Struct CompareOptions 

#[non_exhaustive]
pub struct CompareOptions {
    pub group_by: GroupByOrDefault,
    pub cgroup_flatten: Vec<String>,
    pub no_thread_normalize: bool,
    pub no_cg_normalize: bool,
    pub sort_by: Vec<SortKey>,
}

Options controlling compare.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

group_by: GroupByOrDefault

cgroup_flatten: Vec<String>

Glob patterns that collapse dynamic cgroup path segments to a canonical form before grouping. Tried in listed order; the first pattern that matches a thread’s cgroup path replaces the whole path with the pattern string, so paths differing only in wildcard-matched segments collapse onto one key. A path matching no pattern is returned verbatim. See flatten_cgroup_path for the rewrite rule and examples.

Independent of Self::no_cg_normalize — explicit glob patterns apply first; auto-normalization (token-based) runs after, gated by no_cg_normalize.

no_thread_normalize: bool

When true, disable token-based pattern normalization across every name-family axis: GroupBy::Comm, GroupBy::Pcomm, AND the smaps_rollup keying in collect_smaps_rollup (which keys by pattern_key(&t.pcomm) under default normalization, but reverts to literal pcomm[tgid] when this flag is set so each PID stays attributable).

Under this flag: threads / processes group by their literal name; smaps rows preserve their per-PID identity. The pure-digit/hex/alpha+digits placeholders never fire on any of those axes. Mirror of Self::no_cg_normalize for the thread / process axes. Has no effect under GroupBy::CommExact (already literal) or GroupBy::Cgroup.

no_cg_normalize: bool

When true, disable token-based pattern normalization for cgroup-path grouping (GroupBy::Cgroup). Cgroup paths group by their literal post-flatten path (no Layer 1, 2, or 3 substitutions). Explicit cgroup_flatten glob patterns still apply. Has no effect under other groupings.

sort_by: Vec<SortKey>

Multi-key sort spec for the diff rows. When non-empty, overrides the default delta_pct desc sort. Each SortKey names one metric from CTPROF_METRICS or CTPROF_DERIVED_METRICS and a direction; groups rank by the tuple (metric_1_delta, metric_2_delta, …) under lexicographic order with per-key direction. Within a group, rows appear in registry order. The sort composes with Self::group_by: groups are formed under the chosen axis (pcomm / cgroup / comm / comm-exact) and then ranked by their aggregated metric values, so the same sort_by spec works under every grouping. See parse_sort_by for the CLI string parser.

Trait Implementations

impl Clone for CompareOptions

fn clone(&self) -> CompareOptions

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CompareOptions

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for CompareOptions

fn default() -> CompareOptions

Returns the “default value” for a type.