Fixed some errors in translator for "ABC News Australia" #3222
Conversation
|
Hi! For some reason, my pull request is always failing at the The error message is kind of opaque. Is it a problem that my PR is failing here? |
|
The tests have been broken for a while, work to fix them is in progress in PR #3210 . |
ABC News Australia.js
Outdated
| @@ -38,13 +38,14 @@ | |||
|
|
|||
| function detectWeb(doc, _url) { | |||
| let contentType = attr(doc, 'meta[property="ABC.ContentType"]', 'content'); | |||
| if (contentType == 'CMChannel' && getSearchResults(doc, true)) { | |||
| contentType = contentType.toUpperCase(); // Case-insensitive treatment of content type | |||
There was a problem hiding this comment.
Style: one space before a comment in new code.
ABC News Australia.js
Outdated
| for (let i = 0; i < authorNameLinks.length; i++) { | ||
| let authorName = authorNameLinks[i].innerText; |
There was a problem hiding this comment.
| for (let i = 0; i < authorNameLinks.length; i++) { | |
| let authorName = authorNameLinks[i].innerText; | |
| for (let authorNameLink of authorNameLinks) { | |
| let authorName = authorNameLink.textContent; |
Prefer for..of over indexed for on arrays; textContent instead of innerText for reasons of convention and because innerText includes presentation (e.g. uppercase conversion that's specified in CSS) and textContent doesn't.
ABC News Australia.js
Outdated
| for (let i = 0; i < authorsList.length; i++) { | ||
| item.creators.push(ZU.cleanAuthor(authorsList[i], "author")); |
There was a problem hiding this comment.
| for (let i = 0; i < authorsList.length; i++) { | |
| item.creators.push(ZU.cleanAuthor(authorsList[i], "author")); | |
| for (let author of authorsList) { | |
| item.creators.push(ZU.cleanAuthor(author, "author")); |
|
Tests might work if you rebase. |
- fixed comment styling - prefer `let x of xs` sugar to indexed `for` loops - prefer `textContent` to `innerText`
Contributed the following to
ABC News Australia.js:contentTypecomparison for detection case-insensitive. Case-sensitivity was causing one of the test cases to fail because thecontentTypewasvideorather thanVideo.data-uriattribute starting withcoremedia://person/, when such elements exist. Previous method of author name detection is now used as a fallback. The previous method fails when author titles such as "weather correspondent" or "global affairs editor" are present.Before these modifications, the first test case failed and the second failed detection. Now the first test case passes and the second passes detection, but still fails due to an incorrect
abstractNote. But this appears to be an issue with a different translator that is being used here (whichever one has GUID equal to951c027d-74ac-47d4-a107-9c3069ab7b48).P.S. This is my first contribution, just learning how Zotero translators work! :-)