Skip to content

Building 70 Projects ranging from beginner to advanced so anyone can β€” learn from, build upon, use as a reference, or even copy directly. Owned by my cybersecurity company πŸ‘‡

chore(lisdex): drop the comments B28 added, keep the figures it corrected

CarterPerez-devab1c01c4e690
diff --git a/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/complexity.rs b/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/complexity.rs
index 7a0244c5..d1fff013 100644
--- a/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/complexity.rs
+++ b/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/complexity.rs
@@ -200,11 +200,6 @@ fn pointer_operands(entity: &Entity) -> usize {
.count()
}
-// Leopard-V's pointer-arithmetic term, defined from types alone because clang-2.1.0 exposes
-// EntityKind::BinaryOperator but no accessor for which operator it is. Each arm below is
-// separated by the RESULT type and the operand count rather than by a spelling:
-// `p == q` and `p - q` are integer-typed and never reach here, and `p = q` carries TWO
-// pointer operands where `p + n` carries one.
fn is_pointer_arithmetic(entity: &Entity) -> bool {
match entity.get_kind() {
EntityKind::ArraySubscriptExpr => entity
@@ -219,8 +214,6 @@ fn is_pointer_arithmetic(entity: &Entity) -> bool {
}
}
-// A subscript on a true array decays to a pointer at the cast, so the cast's own type cannot
-// separate `p[i]` from `arr[i]`. The declared type underneath it can.
fn subscript_base_is_pointer(base: &Entity) -> bool {
let mut node = *base;
loop {
diff --git a/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/rank.rs b/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/rank.rs
index 3f4036d6..56fc4f94 100644
--- a/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/rank.rs
+++ b/PROJECTS/advanced/zero-day-vulnerability-scanner/crates/lisdex-surface/src/rank.rs
@@ -124,11 +124,6 @@ fn exclusion_for(role: Role) -> Option<ExclusionReason> {
// top would saturate at 1.0 and stop discriminating, on the one tree where the numbers were
// hardest won. This leaves 1.75x headroom over anything measured.
//
-// B28 re-measured every figure above when Leopard-V's pointer-arithmetic term entered
-// `danger`. The previous readings were libarchive 14646, libpng 5088, libtiff 4543,
-// openjpeg 4402, brotli 832, lz4 439, so the term moved each tree between +18.6% and +123.9%.
-// openjpeg more than doubled and overtook libpng. The ranking did not move at all.
-//
// A saturating ratio was tried first because it can never clamp at all, and was rejected on
// measurement: it cannot discriminate across a range this wide. At the magnitudes a small
// synthetic tree produces it scored a parser-driving reader 0.012 against a trivial
@@ -455,7 +450,7 @@ mod tests {
param(
"readproc",
Shape::Pointer {
- pointee: Pointee::Function,
+ pointee: Pointee::Function(None),
is_const: false,
},
),
@@ -504,7 +499,7 @@ mod tests {
param(
"read_data_fn",
Shape::Pointer {
- pointee: Pointee::Function,
+ pointee: Pointee::Function(None),
is_const: false,
},
),