Struct AnalyzeBuilder 

pub struct AnalyzeBuilder { /* private fields */ }

Implementations

impl AnalyzeBuilder

pub fn extensions( self, exts: impl IntoIterator<Item = impl Into<String>>, ) -> Self

Only files with one of these extensions (case-insensitive, without the leading dot) are matched. Unset matches any extension, including files with none.

pub fn exclude_globs( self, patterns: impl IntoIterator<Item = impl Into<String>>, ) -> Self

Glob patterns (matched against the path relative to the analyzed root) that prune traversal entirely — an excluded directory is never descended into, not merely omitted from the report.

pub fn min_size(self, bytes: u64) -> Self

pub fn max_size(self, bytes: u64) -> Self

pub fn modified_after(self, t: SystemTime) -> Self

Only files modified at or after t are matched. A file with no readable modified time never matches once this is set.

pub fn modified_before(self, t: SystemTime) -> Self

Only files modified at or before t are matched. A file with no readable modified time never matches once this is set.

pub fn max_depth(self, depth: usize) -> Self

Bounds how far the walk descends: the analyzed root is depth 0, its immediate children depth 1, and so on — passed straight through to jwalk’s own max_depth, which prunes traversal past the bound rather than filtering after the fact.

pub fn follow_symlinks(self, follow: bool) -> Self

Off by default. When enabled, jwalk’s own loop detection surfaces a symlink cycle as a per-entry error, handled like any other error via .on_error().

pub fn walk_concurrency(self, n: usize) -> Self

How many worker threads jwalk uses to read directories and stat entries concurrently while walking. Defaults to available_parallelism(), matching hash_concurrency.

Traversal is still reported (and aggregated into AnalysisReport) in a single stream on the calling task — this only parallelizes the underlying directory reads and stat calls, which is where wall time actually goes on large trees or slow/network filesystems.

pub fn top_n_largest(self, n: usize) -> Self

How many of the largest matched files to keep in AnalysisReport::largest_files. 0 disables the tracking entirely.

pub fn detect_mime_types(self, enable: bool) -> Self

Off by default — sniffing every matched file’s header via infer is an extra read per file, on top of the metadata walkdir already reads for every entry.

pub fn on_error(self, strategy: AnalysisErrorStrategy) -> Self

pub fn max_reported_errors(self, n: usize) -> Self

Caps AnalysisReport::errors; AnalysisReport::errors_total stays uncapped regardless of this setting.

pub fn start(self) -> Result<AnalysisHandle>