other_mirrors/ rss-to-activitypub
1
0
Fork 0
mirror of https://github.com/dariusk/rss-to-activitypub.git synced 2024-11-22 17:29:19 +02:00
Code Issues Packages Projects Releases Wiki Activity

iframes

This commit is contained in:
Darius Kazemi 2019-04-29 13:37:07 -07:00
parent 7cfe50605c
commit 01a8a5f336
1 changed files with 15 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
updateFeeds.js
View File

@ -95,7 +95,7 @@ function transformContent(item) {
} }
let $ = cheerio.load(item.content); let $ = cheerio.load(item.content);
// look through all the links // look through all the links to find images
let links = $('a'); let links = $('a');
let urls = []; let urls = [];
//console.log('links', links.length); //console.log('links', links.length);
@ -107,9 +107,10 @@ function transformContent(item) {
urls.push(url); urls.push(url);
} }
}); });
// look through all the images
let images = $('img'); let images = $('img');
images.each((i,e) => { images.each((i,e) => {
console.log(i,e);
let url = $(e).attr('src'); let url = $(e).attr('src');
// if there's an image, add it as a media attachment // if there's an image, add it as a media attachment
if (url && url.match(/(http)?s?:?(\/\/[^"']*\.(?:png|jpg|jpeg|gif|png|svg))/)) { if (url && url.match(/(http)?s?:?(\/\/[^"']*\.(?:png|jpg|jpeg|gif|png|svg))/)) {
@ -122,8 +123,17 @@ function transformContent(item) {
item.urls = urls; item.urls = urls;
// find iframe embeds and turn them into links
let iframes = $('iframe');
iframes.each((i,e) => {
console.log('iframe',i,e);
let url = $(e).attr('src');
$(e).replaceWith($(`<a href="${url}">[embedded content]</a>`));
});
// remove multiple line breaks // remove multiple line breaks
$('br').remove(); //$('br').remove();
$('p').each((i, el) => { $('p').each((i, el) => {
if($(el).html().replace(/\s|&nbsp;/g, '').length === 0) {$(el).remove();} if($(el).html().replace(/\s|&nbsp;/g, '').length === 0) {$(el).remove();}
}); });
@ -131,6 +141,8 @@ function transformContent(item) {
// couple of hacky regexes to make sure we clean up everything // couple of hacky regexes to make sure we clean up everything
item.content = $('body').html().replace(/^(\n|\r)/,'').replace(/>\r+</,' ').replace(/ +/g, ''); item.content = $('body').html().replace(/^(\n|\r)/,'').replace(/>\r+</,' ').replace(/ +/g, '');
item.content = item.content.replace(/^(\n|\r)/,'').replace(/>\r+</,' ').replace(/ +/g, ''); item.content = item.content.replace(/^(\n|\r)/,'').replace(/>\r+</,' ').replace(/ +/g, '');
// remove whitespace nodes from in between paragraphs
item.content = item.content.replace(/p>\s<p/g,'p><p');
return item; return item;
} }