WorldServer - How to control repaired matches for legacy filters and for File Types filters

What are the techniques and limitations to control repaired matches for WorldServer legacy filters and for Studio filters/File Types?

When using the Studio-aligned scoping mode (recommended), the repair functionality is achieved in the tm.properties file by the Recognizers and the Auto-Substitution settings that are enabled by default. The scoping mode is set in the Project Type configuration. These settings are:

### Studio Enabled/Disabled Auto-Substitution Options 

# Enables/disables auto-substitution for date placeables. 
#studio_enable_auto_substitution_for_dates=true 

# Enables/disables auto-substitution for time placeables. 
#studio_enable_auto_substitution_for_times=true 

# Enables/disables auto-substitution for number placeables. 
#studio_enable_auto_substitution_for_numbers=true 

# Enables/disables auto-substitution for measurement placeables. 
#studio_enable_auto_substitution_for_measurements=true

### Studio Enabled/Disabled Recognizers

# Enables/disables recognition for date placeables.
#studio_enable_recognize_dates=true

# Enables/disables recognition for time placeables.
#studio_enable_recognize_times=true

# Enables/disables recognition for number placeables.
#studio_enable_recognize_numbers=true

# Enables/disables recognition for measurement placeables.
#studio_enable_recognize_measurements=true

# Enables/disables recognition for acronym and URL placeables.
# Note that neither placeable type is auto-substitutable.
#studio_enable_recognize_acronyms=true

When using the WorldServer 9.x scoping mode, The condition for the Autorepair settings to be enabled when using the WorldServer 9.x scoping mode is that this property is set false in tm.properties:

prevent_all_segment_repairs=false

If you want to disable Auto repair, set this property to:

prevent_all_segment_repairs=true

and restart the Idiom Service. Note: this property has no impact on the Studio Autorepair functionality. 

The WorldServer 9.x scoping mode is a scoping mode used when using Idiom Legacy filters instead of (Studio) File Type. Note that Idiom Legacy Filters are no longer supported or developed, although they are still enabled in WorldServer. The scoping mode is set in the Project Type configuration. 

These are the repair settings in tm.properties when using the WorldServer 9.x scoping mode:

1- Number repair: This repair attempts to handle number differences between the segment text and the TM match text.
 

# This repair attempts to handle number differences between the
# segment text and the TM match text.
# Requires prevent_all_segment_repairs=false in order to run.
##number_transform.target_languages=all
#number_transform.enabled=true

2- Number deletion repair:  These repairs attempt to remove non-required numerical data

# This repair attempts to delete extra numbers from the TM match
# text.
# Requires prevent_all_segment_repairs=false in order to run.
#number_deletion.target_languages=all
#number_deletion.enabled=true

3- Placeholder repair: These repairs attempt to repair placeholder differences 

# The repairs attempts to address formatting differences between
# the segment text and the corresponding TM match. It 
# allows a penalty to be assessed against placeholder repairs that
# WorldServer does not consider safe.
# Requires prevent_all_segment_repairs=false in order to run.
# placeholder.target_languages=all
#placeholder.enabled=true
#placeholder.unsafe_repair_penalty=

Unsafe repair penalty attribute
The unsafe_repair_penalty attribute (default value is 0.0) – allows the user to set a final scoring penalty to be
assessed if the placeholder repair that has been applied is not deemed by the system as safe. Safe placeholder repair
support is not an explicit repair option, although the system supports identifying these types of repair conditions. This
option effectively allows the user to prevent unsafe repairs from being scored up to 100% matches. Set the value between
0.0 and 1.0. A practical value is 0.01, which translates into a 1% final score penalty.

4- Word transform repair: These repairs attempt to repair untranslated content 

# The current implementation seeks to transform untranslated text in
# the source and target text of the match to match that of the source
# segment. This repair should not be used with languages that share
# common words.
# Requires prevent_all_segment_repairs=false in order to run.
#word_transform.target_languages=all
#word_transform.enabled=false

5- Word deletion repair: These repairs attempt to delete extra words 

# This repair attempts to delete extra words from the TM match
# text.
# Requires prevent_all_segment_repairs=false in order to run.
#word_deletion.target_languages=all
#word_deletion.enabled=true

6- Punctuation repair: These repairs attempt correct boundary punctuation differences 

# This repairs attempts to align the outer punctuation of the TM
# match to that of the source segment.
# Requires prevent_all_segment_repairs=false in order to run.
#punctuation.target_languages=all
#punctuation.enabled=true

7- Capitalization repair: These repairs attempt enforce consistency between replacement text and original text 

# This repair attempts to ensure capitalization consistency of
# replaced text to that of the original text. This works in conjunction
# with the word transform repair.
# Requires prevent_all_segment_repairs=false in order to run.
#capitalization.target_languages=all
#capitalization.enabled=true