From 7b97b52af021500d8085c875d20215e8dc0f53cc Mon Sep 17 00:00:00 2001
From: 魏曹先生 <1992414357@qq.com>
Date: Mon, 17 Nov 2025 11:49:49 +0800
Subject: feat: Add file status tracking and SHA1 hash system

- Implement SHA1 hash calculation module with async support
- Add file status analysis for tracking moves, creates, and modifications
- Enhance local file management with relative path handling
- Update virtual file actions with improved tracking capabilities
---
 crates/utils/sha1_hash/Cargo.toml           |   9 ++
 crates/utils/sha1_hash/res/story.sha1       |   1 +
 crates/utils/sha1_hash/res/story.txt        |  48 +++++++
 crates/utils/sha1_hash/src/lib.rs           | 197 ++++++++++++++++++++++++++++
 crates/utils/string_proc/src/format_path.rs |  44 ++++++-
 5 files changed, 296 insertions(+), 3 deletions(-)
 create mode 100644 crates/utils/sha1_hash/Cargo.toml
 create mode 100644 crates/utils/sha1_hash/res/story.sha1
 create mode 100644 crates/utils/sha1_hash/res/story.txt
 create mode 100644 crates/utils/sha1_hash/src/lib.rs

(limited to 'crates/utils')

diff --git a/crates/utils/sha1_hash/Cargo.toml b/crates/utils/sha1_hash/Cargo.toml
new file mode 100644
index 0000000..e206efd
--- /dev/null
+++ b/crates/utils/sha1_hash/Cargo.toml
@@ -0,0 +1,9 @@
+[package]
+name = "sha1_hash"
+edition = "2024"
+version.workspace = true
+
+[dependencies]
+tokio = { version = "1.48", features = ["full"] }
+sha1 = "0.10"
+futures = "0.3"
diff --git a/crates/utils/sha1_hash/res/story.sha1 b/crates/utils/sha1_hash/res/story.sha1
new file mode 100644
index 0000000..c2e3213
--- /dev/null
+++ b/crates/utils/sha1_hash/res/story.sha1
@@ -0,0 +1 @@
+6838aca280112635a2cbf93440f4c04212f58ee8
diff --git a/crates/utils/sha1_hash/res/story.txt b/crates/utils/sha1_hash/res/story.txt
new file mode 100644
index 0000000..a91f467
--- /dev/null
+++ b/crates/utils/sha1_hash/res/story.txt
@@ -0,0 +1,48 @@
+魏曹者，程序员也，发稀甚于代码。
+忽接神秘电话曰：
+"贺君中彩，得长生之赐。"
+魏曹冷笑曰："吾命尚不及下版之期。"
+
+翌日果得U盘。
+接入电脑，弹窗示曰：
+"点此确认，即获永生。"
+魏曹径点"永拒"。
+
+三月后，U盘自格其盘。
+进度条滞于九九。
+客服电话已成空号。
+魏曹乃知身可不死，然体内癌细胞亦得不灭。
+
+遂谒主请辞。
+主曰："巧甚，公司正欲优化。"
+魏曹曰："吾不死。"
+主目骤亮："则可007至司闭。"
+
+魏曹始试诸死法。
+坠楼，卧医三月，账单令其愿死。
+饮鸩，肝肾永损，然终不得死。
+终决卧轨。
+
+择高铁最速者。
+司机探头曰："兄台，吾亦不死身也。"
+"此车已碾如君者二十人矣。"
+
+二人遂坐轨畔对饮。
+司机曰："知最讽者何？"
+"吾等永存，而所爱者皆逝矣。"
+
+魏曹忽得系统提示：
+"侦得用户消极求生，将启工模。"
+自是无日不毕KPI，否则遍尝绝症之苦。
+
+是日对镜整寿衣。
+忽见顶生一丝乌发。
+泫然泣下，此兆示其将复活一轮回。
+
+--- 忽忆DeepSeek尝作Footer曰：
+"文成而Hash1验，若星河之固。"
+遂取哈希值校之，
+字符流转如天河倒泻，
+终得"e3b0c44298fc1c14"之数。
+然文末数字竟阙如残月，
+方知此篇亦遭永劫轮回。
diff --git a/crates/utils/sha1_hash/src/lib.rs b/crates/utils/sha1_hash/src/lib.rs
new file mode 100644
index 0000000..4373d50
--- /dev/null
+++ b/crates/utils/sha1_hash/src/lib.rs
@@ -0,0 +1,197 @@
+use sha1::{Digest, Sha1};
+use std::path::{Path, PathBuf};
+use std::sync::Arc;
+use tokio::fs::File;
+use tokio::io::{AsyncReadExt, BufReader};
+use tokio::task;
+
+#[derive(Debug, Clone)]
+pub struct Sha1Result {
+    pub file_path: PathBuf,
+    pub hash: String,
+}
+
+/// Calc SHA1 hash of a single file
+pub async fn calc_sha1<P: AsRef<Path>>(
+    path: P,
+    buffer_size: usize,
+) -> Result<Sha1Result, Box<dyn std::error::Error + Send + Sync>> {
+    let file_path = path.as_ref().to_string_lossy().to_string();
+
+    // Open file asynchronously
+    let file = File::open(&path).await?;
+    let mut reader = BufReader::with_capacity(buffer_size, file);
+    let mut hasher = Sha1::new();
+    let mut buffer = vec![0u8; buffer_size];
+
+    // Read file in chunks and update hash asynchronously
+    loop {
+        let n = reader.read(&mut buffer).await?;
+        if n == 0 {
+            break;
+        }
+        hasher.update(&buffer[..n]);
+    }
+
+    let hash_result = hasher.finalize();
+
+    // Convert to hex string
+    let hash_hex = hash_result
+        .iter()
+        .map(|b| format!("{:02x}", b))
+        .collect::<String>();
+
+    Ok(Sha1Result {
+        file_path: file_path.into(),
+        hash: hash_hex,
+    })
+}
+
+/// Calc SHA1 hashes for multiple files using multi-threading
+pub async fn calc_sha1_multi<P, I>(
+    paths: I,
+    buffer_size: usize,
+) -> Result<Vec<Sha1Result>, Box<dyn std::error::Error + Send + Sync>>
+where
+    P: AsRef<Path> + Send + Sync + 'static,
+    I: IntoIterator<Item = P>,
+{
+    let buffer_size = Arc::new(buffer_size);
+
+    // Collect all file paths
+    let file_paths: Vec<P> = paths.into_iter().collect();
+
+    if file_paths.is_empty() {
+        return Ok(Vec::new());
+    }
+
+    // Create tasks for each file
+    let tasks: Vec<_> = file_paths
+        .into_iter()
+        .map(|path| {
+            let buffer_size = Arc::clone(&buffer_size);
+            task::spawn(async move { calc_sha1(path, *buffer_size).await })
+        })
+        .collect();
+
+    // Execute tasks with concurrency limit using join_all
+    let results: Vec<Result<Sha1Result, Box<dyn std::error::Error + Send + Sync>>> =
+        futures::future::join_all(tasks)
+            .await
+            .into_iter()
+            .map(|task_result| match task_result {
+                Ok(Ok(calc_result)) => Ok(calc_result),
+                Ok(Err(e)) => Err(e),
+                Err(e) => Err(Box::new(e) as Box<dyn std::error::Error + Send + Sync>),
+            })
+            .collect();
+
+    // Check for any errors and collect successful results
+    let mut successful_results = Vec::new();
+    for result in results {
+        match result {
+            Ok(success) => successful_results.push(success),
+            Err(e) => return Err(e),
+        }
+    }
+
+    Ok(successful_results)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use std::fs;
+
+    #[tokio::test]
+    async fn test_sha1_accuracy() {
+        // Test file path relative to the crate root
+        let test_file_path = "res/story.txt";
+        let expected_hash_path = "res/story.sha1";
+
+        // Calculate SHA1 hash
+        let result = calc_sha1(test_file_path, 8192)
+            .await
+            .expect("Failed to calculate SHA1");
+
+        // Read expected hash from file
+        let expected_hash = fs::read_to_string(expected_hash_path)
+            .expect("Failed to read expected hash file")
+            .trim()
+            .to_string();
+
+        // Verify the calculated hash matches expected hash
+        assert_eq!(
+            result.hash, expected_hash,
+            "SHA1 hash mismatch for test file"
+        );
+
+        println!("Test file: {}", result.file_path.display());
+        println!("Calculated hash: {}", result.hash);
+        println!("Expected hash: {}", expected_hash);
+    }
+
+    #[tokio::test]
+    async fn test_sha1_empty_file() {
+        // Create a temporary empty file for testing
+        let temp_file = "test_empty.txt";
+        fs::write(temp_file, "").expect("Failed to create empty test file");
+
+        let result = calc_sha1(temp_file, 4096)
+            .await
+            .expect("Failed to calculate SHA1 for empty file");
+
+        // SHA1 of empty string is "da39a3ee5e6b4b0d3255bfef95601890afd80709"
+        let expected_empty_hash = "da39a3ee5e6b4b0d3255bfef95601890afd80709";
+        assert_eq!(
+            result.hash, expected_empty_hash,
+            "SHA1 hash mismatch for empty file"
+        );
+
+        // Clean up
+        fs::remove_file(temp_file).expect("Failed to remove temporary test file");
+    }
+
+    #[tokio::test]
+    async fn test_sha1_simple_text() {
+        // Create a temporary file with simple text
+        let temp_file = "test_simple.txt";
+        let test_content = "Hello, SHA1!";
+        fs::write(temp_file, test_content).expect("Failed to create simple test file");
+
+        let result = calc_sha1(temp_file, 4096)
+            .await
+            .expect("Failed to calculate SHA1 for simple text");
+
+        // Note: This test just verifies that the function works without errors
+        // The actual hash value is not critical for this test
+
+        println!("Simple text test - Calculated hash: {}", result.hash);
+
+        // Clean up
+        fs::remove_file(temp_file).expect("Failed to remove temporary test file");
+    }
+
+    #[tokio::test]
+    async fn test_sha1_multi_files() {
+        // Test multiple files calculation
+        let test_files = vec!["res/story.txt"];
+
+        let results = calc_sha1_multi(test_files, 8192)
+            .await
+            .expect("Failed to calculate SHA1 for multiple files");
+
+        assert_eq!(results.len(), 1, "Should have calculated hash for 1 file");
+
+        // Read expected hash from file
+        let expected_hash = fs::read_to_string("res/story.sha1")
+            .expect("Failed to read expected hash file")
+            .trim()
+            .to_string();
+
+        assert_eq!(
+            results[0].hash, expected_hash,
+            "SHA1 hash mismatch in multi-file test"
+        );
+    }
+}
diff --git a/crates/utils/string_proc/src/format_path.rs b/crates/utils/string_proc/src/format_path.rs
index 2d8927b..a3493f5 100644
--- a/crates/utils/string_proc/src/format_path.rs
+++ b/crates/utils/string_proc/src/format_path.rs
@@ -3,6 +3,7 @@ use std::path::PathBuf;
 /// Format path str
 pub fn format_path_str(path: impl Into<String>) -> Result<String, std::io::Error> {
     let path_str = path.into();
+    let ends_with_slash = path_str.ends_with('/');
 
     // ANSI Strip
     let cleaned = strip_ansi_escapes::strip(&path_str);
@@ -27,10 +28,43 @@ pub fn format_path_str(path: impl Into<String>) -> Result<String, std::io::Error
         .filter(|c| !unfriendly_chars.contains(c))
         .collect();
 
-    if result.ends_with('/') {
-        Ok(result)
+    // Handle ".." path components
+    let path_buf = PathBuf::from(&result);
+    let normalized_path = normalize_path(&path_buf);
+    result = normalized_path.to_string_lossy().replace('\\', "/");
+
+    // Restore trailing slash if original path had one
+    if ends_with_slash && !result.ends_with('/') {
+        result.push('/');
+    }
+
+    Ok(result)
+}
+
+/// Normalize path by resolving ".." components without requiring file system access
+fn normalize_path(path: &PathBuf) -> PathBuf {
+    let mut components = Vec::new();
+
+    for component in path.components() {
+        match component {
+            std::path::Component::ParentDir => {
+                if !components.is_empty() {
+                    components.pop();
+                }
+            }
+            std::path::Component::CurDir => {
+                // Skip current directory components
+            }
+            _ => {
+                components.push(component);
+            }
+        }
+    }
+
+    if components.is_empty() {
+        PathBuf::from(".")
     } else {
-        Ok(result)
+        components.iter().collect()
     }
 }
 
@@ -58,6 +92,10 @@ mod tests {
             format_path_str("/home/user/file.txt")?,
             "/home/user/file.txt"
         );
+        assert_eq!(
+            format_path_str("/home/my_user/DOCS/JVCS_TEST/Workspace/../Vault/")?,
+            "/home/my_user/DOCS/JVCS_TEST/Vault/"
+        );
 
         Ok(())
     }
-- 
cgit