Generate path encoding set using 'const fn'.

Co-authored-by: Jakub Wieczorek <jakub.adam.wieczorek@gmail.com>

core/http/src/parse/uri/mod.rs

@@ -1,6 +1,6 @@
 mod parser;
 mod error;
-mod tables;
+pub(crate) mod tables;
 
 #[cfg(test)] mod tests;
 
@@ -8,7 +8,6 @@ use crate::uri::{Uri, Origin, Absolute, Authority};
 
 use self::parser::{uri, origin, authority_only, absolute_only, rocket_route_origin};
 
-pub use self::tables::{is_pchar, PATH_SET};
 pub use self::error::Error;
 
 type RawInput<'a> = pear::input::Pear<pear::input::Cursor<&'a [u8]>>;

core/http/src/parse/uri/tables.rs

@@ -1,40 +1,4 @@
-use percent_encoding::AsciiSet;
+pub(crate) const PATH_CHARS: [u8; 256] = [
-
-// Generates an AsciiSet from a full character table:
-//
-// table_to_ascci_set!( ExistingAsciiSet, 0, [ 0, 1, b'x', ... ] );
-//
-// A 0 or 1 token in the ith position indicates that the ASCII character i
-// should be included in the set. Any other literal in the ith position
-// indicates that the ASCII character with value i should not be included in the
-// set.
-//
-// The table's last index must be 127 or earlier. All values in the original
-// set, up to the end of the passed-in table, are overwritten.
-macro_rules! table_to_ascii_set {
-    ($base:expr, $i:expr, [ 0,           $($rest:tt,)* ]) => { table_to_ascii_set!($base.add($i),    $i+1, [ $($rest,)* ]); };
-    ($base:expr, $i:expr, [ 1,           $($rest:tt,)* ]) => { table_to_ascii_set!($base.add($i),    $i+1, [ $($rest,)* ]); };
-    ($base:expr, $i:expr, [ $ch:literal, $($rest:tt,)* ]) => { table_to_ascii_set!($base.remove($i), $i+1, [ $($rest,)* ]); };
-    ($base:expr, $i:expr, [ ]) => { $base };
-}
-
-// Generates an AsciiSet accompanying a character table. This is
-// used to keep `PATH_CHARS` in sync with `PATH_SET`.
-//
-// The first block is limited to 128 entries, since it is passed
-// to table_to_ascii_set!
-macro_rules! table_and_asciiset {
-    (
-        const $name:ident: [u8; $size:expr] = [ $($block1:tt)* ] [ $($block2:tt)* ];
-        pub const $setname:ident: AsciiSet;
-    ) => {
-        const $name: [u8; $size] = [ $($block1)* $($block2)* ] ;
-        pub const $setname: AsciiSet = table_to_ascii_set!(percent_encoding::CONTROLS, 0, [ $($block1)* ]);
-    };
-}
-
-table_and_asciiset! {
-const PATH_CHARS: [u8; 256] = [
     //  0      1      2      3      4      5      6      7      8      9
         0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //   x
         0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  1x
@@ -42,14 +6,13 @@ const PATH_CHARS: [u8; 256] = [
         0,     0,     0,  b'!',     0,     0,  b'$',  b'%',  b'&', b'\'', //  3x
      b'(',  b')',  b'*',  b'+',  b',',  b'-',  b'.',  b'/',  b'0',  b'1', //  4x
      b'2',  b'3',  b'4',  b'5',  b'6',  b'7',  b'8',  b'9',  b':',  b';', //  5x
-     // <             > (1 used to indicate these are valid in route URIs only)
+        0,  b'=',     0,     0,  b'@',  b'A',  b'B',  b'C',  b'D',  b'E', //  6x
-        1,  b'=',     1,     0,  b'@',  b'A',  b'B',  b'C',  b'D',  b'E', //  6x
      b'F',  b'G',  b'H',  b'I',  b'J',  b'K',  b'L',  b'M',  b'N',  b'O', //  7x
      b'P',  b'Q',  b'R',  b'S',  b'T',  b'U',  b'V',  b'W',  b'X',  b'Y', //  8x
      b'Z',     0,     0,     0,     0,  b'_',     0,  b'a',  b'b',  b'c', //  9x
      b'd',  b'e',  b'f',  b'g',  b'h',  b'i',  b'j',  b'k',  b'l',  b'm', // 10x
      b'n',  b'o',  b'p',  b'q',  b'r',  b's',  b't',  b'u',  b'v',  b'w', // 11x
-     b'x',  b'y',  b'z',     0,     0,     0,  b'~',     0, ] [ 0,     0, // 12x
+     b'x',  b'y',  b'z',     0,     0,     0,  b'~',     0,     0,     0, // 12x
         0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 13x
         0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 14x
         0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 15x
@@ -64,17 +27,50 @@ const PATH_CHARS: [u8; 256] = [
         0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 24x
         0,     0,     0,     0,     0,     0,                             // 25x
 ];
-pub const PATH_SET: AsciiSet;
-}
 
 #[inline(always)]
 pub fn is_pchar(&c: &u8) -> bool {
-    PATH_CHARS[c as usize] == c
+    PATH_CHARS[c as usize] != 0
+}
+
+pub(crate) const ROUTE_CHARS: [u8; 256] = [
+    //  0      1      2      3      4      5      6      7      8      9
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //   x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  1x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  2x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  3x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  4x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  5x
+     b'<',     0,  b'>',     0,     0,     0,     0,     0,     0,     0, //  6x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  7x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  8x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, //  9x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 10x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 11x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 12x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 13x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 14x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 15x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 16x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 17x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 18x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 19x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 20x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 21x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 22x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 23x
+        0,     0,     0,     0,     0,     0,     0,     0,     0,     0, // 24x
+        0,     0,     0,     0,     0,     0,                             // 25x
+];
+
+#[inline(always)]
+pub fn is_rchar(&c: &u8) -> bool {
+    ROUTE_CHARS[c as usize] != 0
 }
 
 #[inline(always)]
-pub fn is_pchar_or_rchar(&c: &u8) -> bool {
+pub fn is_pchar_or_rchar(c: &u8) -> bool {
-    PATH_CHARS[c as usize] != 0
+    is_pchar(c) || is_rchar(c)
 }
 
 const REG_CHARS: [u8; 256] = [

core/http/src/uri/encoding.rs

@@ -4,7 +4,7 @@ use std::borrow::Cow;
 use percent_encoding::{AsciiSet, utf8_percent_encode};
 
 use crate::uri::{UriPart, Path, Query};
-use crate::parse::uri::PATH_SET;
+use crate::parse::uri::tables::PATH_CHARS;
 
 #[derive(Clone, Copy)]
 #[allow(non_camel_case_types)]
@@ -13,6 +13,24 @@ pub trait EncodeSet {
     const SET: AsciiSet;
 }
 
+const fn set_from_table(table: &'static [u8; 256]) -> AsciiSet {
+    const ASCII_RANGE_LEN: u8 = 0x80;
+
+    let mut set = percent_encoding::CONTROLS.add(0);
+    let mut i: u8 = 0;
+    while i < ASCII_RANGE_LEN {
+        if table[i as usize] == 0 {
+            set = set.add(i);
+        }
+
+        i += 1;
+    }
+
+    set
+}
+
+const PATH_SET: AsciiSet = set_from_table(&PATH_CHARS);
+
 impl<P: UriPart> Default for UNSAFE_ENCODE_SET<P> {
     #[inline(always)]
     fn default() -> Self { UNSAFE_ENCODE_SET(PhantomData) }

core/http/src/uri/origin.rs

@@ -343,7 +343,9 @@ impl<'a> Origin<'a> {
     #[inline]
     pub fn map_path<F: FnOnce(&str) -> String>(&self, f: F) -> Option<Self> {
         let path = f(self.path());
-        if !path.starts_with('/') || !path.bytes().all(|b| crate::parse::uri::is_pchar(&b)) {
+        if !path.starts_with('/')
+            || !path.bytes().all(|b| crate::parse::uri::tables::is_pchar(&b))
+        {
             return None;
         }