defuddle 0.2.3 → 0.3.0

package/README.md

@@ -34,6 +34,19 @@ console.log(article.content);  // HTML string of the main content
 console.log(article.title);    // Title of the article
 ```
 
+### Debug mode
+
+You can enable debug mode by passing an options object when creating a new Defuddle instance:
+
+```typescript
+const article = new Defuddle(document, { debug: true }).parse();
+```
+
+- More verbose console logging about the parsing process
+- Preserves HTML class and id attributes that are normally stripped
+- Retains all data-* attributes
+- Skips div flattening to preserve document structure
+
 ### Server-side usage
 
 When using Defuddle in a Node.js environment, you can use JSDOM to create a DOM document:
@@ -65,10 +78,12 @@ The `parse()` method returns an object with the following properties:
 | `domain` | string | Domain name of the website |
 | `favicon` | string | URL of the website's favicon |
 | `image` | string | URL of the article's main image |
+| `parseTime` | number | Time taken to parse the page in milliseconds |
 | `published` | string | Publication date of the article |
 | `author` | string | Author of the article |
 | `site` | string | Name of the website |
 | `schemaOrgData` | object | Raw schema.org data extracted from the page |
+| `wordCount` | number | Total number of words in the extracted content |
 
 ## HTML standardization
 

package/dist/constants.d.ts

@@ -0,0 +1,14 @@
+export declare const ENTRY_POINT_ELEMENTS: string[];
+export declare const MOBILE_WIDTH = 600;
+export declare const BLOCK_ELEMENTS: string[];
+export declare const PRESERVE_ELEMENTS: Set<string>;
+export declare const INLINE_ELEMENTS: Set<string>;
+export declare const HIDDEN_ELEMENT_SELECTORS: string;
+export declare const EXACT_SELECTORS: string[];
+export declare const PARTIAL_SELECTORS: string[];
+export declare const FOOTNOTE_INLINE_REFERENCES: string;
+export declare const FOOTNOTE_LIST_SELECTORS: string;
+export declare const ALLOWED_EMPTY_ELEMENTS: Set<string>;
+export declare const ALLOWED_ATTRIBUTES: Set<string>;
+export declare const ALLOWED_ATTRIBUTES_DEBUG: Set<string>;
+export declare const SUPPORTED_LANGUAGES: Set<string>;

package/dist/defuddle.d.ts

@@ -13,17 +13,20 @@ export declare class Defuddle {
      * Parse the document and extract its main content
      */
     parse(): DefuddleResponse;
+    private countWords;
     private _log;
     private _evaluateMediaQueries;
     private applyMobileStyles;
     private removeHiddenElements;
     private removeClutter;
+    private flattenDivs;
     private cleanContent;
     private removeTrailingHeadings;
     private handleHeadings;
     private removeHtmlComments;
     private stripUnwantedAttributes;
     private removeEmptyElements;
+    private removeEmptyLines;
     private createFootnoteItem;
     private collectFootnotes;
     private findOuterFootnoteContainer;