# dbs/lucegene/blastxml.properties
# d.gilbert, jan 05

LIB_NAME=blastxml
title = BLAST output XML

# dmel-dvir-tblastn.xml
DATA_ROOT=web/data/blast/
INDEX_PATH=indices/lucene/blastxml/

MIME_TYPE=text/xml

## --------- search/report keys -------------------

searchfield=all

outfields=docid,queryid,Hsp_identity,Hsp_align-len,Hsp_evalue,Hsp_bit-score,url
storefields=docid,queryid,Hit_id,Hsp_identity,Hsp_evalue,Hsp_bit-score,Hsp_align-len,BlastOutput_db,url
# BlastOutput_query-def,

docurl=lookup.jsp?id=
batch.forward=lookup.jsp

## for xslt to produce the result page tables
# resultxsl = conf/blast_result.xsl
# resultspage = resultxsl.jsp

# header.native=<?xml version="1.0"?>\n<BlastOutputs>
header.native=<?xml version="1.0"?>\n\
   <?xml-stylesheet href="conf/blast_report.xsl" type="text/xsl"?>\n\
   <BlastOutputs>\n

footer.native=\n</BlastOutputs>

batchformat = text/xml
batchformats = text/xml, text/plain, text/csv,  text/tsv
title.text/xml=BLAST XML
nativeformat = text/xml 

## difference in header. footer. is for native file data versus lucegene index fields
header.xml=<?xml version="1.0"?>\n<LuceGene>
footer.xml=\n</LuceGene>

# ---- indexing values -----------------------

# locate data with regex file, folder patterns
regex_folder=
regex_file=^.*blast.*\.xml$

regex_skipfile=
regex_skipfolder=.*
# (tmp|.*\.old)


INDEX_CLASS=org.eugenes.index.LuceneXmlIndexer

## append existing index or create new
INDEX_APPEND=false
## index <tag> names as values (as well as field names)?
INDEX_TAGS=false
## index <tag attr="val"> values ?
INDEX_ATTRIBUTES=false
## use fieldname xpath; full top.middle.last field name?
INDEX_XPATH=false

## INDEX_LEVEL=0 means index main records one level below xml <top> tag 
INDEX_LEVEL=0
INDEX_BLANKS=false

## IndexWriter opts
## merge=10 is default; 4 == less mem usage ; 2 minimum  
merge_factor=6
## max_field_length is max # terms/field
max_field_length=1000000
MAX_FIELDS=50000

# to create "contents" field of all text
indexall=false


## field indexing parameters
## sumfields list needs to match field.xxx common summary fields
sumfields=docid,queryid
# special summary fields -- replace w/ fieldalias.TAG=newtag
# field.docid=ID
# field.docclass=CLA
# field.title=RETE
# field.summary=GeneSummary.SUMMARY.text


## need some recoders; want to index each <Hit> as a document
# <BlastOutput.BlastOutput_iterations.Iteration.Iteration_hits.Hit ...
# not good; need to add mainrec fields to subrec docs - query, ...
 
subrecdoc.Hit=Hit

## default -  Text or UnStored = index but dont store text
fieldtype=UnStored

fieldtype.Hit_def=Text
tokenizer.Hit_def=fbacode$deflineTokens
fieldrecoder.Hit_def=fbacode$defline_Recoder
# ^ defline_Recoder makes ID field
fieldalias.ID=docid
tokenizer.docid=fbacode$deflineTokens
tokenfilter.docid=fbacode$DebugEndOfRecordFilter

fieldtype.Hit_id=Text
fieldtype.Hsp_bit-score=Text
fieldtype.Hsp_align-len=Text
fieldtype.Hsp_evalue=Text
fieldtype.Hsp_identity=Text
fieldtype.BlastOutput_db=Text

fieldtype.BlastOutput_query-def=Text
fieldrecoder.BlastOutput_query-def=fbacode$mainrecsave_Recoder
##field.queryid=BlastOutput_query-def
fieldalias.ID2=queryid
fieldtype.queryid=Text
tokenizer.queryid=fbacode$deflineTokens

fieldtype.Hsp_qseq=skip
fieldtype.Hsp_hseq=skip
fieldtype.Hsp_midline=skip

# Summary.text
# fieldtype.text=UnStored

## these must always be stored; Text or UnIndexed
## Keyword is problem as Search wants to lc() all 1st, Keyword is casefull
fieldtype.docid=Text
fieldtype.docclass=Text
##fieldtype.doclink=Text
fieldtype.url=UnIndexed
fieldtype.modified=Keyword
fieldtype.title=UnIndexed
fieldtype.summary=UnIndexed

analyzer=org.eugenes.index.BiodataAnalyzer2

# all field defaults
tokenizer=fbacode$LowerWordTokenizer
tokenfilter=fbacode$DebugFilter


fieldrecoder=fbacode$FBID_Recoder

#----- fix blast.xml output ----

# set xout=data2/xml/dmel-dvir-tblastn.xml

# perlfix= \
# #!/usr/bin/perl
# # slimblastxml.pl -- cut verbosity down for NCBI BLAST -m 7 xml output
# print "<?xml version=\"1.0\"?>\n<BlastOutputs>\n";  
# while(<>){
#    next if (m,^<\?xml, || m,^<\!DOCTYPE,);
#    if (m,<(BlastOutput_reference|BlastOutput_param|Iteration_stat),) { $skipto= $1; }
#    if ($skipto) {
#     $skipto='' if (m,</$skipto>,);
#     next; 
#     }
#   print;
# }
# print "</BlastOutputs>\n"; # trailer