251 lines
8.5 KiB
Ruby
Executable File
251 lines
8.5 KiB
Ruby
Executable File
#!/bin/env ruby
|
|
############################################################################
|
|
#Word sorter for Pat's dissertation #
|
|
#More documentation TBD #
|
|
#Current Usage: #
|
|
# Scans a file for groups of words and counts the totals #
|
|
# Accepts the following options: #
|
|
# --file - the name of the text to sort #
|
|
# --type - iat or pn #
|
|
# --bin-file - the name of the bin csv file #
|
|
# #
|
|
# The input file is human readable, easy to edit. #
|
|
# The output file is single-line JSON, use jq to format for reading. #
|
|
# #
|
|
#To Do: #
|
|
# * add frequency #
|
|
############################################################################
|
|
require 'json'
|
|
require 'pp'
|
|
require 'csv'
|
|
require 'optparse'
|
|
|
|
#bin_counter expects:
|
|
# bin - a hash with strings to search for
|
|
# test - the document text to search through
|
|
#This method returns a hash containing the strings and their frequency
|
|
def bin_counter (bin, text)
|
|
ret = Hash.new
|
|
bin.each do |word|
|
|
ret[word.to_sym] = text.scan(word).count.to_i
|
|
end
|
|
return ret
|
|
end
|
|
|
|
#count_total expects:
|
|
# bin - a hash created by bin_counter
|
|
#This method returns a hash with the total count of all words in a bin
|
|
def count_total (bin)
|
|
count = 0
|
|
bin[:words].each_key do |word|
|
|
count += bin[:words][word].to_i
|
|
end
|
|
return count
|
|
end
|
|
|
|
#write_output expects:
|
|
# output - a hash containing all of our output
|
|
#This method converts the output hash to JSON and writes it to "output.json"
|
|
def write_output_json (output, filename)
|
|
outfile = File.open(filename,'w')
|
|
outfile.write(output.to_json)
|
|
outfile.close
|
|
end
|
|
|
|
def write_output_csv (output, filename)
|
|
CSV.open(filename, 'wb') do |csv|
|
|
csv << ["bin", "words", "total"]
|
|
output.delete(:filename)
|
|
output.delete(:total_words)
|
|
output.each_key do |key|
|
|
line = []
|
|
line.push(key)
|
|
output[key].each_key do |sub_key|
|
|
line.push(output[key][sub_key])
|
|
end
|
|
csv << line
|
|
end
|
|
end
|
|
end
|
|
|
|
#strip_text expects:
|
|
# text - the text we're working on
|
|
# start - the starting string to search for
|
|
# fin - the ending string to search for
|
|
#This method strips out all test before "start" and after "fin"
|
|
def strip_text (text, start, fin)
|
|
text.lines(start,chomp: true)[1].lines(fin,chomp: true)[0]
|
|
end
|
|
|
|
#split_text expects:
|
|
# text - the text we're working on
|
|
# start - the starting string to search for
|
|
# fin the ending string to search for
|
|
# This method returns everything between start and fin and handles multiple sections
|
|
# Example:
|
|
# For the following text:
|
|
# > asdfasdf
|
|
# > AAAAA
|
|
# > testing
|
|
# > abcdefg
|
|
# > BBBBB
|
|
# > nothing
|
|
# > something
|
|
# > AAAAA
|
|
# > moo said the cow.
|
|
# > cluck said the chicken.
|
|
# > BBBBB
|
|
# > bark said the dog
|
|
# Running split_text(text, 'AAAAA', 'BBBBB') would return:
|
|
# > testing
|
|
# > abcdefg
|
|
# > moo said the cow.
|
|
# > cluck said the chicken.
|
|
def split_text (text, start, fin)
|
|
split1 = text.lines(start, chomp: true)[1..-1]
|
|
split2 = []
|
|
ret = ""
|
|
split1.each do |section|
|
|
split2 += section.lines(fin, chomp: true)
|
|
end
|
|
split2.each do |section|
|
|
ret += section
|
|
end
|
|
return ret
|
|
end
|
|
|
|
#process_file expects:
|
|
# file_name - the name of the file to process
|
|
# binfile - the name of the bin file (csv) to use
|
|
# type - which type of file are we processing, must be 'pn' or 'iat'
|
|
#
|
|
#This method is the meat and potatos. Preforms the text stripping, word counting, and creates output files.
|
|
def process_file (file_name, binfile, type)
|
|
#text = split_text(text, 'Narrative:', 'Signatures:') if type == 'pn'
|
|
csv = CSV.read(binfile)
|
|
text = File.read(file_name)
|
|
bins = Hash.new #This hash stores the bins
|
|
csv.each { |bin| bins[bin[0]] = bin[1..].compact } #turn the csv array into a hash, remove nils
|
|
outfile = file_name
|
|
outfile.slice!('.txt')
|
|
if type == 'iat'
|
|
text = strip_text(text, 'PLOVEINTAKE', 'PLOVECLOSING') if type == 'iat'
|
|
output = Hash.new #Creating the output storage object
|
|
#puts outfile
|
|
output[:filename] = outfile
|
|
output[:total_words] = text.split.size
|
|
bins.each_key do |bin_number|
|
|
key = bin_number.to_sym
|
|
output[key] = Hash.new
|
|
output[key][:words] = bin_counter(bins[bin_number], text)
|
|
output[key][:total] = count_total(output[key])
|
|
output[key][:frequency] = output[key][:total].to_f / output[:total_words].to_f
|
|
end
|
|
write_output_json(output,outfile + '-out.json')
|
|
write_output_csv(output,outfile + '-out.csv')
|
|
elsif type == 'pn'
|
|
sections = text.lines("Date and time:", chomp: true) #sections is an arrary of each date section from the text
|
|
sections.delete_at(0) #we can ignore the first chunk of text
|
|
sections.each do |chunk|
|
|
timestamp = chunk.lines.delete_if {|line| line == "\r\n"}[0] #pulling out the timestamp
|
|
timestamp.tr!('/','-').tr!(':','').tr!(' ','_') #remove slashes and colons from timestamp, replaces spaces with underscores
|
|
timestamp.strip!
|
|
output = Hash.new #Creating the output storage object
|
|
outfile = file_name + '_' + timestamp
|
|
outfile.slice!('.txt')
|
|
text = strip_text(chunk, 'Narrative:', 'Signatures:')
|
|
output[:filename] = outfile
|
|
output[:total_words] = text.split.size
|
|
bins.each_key do |bin_number|
|
|
key = bin_number.to_sym
|
|
output[key] = Hash.new
|
|
output[key][:words] = bin_counter(bins[bin_number], text)
|
|
output[key][:total] = count_total(output[key])
|
|
output[key][:frequency] = output[key][:total].to_f / output[:total_words].to_f
|
|
end
|
|
write_output_json(output,outfile + '-out.json')
|
|
write_output_csv(output,outfile + '-out.csv')
|
|
end
|
|
end
|
|
end
|
|
|
|
#process_dir expects:
|
|
# dir_name - a direcroty containing text files to process
|
|
# binfile - the name of the bin file
|
|
# type - which type of file are we processing, must be 'pn' or 'iat'
|
|
#
|
|
#This method will process all .txt files in the supplied directory
|
|
def process_dir(dir_name, binfile, type)
|
|
threads = []
|
|
Dir.glob(dir_name + '*.txt') do |file_name|
|
|
threads << Thread.new do
|
|
puts "Processing " + file_name
|
|
process_file(file_name, binfile, type)
|
|
end
|
|
end
|
|
threads.each { |thr| thr.join }
|
|
end
|
|
|
|
def generate_master_output(dir_name, binfile)
|
|
file=File.open(binfile,"r")
|
|
bin_count = file.readlines.size
|
|
file.close
|
|
bin_header = []
|
|
bin_count.times do |num|
|
|
num += 1
|
|
words_head = "Bin " + num.to_s + " words"
|
|
total_head = "Bin " + num.to_s + " total"
|
|
freq_head = "Bin " + num.to_s + " frequency"
|
|
bin_header.push(words_head)
|
|
bin_header.push(total_head)
|
|
bin_header.push(freq_head)
|
|
end
|
|
CSV.open('master.csv', 'wb') do |csv|
|
|
header = ["File", "Total Words" ] + bin_header
|
|
csv << header
|
|
Dir.glob(dir_name + '*.json') do |file_name|
|
|
puts 'Getting data from: ' + file_name
|
|
csv_row = []
|
|
json_file = File.read(file_name)
|
|
data_hash = JSON.parse(json_file)
|
|
csv_row.push(data_hash["filename"])
|
|
csv_row.push(data_hash["total_words"])
|
|
data_hash.delete("filename")
|
|
data_hash.delete("total_words")
|
|
data_hash.each_key do |key|
|
|
csv_row.push(data_hash[key]["words"])
|
|
csv_row.push(data_hash[key]["total"])
|
|
csv_row.push(data_hash[key]["frequency"])
|
|
end
|
|
csv << csv_row
|
|
end
|
|
end
|
|
end
|
|
|
|
options = Hash.new
|
|
OptionParser.new do |opts|
|
|
opts.banner = 'sorter.rb --options'
|
|
opts.on("-f", "--file file", "Name of the file to process") do |file|
|
|
options[:file] = file
|
|
end
|
|
opts.on("-t", "--type type", "Type of file. Must be \"iat\" or \"pn\"") do |type|
|
|
options[:type] = type
|
|
end
|
|
opts.on("-b", "--bin-file binfile", "Name of the bin file") do |binfile|
|
|
options[:binfile] = binfile
|
|
end
|
|
opts.on("-d", "--directory dir", "Directory containing text files to process") do |dir|
|
|
options[:dir] = dir
|
|
end
|
|
end.parse!
|
|
|
|
if options[:file] && options[:dir]
|
|
puts "Invalid options, you must either a file or a directoy of files."
|
|
elsif options[:file]
|
|
process_file(options[:file], options[:binfile], options[:type])
|
|
elsif options[:dir]
|
|
process_dir(options[:dir], options[:binfile], options[:type])
|
|
generate_master_output(options[:dir], options[:binfile])
|
|
end
|
|
|