Struct Scanner 

pub struct Scanner { /* private fields */ }

Scanner: The Rock Star of our Show

BEHOLD! The Scanner itself! This is the main act. It takes the config, the ignore lists, and a path, and it puts on a spectacular show of directory traversal. It’s fast, it’s smart, and it knows all the best moves.

Implementations

impl Scanner

pub fn root(&self) -> &Path

Returns the canonicalized root path of the scanner

pub fn quick_scan(&self) -> Result<(Vec<FileNode>, TreeStats)>

Quick scan for basic project analysis - lighter weight than full scan Returns only basic stats and key files for faster integration

pub fn find_recent_files(&self, hours_ago: u64) -> Result<Vec<FileNode>>

Find files modified within a specific time range Useful for finding recent activity in projects

pub fn find_key_files(&self) -> Result<Vec<FileNode>>

Get key project files (build configs, main files, etc.) Returns a filtered list of important files for project analysis

pub fn new(root: &Path, config: ScannerConfig) -> Result<Self>

Scanner::new - Constructor

Creates a new Scanner instance. This involves:

1. Storing the provided config and root path.
2. Loading and compiling .gitignore patterns if config.respect_gitignore is true.
3. Compiling the DEFAULT_IGNORE_PATTERNS if config.use_default_ignores is true.
4. Initializing sets of system paths and specific files to always ignore.

This setup prepares the scanner for efficient should_ignore checks during traversal.

pub fn scan_stream(&self, sender: Sender<FileNode>) -> Result<TreeStats>

Stream nodes as they are discovered This version of scan is optimized for the --stream flag. It sends FileNode objects through the sender channel as soon as they are processed. This allows the formatter to start displaying output immediately, which is great for large directories. Returns the final TreeStats once the scan is complete.

pub fn scan(&self) -> Result<(Vec<FileNode>, TreeStats)>

scan - The Full Scan (Non-Streaming)

Performs a complete directory scan, collecting all FileNodes that meet the criteria (not ignored, or shown if ignored, and pass filters if any). This method first traverses the entire directory structure defined by config.max_depth, creating FileNode objects for each entry. It then performs a second pass if filters are active to ensure that directories are only included if they (or their subdirectories) contain files that match the filters. Returns a tuple: (Vec<FileNode>, TreeStats).

scan - The “Scan-It-All-Then-Sort-It-Out” Method

This is the classic way to scan. It’s a two-act show:

1. Act I: Walk through every single file and directory, collecting a huge list of FileNodes.
2. Act II: If there are filters, go through that huge list and pick out only the ones that match, making sure to keep their parent directories so the tree still makes sense. It’s thorough and great for when you need the whole picture before making decisions.