coming out with agent, disable text extraction from inside script tags

Author: oiwn

CLAUDE.md

@@ -0,0 +1,82 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Project Overview
+
+This is a Rust library implementing the Content Extraction via Text Density (CETD) algorithm for extracting main content from web pages. The core concept analyzes text density patterns to distinguish content-rich sections from navigational elements.
+
+## Architecture
+
+### Core Components
+
+- **`DensityTree`** (`src/cetd.rs`): Main structure representing text density analysis of HTML documents. Contains methods for building density trees, calculating metrics, and extracting content.
+- **`DensityNode`** (`src/cetd.rs`): Individual nodes containing text density metrics (character count, tag count, link density).
+- **Tree operations** (`src/tree.rs`): HTML document traversal and node metrics calculation.
+- **Unicode handling** (`src/unicode.rs`): Proper character counting using grapheme clusters and Unicode normalization.
+- **Utilities** (`src/utils.rs`): Helper functions for text extraction and link analysis.
+
+### Algorithm Flow
+
+1. Parse HTML document using `scraper::Html`
+2. Build density tree mirroring HTML structure (`DensityTree::from_document`)
+3. Calculate text density metrics for each node
+4. Compute composite density scores (`calculate_density_sum`)
+5. Extract high-density regions as main content
+
+### Binary Tool
+
+The `dce` binary (`src/main.rs`) provides CLI access to the library functionality, supporting both local files and URL fetching.
+
+## Development Commands
+
+### Build and Test
+```bash
+cargo build              # Build library
+cargo build --release    # Optimized build
+cargo test               # Run tests
+cargo bench              # Run benchmarks
+```
+
+### Code Quality
+```bash
+cargo fmt                # Format code (max_width = 84, see rustfmt.toml)
+cargo clippy             # Lint code
+cargo tarpaulin          # Generate coverage report (target: 80%+, see .tarpaulin.toml)
+just coverage            # Alternative coverage command (requires just)
+```
+
+### Examples
+```bash
+cargo run --example check -- lorem-ipsum    # Extract from generated lorem ipsum
+cargo run --example check -- test4          # Show highest density node
+cargo run --example ce_score                # Benchmark against CleanEval dataset
+```
+
+### Binary Usage
+```bash
+cargo run --bin dce -- --url "https://example.com"        # Extract from URL
+cargo run --bin dce -- --file input.html --output out.txt # Extract from file
+```
+
+## Project Structure
+
+- `src/lib.rs` - Main library interface and public API
+- `src/cetd.rs` - Core CETD algorithm implementation
+- `src/tree.rs` - HTML tree traversal and metrics
+- `src/unicode.rs` - Unicode-aware text processing
+- `src/utils.rs` - Text extraction utilities
+- `src/main.rs` - CLI binary implementation
+- `examples/` - Usage examples and benchmarking tools
+
+## Key Dependencies
+
+- `scraper` - HTML parsing and CSS selector support
+- `ego-tree` - Tree data structure for density calculations
+- `unicode-segmentation` - Proper Unicode grapheme handling
+- `unicode-normalization` - Text normalization for consistent processing
+
+## Features
+
+- Default features include CLI functionality (`cli` feature)
+- Library can be used without CLI dependencies by disabling default features

src/utils.rs

@@ -46,16 +46,36 @@ pub fn get_node_text(
 ) -> Result<String, DomExtractionError> {
     let mut text_fragments: Vec<String> = vec![];
     let root_node = get_node_by_id(node_id, document)?;
-    for node in root_node.descendants() {
-        if let Some(txt) = node.value().as_text() {
+    collect_text_filtered(&root_node, &mut text_fragments);
+    // Use the Unicode join function instead of simple join
+    Ok(crate::unicode::join_text_fragments(text_fragments))
+}
+
+/// Recursively collect text from nodes while filtering out script/style content
+fn collect_text_filtered(node: &ego_tree::NodeRef<'_, scraper::node::Node>, text_fragments: &mut Vec<String>) {
+    match node.value() {
+        scraper::Node::Text(txt) => {
             let clean_text = txt.trim();
             if !clean_text.is_empty() {
                 text_fragments.push(clean_text.to_string());
-            };
-        };
+            }
+        }
+        scraper::Node::Element(elem) => {
+            // Skip script, noscript, and style elements entirely
+            if !matches!(elem.name(), "script" | "noscript" | "style") {
+                // Process children only if this isn't a filtered element
+                for child in node.children() {
+                    collect_text_filtered(&child, text_fragments);
+                }
+            }
+        }
+        _ => {
+            // For other node types, process children
+            for child in node.children() {
+                collect_text_filtered(&child, text_fragments);
+            }
+        }
     }
-    // Use the Unicode join function instead of simple join
-    Ok(crate::unicode::join_text_fragments(text_fragments))
 }
 
 /// Helper function to extract all links (`href` attributes) from a `scraper::Html`