open-graph-scraper - npm Package Compare versions

Comparing version 3.0.2 to 3.1.0

CHANGELOG.md

@@ -0,1 +1,4 @@
+### 3.1.0
+- Adding ability to extract meta from HTML string
+
 ### 3.0.2
@@ -2,0 +5,0 @@ - Adding CHANGELOG.md

lib/openGraphScraper.js

@@ -38,2 +38,10 @@ 'use strict';
 const setOptionsAndReturnOpenGraphResults = function (options, callback) {
+  if (options.html) {
+    if (options.url) {
+      return callback(true, { error: 'Must specify either `url` or `html`, not both', success: false, requestUrl: options.url, errorDetails: 'Must specify either `url` or `html`, not both' }, null);
+    }
+    const ogObject = extractMetaTags(options.html, options);
+    return callback(false, { data: ogObject, success: true }, null);
+  }
+
   let validate = utils.validate(options.url, options.timeout);
@@ -101,3 +109,2 @@
   const peekSize = options.peekSize || 1024;
-  let ogObject = {};
 
@@ -119,66 +126,75 @@ request(options, function (error, response, body) {
 
-      const $ = cheerio.load(body);
-      const meta = $('meta');
-      const keys = Object.keys(meta);
-
+      let ogObject = extractMetaTags(body, options);
       if (options.withCharset) {
         ogObject.charset = charset(response.headers, body, peekSize);
       }
+      return callback(null, ogObject, response);
+    }
+  });
+};
 
-      keys.forEach(function (key) {
-        if (!(meta[key].attribs && (meta[key].attribs.property || meta[key].attribs.name))) {
-          return;
+/*
+ * extract meta tags from html string
+ * @param string body - html string
+ * @param string options - options the user has set
+ */
+const extractMetaTags = (body, options) => {
+  let ogObject = {};
+  const $ = cheerio.load(body);
+  const meta = $('meta');
+  const keys = Object.keys(meta);
+
+  keys.forEach(function (key) {
+    if (!(meta[key].attribs && (meta[key].attribs.property || meta[key].attribs.name))) {
+      return;
+    }
+    const property = meta[key].attribs.property || meta[key].attribs.name;
+    const content = meta[key].attribs.content || meta[key].attribs.value;
+    fields.forEach(function (item) {
+      if (property === item.property) {
+        if (!item.multiple) {
+          ogObject[item.fieldName] = content;
+        } else if (!ogObject[item.fieldName]) {
+          ogObject[item.fieldName] = [content];
+        } else if (Array.isArray(ogObject[item.fieldName])) {
+          ogObject[item.fieldName].push(content);
         }
-        const property = meta[key].attribs.property || meta[key].attribs.name;
-        const content = meta[key].attribs.content || meta[key].attribs.value;
-        fields.forEach(function (item) {
-          if (property === item.property) {
-            if (!item.multiple) {
-              ogObject[item.fieldName] = content;
-            } else if (!ogObject[item.fieldName]) {
-              ogObject[item.fieldName] = [content];
-            } else if (Array.isArray(ogObject[item.fieldName])) {
-              ogObject[item.fieldName].push(content);
-            }
-          }
-        });
-      });
-
-      // set the ogImage or fallback to ogImageURL or ogImageSecureURL
-      ogObject.ogImage = ogObject.ogImage ? ogObject.ogImage : (ogObject.ogImageURL ? ogObject.ogImageURL : (ogObject.ogImageSecureURL ? ogObject.ogImageSecureURL : []));
-      if (!ogObject.ogImage || !ogObject.ogImage.length) {
-        delete ogObject['ogImage'];
       }
+    });
+  });
 
-      // sets up all the media stuff
-      ogObject = media.mediaSetup(ogObject, options);
+  // set the ogImage or fallback to ogImageURL or ogImageSecureURL
+  ogObject.ogImage = ogObject.ogImage ? ogObject.ogImage : (ogObject.ogImageURL ? ogObject.ogImageURL : (ogObject.ogImageSecureURL ? ogObject.ogImageSecureURL : []));
+  if (!ogObject.ogImage || !ogObject.ogImage.length) {
+    delete ogObject['ogImage'];
+  }
 
-      // Check for 'only get open graph info'
-      if (!options.onlyGetOpenGraphInfo) {
-        // Get title tag if og title was not provided
-        if (!ogObject.ogTitle && $('head > title').text() && $('head > title').text().length > 0) {
-          ogObject.ogTitle = $('head > title').text();
+  // sets up all the media stuff
+  ogObject = media.mediaSetup(ogObject, options);
+
+  // Check for 'only get open graph info'
+  if (!options.onlyGetOpenGraphInfo) {
+    // Get title tag if og title was not provided
+    if (!ogObject.ogTitle && $('head > title').text() && $('head > title').text().length > 0) {
+      ogObject.ogTitle = $('head > title').text();
+    }
+    // Get meta description tag if og description was not provided
+    if (!ogObject.ogDescription && $('head > meta[name="description"]').attr('content') && $('head > meta[name="description"]').attr('content').length > 0) {
+      ogObject.ogDescription = $('head > meta[name="description"]').attr('content');
+    }
+    // Get first image as og:image if there is no og:image tag.
+    const ogImageFallback = options.ogImageFallback === undefined ? true : options.ogImageFallback;
+    if (!ogObject.ogImage && ogImageFallback) {
+      const supportedImageExts = ['jpg', 'jpeg', 'png'];
+      $('img').each(function (i, elem) {
+        if ($(elem).attr('src') && $(elem).attr('src').length > 0 && supportedImageExts.indexOf($(elem).attr('src').split('.').pop()) !== -1) {
+          ogObject.ogImage = {
+            url: $(elem).attr('src')
+          };
+          return false;
         }
-        // Get meta description tag if og description was not provided
-        if (!ogObject.ogDescription && $('head > meta[name="description"]').attr('content') && $('head > meta[name="description"]').attr('content').length > 0) {
-          ogObject.ogDescription = $('head > meta[name="description"]').attr('content');
-        }
-        // Get first image as og:image if there is no og:image tag.
-        const ogImageFallback = options.ogImageFallback === undefined ? true : options.ogImageFallback;
-        if (!ogObject.ogImage && ogImageFallback) {
-          const supportedImageExts = ['jpg', 'jpeg', 'png'];
-          $('img').each(function (i, elem) {
-            if ($(elem).attr('src') && $(elem).attr('src').length > 0 && supportedImageExts.indexOf($(elem).attr('src').split('.').pop()) !== -1) {
-              ogObject.ogImage = {
-                url: $(elem).attr('src')
-              };
-              return false;
-            }
-          });
-        }
-      }
-      // console.log('ogObject',ogObject);
-      return callback(null, ogObject, response);
+      });
     }
-  });
+  }
+  return ogObject;
 };
@@ -185,0 +201,0 @@

package.json

 {
   "name": "open-graph-scraper",
   "description": "Node.js scraper module for Open Graph and Twitter Card info",
-  "version": "3.0.2",
+  "version": "3.1.0",
   "license": "MIT",
@@ -6,0 +6,0 @@ "main": "index.js",

README.md

@@ -101,2 +101,14 @@ openGraphScraper
 
+It's possible to pass in an HTML string instead of a URL. There won't be a resonse object.
+```
+const htmlString = /* html string goes here */;
+const ogs = require('open-graph-scraper');
+const options = {'html': htmlString};
+ogs(options, function (error, results) {
+  console.log('error:', error); // This is returns true or false. True if there was a error. The error it self is inside the results object.
+  console.log('results:', results);
+});
+```
+
+
 ### Results JSON
@@ -103,0 +115,0 @@ Check the return for a ```success``` flag. If success is set to true, then the url input was valid. Otherwise it will be set to false. The above example will return something like...

tests/index.spec.js

@@ -5,2 +5,10 @@ 'use strict';
 const expect = require('expect.js');
+const HTML_STRING = `
+<html>
+<head>
+  <meta property="og:title" content="Test page"/>
+</head>
+<body></body>
+</html>
+`;
 
@@ -788,2 +796,28 @@ describe('GET OG', function () {
   });
+  it('Invalid Call - Can\'t request URL and pass in HTML string', function (done) {
+    app({
+      'url': 'https://upload.wikimedia.org/wikipedia/commons/a/a2/Overlook_Hong_Kong_Island_north_coast,_Victoria_Harbour_and_Kowloon_from_middle_section_of_Lugard_Road_at_daytime_(enlarged_version_and_better_contrast,_revised).jpg',
+      'html': HTML_STRING
+    }, function (error, result) {
+      console.log('error:', error);
+      console.log('result:', result);
+      expect(error).to.be(true);
+      expect(result.success).to.be(false);
+      expect(result.requestUrl).to.be('https://upload.wikimedia.org/wikipedia/commons/a/a2/Overlook_Hong_Kong_Island_north_coast,_Victoria_Harbour_and_Kowloon_from_middle_section_of_Lugard_Road_at_daytime_(enlarged_version_and_better_contrast,_revised).jpg');
+      expect(result.error).to.be('Must specify either `url` or `html`, not both');
+      done();
+    });
+  });
+  it('Valid Call - pass in HTML string', function (done) {
+    app({
+      'html': HTML_STRING
+    }, function (error, result) {
+      console.log('error:', error);
+      console.log('result:', result);
+      expect(error).to.be(false);
+      expect(result.success).to.be(true);
+      expect(result.data.ogTitle).to.be('Test page');
+      done();
+    });
+  });
 });

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

90081

increased by2.68%

Lines of code

1909

increased by2.63%

Number of lines in readme file

152

increased by8.57%

No dependency changes