Zoter Connector: Bug when detecting URL

l0o0

Hello, I tested a translator on WanFang Medical, this is a demo link: https://med.wanfangdata.com.cn/Paper/Detail?id=PeriodicalPaper_zhyfyx200202011&dbid=WF_QK

This is the log from Zotero Connector when opening the demo link.

(3)(+0025153): Connector_Browser: onActivated for https://med.wanfangdata.com.cn/Paper/Detail?id=PeriodicalPaper_zhyfyx200202011&dbid=WF_QK

(3)(+0001834): HTTP GET moz-extension://69a3dbe3-db1b-4c79-8623-9dbcca705bc6/utilities/resource/dateFormats.json

******************** A unexpected URL is detected
(3)(+0000283): Connector_Browser: onFrameLoaded for https://med.wanfangdata.com.cn/Paper/Detail?id=PeriodicalPaper_zhyfyx200202011&dbid=WF_QK; https://medad.med.wanfangdata.com.cn/www/delivery/afr.php?zoneid=74&cb=INSERT_RANDOM_NUMBER_HERE
************************************

(3)(+0000000): Translators: Looking for translators for https://medad.med.wanfangdata.com.cn/www/delivery/afr.php?zoneid=74&cb=INSERT_RANDOM_NUMBER_HERE

(3)(+0000000): Not injecting. No translators found for [tab.url, url]: https://med.wanfangdata.com.cn/Paper/Detail?id=PeriodicalPaper_zhyfyx200202011&dbid=WF_QK , https://medad.med.wanfangdata.com.cn/www/delivery/afr.php?zoneid=74&cb=INSERT_RANDOM_NUMBER_HERE

(3)(+0000852): Translators: Looking for translators for https://med.wanfangdata.com.cn/Paper/Detail?id=PeriodicalPaper_zhyfyx200202011&dbid=WF_QK

(4)(+0000045): Translate: Binding sandbox to https://med.wanfangdata.com.cn/Paper/Detail?id=PeriodicalPaper_zhyfyx200202011&dbid=WF_QK


Zotero Connector found an ad iframe on the page, and use this ad URL(https://medad.med.wanfangdata.com.cn) to find available translators. However, the correct target URL is https://med.wanfangdata.com.cn/Paper/Detail?id.

I just search code in the Zotero Connector repo, it seems like an iframe URL popups when this iframe is activated.

l0o0

I know a quick way to fix this problem is to update the target regex. But I think is better to fix it on Zotero Connector side, maybe other sites will throw a different iframe.

dstillman

I'm not sure what you're saying the bug is here? Zotero will check all frames for one that matches a translator. As you can see from the debug output, it checks the /Paper/Detail?id URL as well. That just doesn't match the current target regex in the WanFang Data translator.

l0o0

@dstillman It's my fault. I only focus on the URL https://medad.med.wanfangdata.com.cn in the log. The target regex works fine in the translator scaffold but doesn't work in a real browser. I should have to double-check this problem.

ZoeCMa

Updating the target regex won't be a quick fix for this particular translator. There has to be an update to the translator itself.

l0o0

@ZoeCma Thanks for the comment. What do you mean update the translator itself? The translator code works well both with Wan Fang Data and Wan Fang Medical. It looks like the target regex doesn't match the Wan Fang Medical URL. I don't know why, but the target regex string is matched before.

Target regex: "^https?://[a-z]+\\.wanfangdata\\.com\\.cn"
Wan Fang Medical URL like: https://med.wanfangdata.com.cn/Paper/Detail?id=PeriodicalPaper_zhyfyx200202011&dbid=WF_QK

ZoeCMa

It's possible that we're either being served with different content at the same URL, or we're looking at different translators (mine has uuid eb876bd2-644c-458e-8d05-bf54b10176f3, last updated 2019-06-10 22:46:01).

This translator works on a page only if its getItemId() function returns truthy on the doc. This is not the case with the document I received at the URL. In addition the old export "endpoint" doesn't respond anymore.

l0o0

@ZoeCMa I think you refer to a translator in the Zotero Translator repo https://github.com/zotero/translators/blob/cc7c3c75a068aab2c2aa087f3ecbc72cf8eb72d4/Wanfang Data.js#L121 This is quite outdated(4 years ago). As I know, Wanfang updated their sites years ago. The old translator doesn't work anymore.

I am trying to update Wanfang Data.js in https://github.com/l0o0/translators_CN/blob/d1eb5ecbf75ff57876aee82d95756d453bb5920c/translators/Wanfang Data.js#L4

The target regex is updated and works well. After days of testing, I will make a pr to Zotero repo.