Auto Translate

aaron's picture
public
aaron - Fri, 2007-06-01 20:58

I just started the Auto Translate project. This module, when complete, will add a button to the node translation tab that will fill in a textarea with a (very rough) translation from a third party automatic translator, such as Babelfish or Google. The module may also be configured to translate the content within a textarea.

Obviously, this is not intended to create automatic content for a site -- the current state of the art of automatic translation services are poor at best. Instead, it should hopefully make the translation process just a little easier for editors, particularly if they're not fluent in another language.

I'm not sure of any legal issues that might arise from the use of the module -- as I can't find an API for this sort of thing, I'm writing the module to scrape the other sites. To try to stay within a reasonable interpretation of Fair Use, I'm only configuring the module to translate at the node submission level, and including a logo and/or link to the service(s) provided on the node submission page.

I'm using the Video CCK as a base for the heavy stuff, it is intended to be easily expandable to include future third party translators.

Any comments or suggestions as I proceed with writing the module? Is this duplicating work that someone else has already created? Any valid reason I shouldn't do it this way? Anyone want to help?

I should have something working in a week or three, or in a couple of days if I can find a sponsor for the work. (I'm not really looking for sponsors by the way, but if you're interested and need something really, really quickly, contact me or Advomatic.)

Thanks,
Aaron Winborn


Google Translate module, TOS

Gábor Hojtsy's picture
Gábor Hojtsy - Sat, 2007-06-02 08:40

Well, there is a http://drupal.org/project/gtrans module, but it works on the page display level, not on node submission.

As far as legal stuff goes, I have looked into this sort of feature before and went to the Google Translate FAQ page (http://www.google.com/intl/en/help/faq_translation.html) which links to the generic Google Terms of Service page (http://www.google.com/accounts/TOS). It says the following:

5.3 You agree not to access (or attempt to access) any of the Services by any means other than through the interface that is provided by Google, unless you have been specifically allowed to do so in a separate agreement with Google. You specifically agree not to access (or attempt to access) any of the Services through any automated means (including use of scripts or web crawlers) and shall ensure that you comply with the instructions set out in any robots.txt file present on the Services.

I thought that this is pretty clearly against automated scraping of Google Translate pages. I also went on to search for web services for automated translations, and found some which were similar scrapers of Google Translate and were asked to be taken down by Google, so they are not working anymore.


Yeah, that seems pretty

aaron's picture
aaron - Wed, 2007-06-06 20:51

Yeah, that seems pretty clear. I've found at least one other service that looks like it might work, and will look at the other TOS's. However, at this point, I'll leave it to individual administrators to research the terms of service, as they may wish to contact Google (or other services) directly. If they get the permission, they'll have a working script. I'll make sure to have a handy link to the TOS there, leave the service(s) disabled by default, and make sure the module by itself doesn't abuse a service.

Aaron Winborn
Advomatic, Web Design for Progressive Advocacy, Grassroots Movements, and Really Cool Causes


I have a pre-alpha release

aaron's picture
aaron - Wed, 2007-06-06 20:47

I have a pre-alpha release of the module now from work last night at Auto Translate. It will add buttons to textareas on ALL node submission forms right now. Note that only the 'google' and 'revert' buttons actually do anything, and the 'google' button ONLY translates english to spanish right now. This is just to get a framework in place.

It all uses fancy jquery, which is what I worked on last night mostly -- so when you click a button, the content in the form changes automatically.

Still TODO:
add other services
work on ui
figure out why accented characters are coming up screwy

On the last item, I'm really scratching my head over it. Doesn't matter if I use check_plain or not, the accents don't show up right. Anyone know why '¡Hola, mundo!' would come up as '�Hola, mundo!', even if I'm examining the raw HTML returned by drupal_http_request?

Aaron Winborn
Advomatic, Web Design for Progressive Advocacy, Grassroots Movements, and Really Cool Causes


Any progress with this

arbel - Tue, 2007-10-09 22:13

Any progress with this module?

thanks

Idan