public/browser
1
0
Fork 0
mirror of https://github.com/lightpanda-io/browser.git synced 2025-10-29 15:13:28 +00:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

double quotes must be first and last argument of slice if provided

Browse Source
This commit is contained in:
nikneym
2025-09-16 10:40:38 +03:00
parent 2e68407fbe
commit c05470515f
1 changed files with 9 additions and 75 deletions
Download Patch File Download Diff File Expand all files Collapse all files

84
src/browser/mime.zig
View File

@@ -69,83 +69,17 @@ pub const Mime = struct {
// https://datatracker.ietf.org/doc/rfc2978/
if (value.len > 40) return error.CharsetTooBig;
// If the first char is not a quote, value can be used directly.
// Whitespace is not allowed.
if (value[0] != '"') {
return value;
// If the first char is a quote, look for a pair.
if (value[0] == '"') {
if (value.len < 3 or value[value.len - 1] != '"') {
return error.Invalid;
}
return value[1 .. value.len - 1];
}
// Search for second quote begins.
// Skip the first character.
var offset: usize = 1;
// Charset values are not so large; 128-bit registers should be
// more than enough.
const vec_size = 16;
const Vec = @Vector(vec_size, u8);
const UInt = std.meta.Int(.unsigned, vec_size);
const block_size = @sizeOf(u64);
const charset = blk: {
// Vector search.
while (value.len - offset >= vec_size) : (offset += vec_size) {
// Fill a vector with quotes.
const quotes: Vec = @splat('"');
const chunk: Vec = value[offset..][0..vec_size].*;
// Check if chunk has double quote byte.
const match = @intFromBool(chunk == quotes);
// Create an integer out of match and count how much to skip.
const skip_by = @ctz(@as(UInt, @bitCast(match)));
// Found a match.
if (skip_by != vec_size) {
break :blk value[1 .. offset + skip_by];
}
}
// SWAR search.
while (value.len - offset >= block_size) : (offset += block_size) {
// Magic number for integer filled with double quote.
// [8]u8{ '"', '"', '"', '"', '"', '"', '"', '"' }.
const quotes: u64 = 0x2222222222222222;
// Load the next chunk as unsigned 64-bit integer.
const chunk: u64 = @bitCast(value[offset..][0..block_size].*);
// XOR with the pattern - bytes equal to quote become 0.
const xor_result = chunk ^ quotes;
const magic: u64 = 0x8080808080808080; // High bit mask for each byte.
const sub_result = xor_result -% 0x0101010101010101; // Subtract 1 from each byte.
const and_result = sub_result & (~xor_result); // AND with inverted original.
const zero_mask = and_result & magic; // Extract high bits (indicates zero bytes).
// Found a match.
if (zero_mask != 0) {
// * Count trailing zeroes.
// * Dividing by byte size (>> 3) converts the bit position to byte index.
const skip_by = @ctz(zero_mask) >> 3;
break :blk value[1 .. offset + skip_by];
}
}
// Fallback to scalar search.
for (value[offset..], 0..) |c, i| {
if (c == '"') {
break :blk value[1 .. offset + i];
}
}
// No quote pairs, something is wrong.
return error.Invalid;
};
// Make sure we don't end up w/ empty buffer.
if (charset.len == 0) {
return error.Invalid;
}
return charset;
// No quotes.
return value;
}
pub fn parse(input: []u8) !Mime {