Merge pull request #36 from oiwn/dev

update deps, fix clippy warning

Author: oiwn

.deny.toml

@@ -4,5 +4,11 @@ allow = [
     "MIT",
     "Unicode-3.0",
     "MPL-2.0",
-    "ISC"
+    "ISC",
+    "BSD-3-Clause"
+]
+
+[advisories]
+ignore = [
+    "RUSTSEC-2025-0057"
 ]

.github/workflows/ci.yml

@@ -45,7 +45,7 @@ jobs:
     steps:
       - uses: actions/checkout@v4
       - uses: EmbarkStudios/cargo-deny-action@v2
-  
+
   # Run cargo clippy.
   lint-clippy:
     name: Check Clippy
@@ -68,15 +68,15 @@ jobs:
           globs: |
             '**/*.md'
             '!target'
-           
+
    # Run cargo check. This is a fast way to catch any obvious errors in the code.
   check:
     name: Check ${{ matrix.os }} ${{ matrix.toolchain }}
     strategy:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, windows-latest, macos-latest]
-        toolchain: ["1.85", "stable"]
+        toolchain: ["1.89", "stable"]
     runs-on: ${{ matrix.os }}
     steps:
       - uses: actions/checkout@v4

.tmuxp.yaml

@@ -7,12 +7,12 @@ windows:
     panes:
       - shell_command:
         - helix
-  - window_name: cmd
+  - window_name: agents
     start-directory: ./
     panes:
       - shell_command:
-        - eza
-  - window_name: srv
+        - claude
+  - window_name: cmd
     start-directory: ./
     panes:
       - shell_command:
@@ -21,3 +21,7 @@ windows:
     panes:
       - shell_command:
         - emacs -nw notes.org
+  - window_name: files
+    panes:
+      - shell_command:
+        - yazi

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "dom-content-extraction"
-version = "0.3.11"
+version = "0.3.12"
 
 description = "Rust implementation of Content extraction via text density paper"
 license = "MPL-2.0"
@@ -32,21 +32,20 @@ panic = "abort"
 
 [dependencies]
 ego-tree = "0.10"
-scraper = "0.23"
+scraper = "0.24"
 thiserror = "2"
 # binary
 clap = { version = "4.5", features = ["derive"], optional = true }
 reqwest = { version = "0.12", features = ["blocking"], optional = true }
-tempfile = { version = "3.19", optional = true }
+tempfile = { version = "3.22", optional = true }
 url = { version = "2.5", optional = true }
-anyhow = { version = "1.0", optional = true }
+anyhow = { version = "1", optional = true }
 unicode-normalization = "0.1"
 unicode-segmentation = "1.12"
 
 [dev-dependencies]
-criterion = "0.5"
-zip = "2.2"
-clap = { version = "4.5", features = ["derive"] }
+criterion = "0.7"
+zip = "5"
 anyhow = "1"
 regex = "1"
 rayon = "1"

benches/simple.rs

@@ -1,4 +1,4 @@
-use criterion::{black_box, criterion_group, criterion_main, Criterion};
+use criterion::{Criterion, black_box, criterion_group, criterion_main};
 use std::{fs, io::Read, path};
 use zip::read::ZipArchive;
 

examples/check.rs

@@ -1,6 +1,6 @@
 use clap::{Parser, Subcommand};
 use dom_content_extraction::{
-    get_content, get_node_text, scraper::Html, DensityTree,
+    DensityTree, get_content, get_node_text, scraper::Html,
 };
 use std::fs;
 

src/cetd.rs

@@ -221,13 +221,12 @@ impl<'a> DensityTree {
         }
 
         // Handle link char count for text within links
-        if let Some(parent) = node.parent() {
-            if let Some(element) = parent.value().as_element() {
-                if element.name() == "a" {
-                    density_node.value().metrics.link_char_count +=
-                        density_node.value().metrics.char_count;
-                }
-            }
+        if let Some(parent) = node.parent()
+            && let Some(element) = parent.value().as_element()
+            && element.name() == "a"
+        {
+            density_node.value().metrics.link_char_count +=
+                density_node.value().metrics.char_count;
         }
 
         // Update parent metrics by combining current node's metrics
@@ -280,7 +279,7 @@ impl<'a> DensityTree {
     ///     println!("Max density sum: {:?}", max_node.value().density_sum);
     /// }
     /// ```
-    pub fn get_max_density_sum_node(&self) -> Option<NodeRef<DensityNode>> {
+    pub fn get_max_density_sum_node(&self) -> Option<NodeRef<'_, DensityNode>> {
         self.tree.nodes().max_by(|a, b| {
             a.value()
                 .density_sum

src/utils.rs

@@ -78,10 +78,10 @@ pub fn get_node_links(
     let mut links: Vec<String> = vec![];
     let root_node = get_node_by_id(node_id, document)?;
     for node in root_node.descendants() {
-        if let Some(elem) = node.value().as_element() {
-            if let Some(link) = elem.attr("href") {
-                links.push(link.trim().to_string());
-            };
+        if let Some(elem) = node.value().as_element()
+            && let Some(link) = elem.attr("href")
+        {
+            links.push(link.trim().to_string());
         };
     }
     Ok(links)