Easyling Release Notes - 2014 July

With the heat of July bearing down on us, our office becomes more of a refuge than a workplace - without our new A/C systems, development would quickly grind to a halt due to heat strokes. But since we’re running those A/Cs on full blast, we can afford to keep rolling out new features and abilities for Easyling as fast as we can. Some of these new features include the ability to completely delete source segments via a regular expression, a new, advanced Discovery system, and with this, a new interface that provides more detailed statistics, as well as a more detailed CSV export, useful for tracking changes. See the full list after the jump!

Maybe the most important feature rolled out this month is the Advanced Discovery system. This is an upgrade to our existing Discovery, which has now been given the ability to give word and character (for Asian languages) counts for the sites visited. Furthermore, it provides detailed statistics that also include leverage metrics: five levels of confidence in repetitions, from 98% to 102%, each with its own criteria.

  1. 102%: Strong contextual matches: every segment in the block level element (~paragraph) is a 101% match, where all the tags are identical.
  2. 101%: Contextual matches: both tags in the segment, and contexts (segments immediately before and after) match.
  3. 100%: Regular matches: the segment is repeated exactly, including all tags.
  4. 99%: Strong fuzzy matches: tags from the ends are stripped out, words lowercased, numbers ignored.
  5. 98%: Weak fuzzy matches: all tags are stripped out (may have to be adjusted manually afterwards!), words lowercased, numbers ignored.

These metrics are their most useful when combined with a translation memory (either offline, or uploaded into Easyling), because it gives you a rather accurate picture of the amount of work the TM can do for you and the amount you have to check over. And all this information is available right from the start.

[caption id=“attachment_1335” align=“aligncenter” width=“987”] Discovery (and Scan) now displays detailed repetition statistics across five different confidence levels, useful for leveraging TMs[/caption]

Another new feature is the ability to delete source segments from individual pages or the entire project. This feature is implemented in the Workbench released last month, when searching in pages views or the All Entries view, reached by clicking the search icon (the magnifying glass). Be advised, however, that this is a destructive operation, that requires forethought before using! In the pages view, you can search for given terms, and use the filters to further restrict the operation; while the All Entries view only supports deleting via a regular expression, but will delete everything that matches the expression (commonly called a “regex(p)”). Once the operation completes, the matching entries will be deleted from our database, and will need to be scanned  again if they do need to be translated.

[caption id=“attachment_1337” align=“aligncenter” width=“1920”] The Workbench can now delete entries from the database. The deleted entries will match the regular expression entered into the search field.[/caption]

Lastly, we have enhanced our secondary export file type, the CSV (short for “comma-separated values”, a universal file type that can be interpreted by Excel to produce spreadsheets on the topic) by adding new fields detailing when each entry was scanned and added to the database. This is useful for continuously changing sites, where the dates can be used to verify/support SLAs.