-
Improvement
-
Resolution: Unresolved
-
None
-
None
-
High
-
Emptyshow more show less
Problem
With the new feature resname context matches the whole matchrate system needs to be renewed.
Solution
Introduce new system.
matchrate | conditions met |
---|---|
100% | same source and target text ( + optional same resname*) |
101% | same source and target text + repeated within task (repetition)** |
102% | same source and target text + same file name |
103% | same source and target text from term collection |
104% | same source and target text + same above and below segments (context) |
105% | same source and target text + same above and below segments (context) + same file name |
106% | same source and target text + same resname + same file name |
- same resname without same file name is without any context value and therefore considered like a simple 100% match
- * for a more consistent logic switch repetitions with same filename match
Provide a migration script for all pricing presets and all existing analysis
Discuss the change for the xml report. This is the current state:
This file was generated with translate5.
- there are no file elements added, since translate5 does not support a file specific analysis right now
- for the settings element the following values are set:
reportInternalFuzzyLeverage="AS SET IN Translate5"
reportLockedSegmentsSeparately="no"
reportCrossFileRepetitions="yes"
minimumMatchScore="LOWEST FUZZY RANGE AS USED IN Translate5 analysis"
searchMode="bestWins"
all other settings are not available and marked as n/a
- In the batchTotal analysis section the following applies:
For the following elements all numeric attributes are always set to "0",
because they currently have no analogon in translate5:
# locked
# inContextExact (Exception: if configured xmlInContextUsage then 101% matches are set as inContextExact)
# repeated (translate5 will only have crossFileRepeated)
# newBaseline (this is specific to SDL MT)
# newLearnings (this is specific to SDL MT)
- the number and definitions of fuzzy elements will reflect the fuzzy ranges as defined in translate5
- all MT matches and matches not listed as fuzzy or better match will always be counted within "new"
- crossFileRepeated are translate5s repetitions (which are represented by 102% matches)
- exact are 100% and 101% and 104%-Matches from translate5 (101% may also be used as inContextExact if configured)
- perfect are 103%-Matches from translate5
- The following attributes will always have the value "0", since translate5 does not support them right now:
# characters="0"
# placeables="0"
# tags="0"
# repairWords="0"
# fullRecallWords="0"
# partialRecallWords="0"
# edits="0"
# adaptiveWords="0"
# baselineWords="0"
Tnis config must be changed reflecting the swap of 101/102% match:
runtimeOptions.plugins.MatchAnalysis.xmlInContextUsage
- mentioned in
-
Page Loading...