in match-assoc-panel (panel to associate a match resource to a task) a check-box is added next to the "new analysis" button/check-box (see below): count internal fuzzies
an internal fuzzy is a fuzzy that will be generated, when starting to translate a task. So if segment 1 is translated, the translation may lead to a new fuzzy-match for segment 2. Ect.
In other words: we save the source of the untranslated segment to a temporary TM, so that it is assumed that this source text is already translated
To get correct analysis data one segment after another must be send to the the temp TM. Sending all segments in a step before would not reflect the translators real work.
if the check-box is checked, the internal fuzzies can be calculated by
temporary copying the match-resource
analysing against the copy
after anaylsing a segment, pseudo-translate it
for pre-translation (see below) do not use pseudo-translated segments. To achieve this, pseudo-translate the segments with a unique identifier, that will never occur in target segments, such as "translate5-unique-id[TASKGUID_OF_THE_TASK]". And when receiving a 100%-match with such a target: Do not pre-translate the segment. And do not count them, since they are repetitions (see above)
Sub-task