com.google.enterprise.adaptor.prebuilt
Class PrebuiltTransforms

java.lang.Object
  com.google.enterprise.adaptor.prebuilt.PrebuiltTransforms

public class PrebuiltTransforms
extends Object

Common transforms that you would expect to have available.

copyMetadata

public static DocumentTransform copyMetadata(Map<String,String> config)

Returns a transform that copies metadata values from one key to another. The "overwrite" key can be set to "true" to cause the destination to be replaced; otherwise the destination key is supplemented.

Copies are defined by pairs of "X.from" and "X.to" configuration entries (where X is an integer). The value for each is a metadata key. Copies are applied in the increasing order of the integers.

Example configuration:

overwrite=false
3.from=colour
3.to=color
5.from=author
5.to=contributors

moveMetadata

public static DocumentTransform moveMetadata(Map<String,String> config)

Returns a transform that moves metadata values from one key to another. This method returns a transform that behaves identically to copyMetadata(java.util.Map), except that the source keys are removed. If the source key has no metadata values then the destination is left as-is.

deleteMetadata

public static DocumentTransform deleteMetadata(Map<String,String> config)

Returns a transform that deletes metadata keys. The keys to be deleted are defined by "keyX" configuration entries (where X is an integer).

Example configuration:

key2=sensitive
key4=unhelpful

replaceMetadata

public static DocumentTransform replaceMetadata(Map<String,String> config)

Returns a transform that preforms string replacements within metadata values. The keys to have replacements done on their values are defined by "keyX" configuration entries (where X is an integer). The "overwrite" configuration key can be set to "false" to cause the original string to be left intact; otherwise the original string is replaced.

The needle to be found is configured via a "string" or "pattern" configuration key. string's value is treated as a literal string to be found whereas pattern's value is treated as a regular expression. The replacement is defined by "replacement" and is interpreted as a literal string if "string" was provided and a regular expression replacement if "pattern" was provided.

Example configuration:

overwrite=false
key1=favorite
key5=least favorite
pattern=(Java|C|Perl)
replacement=$1 (but it should be x86 assembler)