sec-data-fetcher - npm Package Compare versions

Comparing version 1.0.1 to 1.0.2

dist/example.js

@@ -36,3 +36,7 @@ "use strict";
     }
+    // Extract tables from a URL
+    const filingUrl = 'https://www.sec.gov/Archives/edgar/data/0000320193/0000320193-21-000010.txt';
+    const tablesFromUrl = yield secClient.extractTablesFromFilingUrl(filingUrl);
+    //console.log('Extracted Tables from URL:', tablesFromUrl);
 }))();
 //# sourceMappingURL=example.js.map

dist/secAPI.d.ts

@@ -60,2 +60,20 @@ export interface CompanyTicker {
     /**
+     * Fetches the SEC filing content from a given URL.
+     * @param url - The URL of the SEC filing.
+     * @returns The raw HTML content of the filing.
+     */
+    fetchFiling(url: string): Promise<string>;
+    /**
+     * Extracts tables from a given SEC filing URL.
+     * @param url - The URL of the SEC filing.
+     * @returns An array of tables extracted from the filing.
+     */
+    extractTablesFromFilingUrl(url: string): Promise<Array<Array<Array<string>>>>;
+    /**
+     * Extracts tables from the provided HTML content of an SEC filing.
+     * @param filingContent - The raw HTML content of the SEC filing.
+     * @returns An array of tables extracted from the content.
+     */
+    extractTablesFromContent(filingContent: string): Array<Array<Array<string>>>;
+    /**
      * Parses SEC filing content from a string into an object.
@@ -62,0 +80,0 @@ * @param content - The SEC filing content as a string.

dist/secAPI.js

@@ -17,2 +17,3 @@ "use strict";
 const utils_1 = require("./utils");
+const tableParser_1 = require("./parsers/tableParser");
 class SECClient {
@@ -122,2 +123,44 @@ /**
     /**
+     * Fetches the SEC filing content from a given URL.
+     * @param url - The URL of the SEC filing.
+     * @returns The raw HTML content of the filing.
+     */
+    fetchFiling(url) {
+        return __awaiter(this, void 0, void 0, function* () {
+            try {
+                const response = yield this.http.get(url, {
+                    headers: {
+                        'User-Agent': this.userAgent,
+                        'Accept-Encoding': 'gzip, deflate',
+                        Host: 'www.sec.gov',
+                    },
+                });
+                return response.data; // The raw HTML of the SEC filing
+            }
+            catch (error) {
+                throw new Error(`Failed to fetch SEC filing from URL: ${url}. Error: ${error}`);
+            }
+        });
+    }
+    /**
+     * Extracts tables from a given SEC filing URL.
+     * @param url - The URL of the SEC filing.
+     * @returns An array of tables extracted from the filing.
+     */
+    extractTablesFromFilingUrl(url) {
+        return __awaiter(this, void 0, void 0, function* () {
+            const filingContent = yield this.fetchFiling(url); // Fetch the raw HTML content
+            const tables = (0, tableParser_1.extractTables)(filingContent); // Extract tables using the parsing function
+            return tables;
+        });
+    }
+    /**
+     * Extracts tables from the provided HTML content of an SEC filing.
+     * @param filingContent - The raw HTML content of the SEC filing.
+     * @returns An array of tables extracted from the content.
+     */
+    extractTablesFromContent(filingContent) {
+        return (0, tableParser_1.extractTables)(filingContent); // Extract tables from the provided HTML content
+    }
+    /**
      * Parses SEC filing content from a string into an object.
@@ -124,0 +167,0 @@ * @param content - The SEC filing content as a string.

package.json

 {
   "name": "sec-data-fetcher",
-  "version": "1.0.1",
+  "version": "1.0.2",
   "description": "An npm library to fetch SEC data from all supported endpoints with rate limiting.",
@@ -13,2 +13,3 @@ "main": "dist/index.js",
     "lint:fix": "eslint 'src/**/*.ts' --fix",
+    "example": "node dist/example.js",
     "clean": "rm -rf dist",
@@ -23,2 +24,4 @@ "prepare": "npm run build"
     "finance",
+    "parsing",
+    "formatting",
     "api",
@@ -33,2 +36,3 @@ "data",
     "axios-rate-limit": "^1.3.0",
+    "cheerio": "^1.0.0",
     "fast-xml-parser": "^4.2.4"
@@ -38,2 +42,3 @@ },
     "@types/axios": "^0.14.0",
+    "@types/cheerio": "^0.22.35",
     "@types/jest": "^29.5.13",
@@ -40,0 +45,0 @@ "@types/node": "^20.4.2",

README.md

@@ -79,3 +79,33 @@ # SEC Data Fetcher
 - **getObjectFromUrl(url: string):** Promise<FilingObject>
+- **fetchFiling(url: string):** Promise<string>  
+  Fetches the raw HTML content of an SEC filing from the provided URL.
+- **extractTablesFromFilingUrl(url: string):** Promise<Array<Array<Array<string>>>>  
+  Fetches the filing content from a URL and extracts tables from the filing.
+- **extractTablesFromContent(filingContent: string):** Array<Array<Array<string>>>  
+  Extracts tables from the provided raw HTML content of an SEC filing.
 
+### Extract Tables from SEC Filings
+
+The package provides two ways to extract tables from SEC filings:
+
+1. **Extract from a URL**: Fetch and extract tables directly from an SEC filing URL.
+2. **Extract from provided HTML content**: Extract tables from the provided HTML content when you already have the filing data.
+
+```typescript
+import { SECClient } from 'sec-data-fetcher';
+
+const secClient = new SECClient({
+  userAgent: 'Your Company <[e-mail]>',
+});
+
+// Extract tables from an SEC filing URL
+const tablesFromUrl = await secClient.extractTablesFromFilingUrl(
+  'https://www.sec.gov/...',
+);
+
+// Extract tables from provided HTML content
+const filingContent = '<html>...</html>';
+const tablesFromContent = secClient.extractTablesFromContent(filingContent);
+```
+
 ## Notes
@@ -82,0 +112,0 @@

src/example.ts

@@ -35,2 +35,8 @@ import { SECClient } from './secAPI'; // Adjust the import path as needed
   }
+
+  // Extract tables from a URL
+  const filingUrl =
+    'https://www.sec.gov/Archives/edgar/data/0000320193/0000320193-21-000010.txt';
+  const tablesFromUrl = await secClient.extractTablesFromFilingUrl(filingUrl);
+  //console.log('Extracted Tables from URL:', tablesFromUrl);
 })();

src/secAPI.ts

@@ -7,2 +7,3 @@ // src/secAPI.ts
 import { padCik } from './utils';
+import { extractTables } from './parsers/tableParser';
 
@@ -169,2 +170,48 @@ export interface CompanyTicker {
   /**
+   * Fetches the SEC filing content from a given URL.
+   * @param url - The URL of the SEC filing.
+   * @returns The raw HTML content of the filing.
+   */
+  public async fetchFiling(url: string): Promise<string> {
+    try {
+      const response = await this.http.get<string>(url, {
+        headers: {
+          'User-Agent': this.userAgent,
+          'Accept-Encoding': 'gzip, deflate',
+          Host: 'www.sec.gov',
+        },
+      });
+      return response.data; // The raw HTML of the SEC filing
+    } catch (error) {
+      throw new Error(
+        `Failed to fetch SEC filing from URL: ${url}. Error: ${error}`,
+      );
+    }
+  }
+
+  /**
+   * Extracts tables from a given SEC filing URL.
+   * @param url - The URL of the SEC filing.
+   * @returns An array of tables extracted from the filing.
+   */
+  public async extractTablesFromFilingUrl(
+    url: string,
+  ): Promise<Array<Array<Array<string>>>> {
+    const filingContent = await this.fetchFiling(url); // Fetch the raw HTML content
+    const tables = extractTables(filingContent); // Extract tables using the parsing function
+    return tables;
+  }
+
+  /**
+   * Extracts tables from the provided HTML content of an SEC filing.
+   * @param filingContent - The raw HTML content of the SEC filing.
+   * @returns An array of tables extracted from the content.
+   */
+  public extractTablesFromContent(
+    filingContent: string,
+  ): Array<Array<Array<string>>> {
+    return extractTables(filingContent); // Extract tables from the provided HTML content
+  }
+
+  /**
    * Parses SEC filing content from a string into an object.
@@ -171,0 +218,0 @@ * @param content - The SEC filing content as a string.

tests/secAPI.test.ts

@@ -29,2 +29,31 @@ import { SECClient } from '../src/secAPI';
   });
+
+  it('should fetch and extract tables from a filing URL', async () => {
+    // Mock filing URL from SEC (you can use a real one or mock the response)
+    const filingUrl =
+      'https://www.sec.gov/Archives/edgar/data/0000320193/0000320193-21-000010.txt';
+
+    const tables = await secClient.extractTablesFromFilingUrl(filingUrl);
+    expect(tables.length).toBeGreaterThan(0); // Ensure at least one table is extracted
+  });
+
+  it('should extract tables from provided filing content', () => {
+    const filingContent = `
+      <html>
+        <body>
+          <table>
+            <tr><th>Header 1</th><th>Header 2</th></tr>
+            <tr><td>Row 1 Col 1</td><td>Row 1 Col 2</td></tr>
+            <tr><td>Row 2 Col 1</td><td>Row 2 Col 2</td></tr>
+          </table>
+        </body>
+      </html>
+    `;
+
+    const tables = secClient.extractTablesFromContent(filingContent);
+    expect(tables.length).toBe(1);
+    expect(tables[0][0]).toEqual(['Header 1', 'Header 2']);
+    expect(tables[0][1]).toEqual(['Row 1 Col 1', 'Row 1 Col 2']);
+    expect(tables[0][2]).toEqual(['Row 2 Col 1', 'Row 2 Col 2']);
+  });
 });

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

48077

increased by36.16%

Number of package files

33

increased by17.86%

Lines of code

876

increased by34.56%

Number of lines in readme file

127

increased by30.93%

Worsened metrics

Dependency count

4

increased by33.33%

Dev dependency count

13

increased by8.33%

Dependency changes

• + Addedcheerio@^1.0.0

• + Addedboolbase@1.0.0(transitive)

• + Addedcheerio@1.0.0(transitive)

• + Addedcheerio-select@2.1.0(transitive)

• + Addedcss-select@5.1.0(transitive)

• + Addedcss-what@6.1.0(transitive)

• + Addeddom-serializer@2.0.0(transitive)

• + Addeddomelementtype@2.3.0(transitive)

• + Addeddomhandler@5.0.3(transitive)

• + Addeddomutils@3.1.0(transitive)

• + Addedencoding-sniffer@0.2.0(transitive)

• + Addedentities@4.5.0(transitive)

• + Addedhtmlparser2@9.1.0(transitive)

• + Addediconv-lite@0.6.3(transitive)

• + Addednth-check@2.1.1(transitive)

• + Addedparse5@7.2.1(transitive)

• + Addedparse5-htmlparser2-tree-adapter@7.1.0(transitive)

• + Addedparse5-parser-stream@7.1.2(transitive)

• + Addedsafer-buffer@2.1.2(transitive)

• + Addedundici@6.21.0(transitive)

• + Addedwhatwg-encoding@3.1.1(transitive)

• + Addedwhatwg-mimetype@4.0.0(transitive)