1. Use the adaptAlSetToBilCorpus.pl routine of the 
Lingua::AlignmentSet toolkit. The manual contains some 
documentation and code example:

perl adaptAlSetToBilCorpus.pl -man

The script passed to standard output the list of minimum 
editions necessary to pass from a corpus to the other:

pair 21 source: substitute 'II.' at positions { 1 } by 'II .'
pair 21 source: substitute '))' at positions { 42 } by ') )'
pair 21 target: substitute 'II.' at positions { 1 } by 'II .'
pair 21 target: substitute '))' at positions { 23 } by ') )'
pair 21 target: substitute '91/494/' at positions { 31 } by '91/494 /'

When doing insertions or replacing a word by various, many-to-many 
links are introduced. In this example, in the links output file, 
the two tokens 'II .' would be aligned as a group to the two 
tokens 'II .' because in the original file, the token 'II.' was linked 
to 'II.'. You may want to edit these links to link 'II' to 'II' 
and '.' to '.'.

2. In order to edit these links:

2.1 Change the adapted reference to alpaco format. Example:
perl chFormat_alSet-1.1.pl -ist adaptedref.engspa.naacl -is adaptedref.eng.iso.naacl -it adaptedref.spa.iso.naacl -if naacl -os adaptedref.eng.iso -ot adaptedref.spa.iso -ost A1 -of blinker
ln -sf adaptedref.eng.iso EN.sample.1
ln -sf adaptedref.spa.iso FR.sample.1

2.2 Run alpaco_sp.pl to edit the necessary links, and change again to a more handy
format.