Sha256: 22f975d3111623dac79ddb8c676f1e9ae1848c406401b2d0ececc78d198de25e
Contents?: true
Size: 1.59 KB
Versions: 133
Compression:
Stored size: 1.59 KB
Contents
#!/usr/bin/env ruby
require 'rbbt-util'
require 'rbbt/util/simpleopt'
$0 = "rbbt #{$previous_commands*""} #{ File.basename(__FILE__) }" if $previous_commands
options = SOPT.setup <<EOF
Assemble the TSV from a table extracted from a PDF
$ rbbt tsv assemble_pdf_table file.txt
When extracting tables from PDF they are often laid out one column at a time, divided by pages.
This command takes a file with the following structure:
1 A few lines containing table headers, one per line
2 A group of lines containing the values for the first column of the first page, ending in an empty line
3 More groups of lines corresponding to other columns
4 Repetitions of 2 and 3 for more pages
This script will take care of matching the columns read with the headers specified
-h--help Help
EOF
SOPT.usage if options[:help]
file = ARGV.shift
file = STDIN if file == '-' or file.nil?
txt = Misc.fixutf8(TSV.get_stream(file).read)
header, _sep, rest = txt.strip.partition("\n\n")
fields = header.split("\n")
num_columns = fields.length
columns = {}
num_columns.times do |i|
columns[i] = []
end
lines = rest.split("\n")
while lines and lines.any?
first_block = lines[0..lines.index("")-1]
block_size = first_block.length
lines = lines[block_size+1..-1]
columns[0] << first_block
(1..num_columns-1).each do |pos|
block = lines[0..block_size-1]
lines = lines[block_size+1..-1]
columns[pos] << block
end
end
full_columns = []
num_columns.times do |i|
column = columns[i]
full_columns << column.flatten
end
puts "#" << fields * "\t"
Misc.zip_fields(full_columns).zip do |values|
puts values * "\t"
end
Version data entries
133 entries across 133 versions & 1 rubygems