eme.properties

# Use post tokenizer which understands EEBO/TCP encodings.

posttokenizer.class=EEBOPostTokenizer

# Use simple XML text inputter.  For very large texts, it is better (but much slower)
# to use the DiskBasedXMLTextInputter which shuttles portions of the text to and
# from disk as needed.

#textinputter.class=DiskBasedXMLTextInputter
textinputter.class=SimpleXMLTextInputter

# Use simple name standardizer.

namestandardizer.class=EEBOSimpleNameStandardizer

# Use Early Modern English lexicons and word lists.

partofspeechtagger.transition_matrix=data/emetransmat.mat
lexicon.word_lexicon=data/emelexicon.lex
lexicon.suffix_lexicon=data/emesuffixlexicon.lex
spelling.spelling_pairs=data/ememergedspellingpairs.tab
abbreviations.abbreviations_url=data/emeabbreviations.txt

# Use latin word list.

wordlists.use_latin_word_list=true

# Fix selected split words in XML input.

xml.fix_split_words=true

# Specify the Java style regular expression pattern(s) for matching and 
# replacement of split words.  Currently these patterns only fix up
# split reflexive pronouns and some leading periods.  
# The zzzzsw is used as a marker tag internally. 

xml.fix_split_words.match1 = \\s\u0304
xml.fix_split_words.replace1 = \u0304
xml.fix_split_words.match2 = (?iu)(here|her|him|hir|hym|It|it|me|mi|my|not|one|oure|our|owne|own|their|theim|theym|theyr|them|the\u0304|thy|us|youre|your)\\s(selfes|selues|selves|selfe|selfs|self)
xml.fix_split_words.replace2 = $1<zzzzsw>$2</zzzzsw>
xml.fix_split_words.match3 = (?iu)(\\s)(ofthe)(\\s)
xml.fix_split_words.replace3 = <choice><sic> ofthe </sic><corr> of the </corr></choice>
xml.fix_split_words.match4 = the\\.(\\s<hi>[0123456789.]+</hi>)
xml.fix_split_words.replace4 =<choice><sic>the.</sic><corr>the</corr></choice>$1
xml.fix_split_words.match5 = the\\.(\\s[0123456789.]+)
xml.fix_split_words.replace5 =<choice><sic>the.</sic><corr>the</corr></choice>$1