Improve html handling from Drupal plugin
Description
Activity
Sundeep Anand 31 May 2016 at 10:20
Please have a look.
https://github.com/zanata/tmgmt_zanata/pull/8
Former user 26 May 2016 at 00:15
best to go with that option if it is the only one that works.
I would like to know more about the 500 server error, since it means there is probably a bug in the server that we need to fix.
Sundeep Anand 25 May 2016 at 11:42
while downloading translated file, it fails causing a 500 server error with drupal_http_request whereas it worked well with curl. how do you feel, is it okay? plus some code refactor may be required in push pull process.
Former user 17 May 2016 at 23:35
it would be best if it can work without saving a file on disk. Could the hash difference be caused by using a different encoding for the string when calculating the hash?
If it is too complicated or impossible to send the contents as-is, I think it is fine to use file_save_data as long as we clean up the files reliably.
Sundeep Anand 17 May 2016 at 11:14
Upload API is now working for me. I think we need a file object for multipart/form-data content. (tried with putting contents as is but hash differs and push fails). how do you think, should we look at drupal's file_save_data?
Details
Details
Assignee
Sundeep AnandReporter
Carlos MunozTested Version/s
Components
Priority
High
The Drupal plugin currently breaks down the html by "paragraphs" using new line characters as a guide. This results in a lot of html 'noise' being sent to Zanata making it difficult for translators to operate on actual translatable content. One alternative would be grab the complete html content and send it to Zanata using the file upload endpoint as html content, taking advantage of the processing used by Zanata's html processor.