Struct FilesystemAdapter 

pub struct FilesystemAdapter { /* private fields */ }

Filesystem source adapter.

Recursively walks a root directory, applying extension and glob filters during enumeration. Fetches file content and computes blake3 hashes.

Implementations

impl FilesystemAdapter

pub fn from_spec( source_name: &str, spec: &SourceSpec, ) -> Result<Self, ResolveError>

Create a new FilesystemAdapter from a SourceSpec.

Errors

Returns ResolveError::UnknownAdapter if the spec is not a filesystem source. Returns ResolveError::Io if a glob pattern is invalid.

pub fn from_fs_spec( source_name: &str, spec: &FilesystemSourceSpec, ) -> Result<Self, ResolveError>

Create a new FilesystemAdapter directly from a FilesystemSourceSpec.

Errors

Returns ResolveError::Io if a glob pattern is invalid.

Trait Implementations

impl Debug for FilesystemAdapter

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

Formats the value using the given formatter.

impl SourceAdapter for FilesystemAdapter

fn source_kind(&self) -> &str

Human-readable name of the source type (e.g., “Google Drive”).

fn enumerate<'life0, 'async_trait>( &'life0 self, ) -> Pin<Box<dyn Future<Output = Result<Vec<SourceItem>, SourceError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Enumerate items available from this source. Returns lightweight descriptors (no content) for filtering and hash comparison. This is the “what’s there?” step.

fn fetch<'life0, 'life1, 'async_trait>( &'life0 self, item: &'life1 SourceItem, ) -> Pin<Box<dyn Future<Output = Result<ExtractedDocument, SourceError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Fetch the full content of a single item. Separate from enumerate() because fetching is expensive and we want to skip unchanged items before paying this cost.