (n[0], n[3], n[6]);
+
+    (valid & (1..=12).contains(&m) & (1..=31).contains(&d))
+        .then(|| (m.into(), d.into(), y.into()))
+}