# https://www.linuxjournal.com/content/downloading-entire-web-site-wget
# https://linuxreviews.org/Wget:_download_whole_or_parts_of_websites_with_ease
# https://www.webhostface.com/kb/knowledgebase/examples-using-wget/
# "You can replicate the HTML content of a website with the –mirror option (or -m for short)
#       wget -m http://domain.com"
# https://www.linuxquestions.org/questions/linux-server-73/wget-how-to-download-more-than-one-file-at-once-instead-of-file-after-file-704693/
wget.mirror.cmd=wget -Q10m -m %%BASE_URL%%

# for downloading a single file
wget.file.cmd=wget %%FILE_URL%%

# Arbitrary cutoff values for WETProcessor.java
WETprocessor.min.content.length=100
WETprocessor.min.line.count=2
WETprocessor.min.content.length.wrapped.line=500
WETprocessor.min.spaces.per.wrapped.line=10

# Arbitrary cutoff values for WETProcessor.java
# for determining whether a WET record has sufficient and sensible content
WETprocessor.max.word.length=15
WETprocessor.min.num.words=20
WETprocessor.max.words.camelcase=10


mongodb.user=anupama
mongodb.pwd=chang3m3
# default mongodb port is 27017. Don't change the port unless you really have configured
# your mongodb server to listen at some other port
mongodb.port=27017
mongodb.host=mongodb.cms.waikato.ac.nz
#mongodb.dbname=ateacrawldata
mongodb.dbname=anupama