FOSS
/
urltomarkdown
mirror of https://github.com/macsplit/urltomarkdown
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

further adjusted filters for wikipedia

main
Lee Hanken 2022-02-03 18:42:54 +00:00
parent 04785ccd02
commit 46c6bb4e6a
1 changed files with 22 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
url_to_markdown_common_filters.js
View File

@ -1,5 +1,3 @@
var urlparser = require('url');
module.exports = { module.exports = {
list: [ list: [
@ -11,7 +9,7 @@ module.exports = {
replace: [ replace: [
{ {
find: /\[([^\]]*)\]\(\/\/([^\)]*)\)/g, find: /\[([^\]]*)\]\(\/\/([^\)]*)\)/g,
replacement: '[$1](https://$2)' replacement: '[$1](https://$2)'
} }
] ]
}, },
@ -19,7 +17,9 @@ module.exports = {
domain: /.*\.wikipedia\.org/, domain: /.*\.wikipedia\.org/,
remove: [ remove: [
/\*\*\[\^\]\(#cite_ref[^\)]+\)\*\*/g, /\*\*\[\^\]\(#cite_ref[^\)]+\)\*\*/g,
/(?:\\\[)?\[edit\]\([^\s]+\s+"[^"]+"\)(?:\\\])?/ig /(?:\\\[)?\[edit\]\([^\s]+\s+"[^"]+"\)(?:\\\])?/ig,
/\^\s\[Jump up to[^\)]*\)/ig,
/\[[^]*\]\(#cite_ref[^\)]+\)/g
] ]
}, },
{ {
@ -34,7 +34,16 @@ module.exports = {
], ],
filter: function (url, data) { filter: function (url, data) {
let domain = urlparser.parse(url).hostname let domain='';
let base_address='';
if (url) {
url = new URL(url);
if (url) {
base_address = url.protocol+"//"+url.hostname;
domain = url.hostname;
}
}
for (let i=0;i<this.list.length;i++) { for (let i=0;i<this.list.length;i++) {
if (domain.match(this.list[i].domain)) { if (domain.match(this.list[i].domain)) {
if (this.list[i].remove) { if (this.list[i].remove) {
@ -50,6 +59,14 @@ module.exports = {
} }
} }
} }
// this filter needs to be defined here, to access the URL
data = data.replaceAll(/\[([^\]]*)\]\(\/([^\/][^\)]*)\)/g,
(match, title, address) => {
return "["+title+"]("+base_address+"/"+address+")";
}
);
return data; return data;
} }