I am using the generic NMT and found a serious mistranslation:
Source- 若中英文条款有不一致之处，以中文条款为准
Target- In the event of any inconsistency between the Chinese and English terms, the English terms shall prevail
The correct translation should be "the Chinese terms shall prevail".
Not sure if there is a official way to report/ improve MT result, if so, could you please also provide me the link?
Thanks
You should raise this point in uservoice, as suggested on their portal here:
But it looks like the link is broken on this page, so I suggest using this uservoice forum: https://cognitive.uservoice.com/forums/558796-translator-api
Related
I've noticed a trend of using emojis for CSS classnames.
.🖖🏾-Vg{color:#ff4040}
.🖖🏾VDe{padding:.75rem 0;font-size:1rem}
It makes certain things more difficult. e.g. writing Selenium tests over these pages.
Is there a real benefit to using them? Security? Filesize?
Or are developers just doing this for kicks?
Edit: For the "Close (Opinion Base)" voters. I genuinely want to know if there's a development reason for doing this. I'm not looking for people's opinions here.
Im going to tenatively answer this while trying to not to be too 'Opinion based'
The 'emoji' support is a feature of supporting all Unicode characters, this was to support Chinese charachter support, which makes perfect sense.
As Emojis have been mapped to the Unicode chars, they came out of the wash too.
I have trouble finding legitimate references to bytes saved with emojis in-lieu of another method. So if someone could correct me that would be helpful.
The closest I found was a gitLab document from 2018 which moreso speaks to the performace improvements they saw implementing the native Unicode emojis.
GitLab Emoji Unicode
Appart from anything else though, I have seen some companies throw them into CSS files to attract some 'UI' enthusiasts while browsing the source of a site, for hiring purposes.
Opinion Spoiler - If I saw this in a company content, the last thing I would be doing is applying to work with that.
Final Note
This really is not useful in any practical way, if you are working as part of a team, ask them yourself how they would feel about searching through a source base using an emoji/unicode instead of some readable class/reference.
🥑
Reading Material
Browser Support SO Question
CanIUse Unicode
Unicode Release with Emoji Support
From what I have read from a forum, the reason people use emojis is because it can shave bytes off of files and it is easy to understand.
As far as I know this is not a security thing.
I'm trying to implement my own little reader view app (an app that would do the same thing as reader-mode on safari), and there are a few things I find asking myself:
Is there a technical term for this feature (reader-view doesn't really cut it)?
Is there a standard that websites are supposed to follow in order to indicate the content they would like to have in their reader views
Is there an open-source set of HTML parsing rules to pull the "readable" content from a website?
Is the effort to implement such a thing simply too big for a single person in a few weeks and if so should I opt for services such as Instaparser?
I believe the original to be implemented by arc90, and they called it readability. You can check out their page here.
It's been ported to many different languages over time, so you could take a look at the different implementations to learn more about it, how it's done etc.
Python readability
JReadability
JavaScript
Ruby
This is just a small sample here, there's many more examples if you would like to find more.
Edit: Oops, after some more Googling I found this question with an answer that explains it very well.
I am curious that why there are some websites that chrome can translate automatically to the chosen language while it cannot to others?
I have tried using several extensions but the website is not getting translated. Is there any possible explanation for this?
Having a link to a particular case will help greatly but I can try to answer the general case. There are two main reasons:
Google does not know how to translate from a particular language
The webpage has confused the CLD.
The first is obvious so let me explain the second one. Chrome detects the webpage language using heuristics. They are in a package known as the compact language detector (CLD). It is compact, which means it trades database size for accuracy; the non-compact, more accurate version being many times larger.
So assuming that the page in question is in a popular language, there must be something in the sequence of text characters that is confusing the CLD which either does not know what language the page is using or it thinks is using a language that Google does not know how to translate.
I have a website in English which I want to translate to Spanish. Is there a fast way of doing this with keeping the HTML tags(<strong>,<span>) in tact?. I know I can just copy the parsed TEXT into a translator but this will take a lot of time.
I plan to rewrite all the needed parts to provide a high translation quality, but I dont want to translate it from scratch.
Disclaimer: A similar question has been asked but it didnt get a real answer
UPDATE
Old question
I posted a new question because the previous question is different in the sense that the owner poster accepted that translation by scratch as an answer and that is what I'm trying to avoid.
Google Translate would be fine, but it doesnt support translating batch files, and API is paid. I'm looking for a free solution.
I want to translate a website with ~100 articles so I was looking for a automatic translation as a way to begin.
I hope the question is clearer now.
Thanks for reading.
Not sure what translator you are using but Google Translate will not translate anything that are inside these < >
<strong>I'm Awesome</strong> will be <strong> Soy impresionante </ strong>
Google translate is your best bet, it won't translate any HTML within < and > tags
Some time ago I had accidentally found website which was apparently product of Microsoft research on fact extraction from the web, more specifically from the Wikipedia. Right now I would like to have better look into how it works, but the problem is i can't find it, maybe somebody knows what I am talking about and could give me directions or link to it?
Thank you!
There is a hint of it here:
http://research.microsoft.com/en-us/collaboration/global/scholar2005.aspx#Kaisser
Looks like MSR where sponsoring a Phd student who was working on it a few years ago. There is a link to his website, might be worth giving him a shout.