iguvium 0.8.2 → 0.8.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: b453936d07d4aea3f6eb682ba90f6c6642f83520b14462b8c46a254c6dc538f7
-  data.tar.gz: 27fc73360191edbb1185a073d6cf4b644455af682764ff8f2cdbaeb9a5b1b973
+  metadata.gz: a70226f57667319ee056c909d177e04e816721aee9706f8ba6351e592cc4c209
+  data.tar.gz: dc3a3ccff9aee45b27224656a9a7e6b2a4d758ee0220697b145bf6ae3918e747
 SHA512:
-  metadata.gz: de9c944ba0504a9c8bc4badcfe2bee974cc5244111999f0109939009fdbb9c167962a51a17e181763138b04909d01c1f2ccfff8a3317ebab1cd9bf1b6e26e9e6
-  data.tar.gz: 7edeb83865b934edd01d509d0024241f6dacf727394d2b72077ff9c0e02189fdae7f1b81fcca961602ab8d2bcf42fa3fb9657bf627ce1ccdd2b2649d01ee97bf
+  metadata.gz: 59aa03ad81d1c9c085fb0dd3b66a48e88d44a4a8b516e83f0160a47de8bde32ae58769efa2f700276779695e9286c4eaaa6a76db432adeb1df4ac5f0be8d90fb
+  data.tar.gz: 58e8f45ca66aa7841207c50597d400a618e38aafa091608fd4a4db7cb7406c0473d884320e08d4cc67b26a92ba5d7c64e6aa8ef5ab1bcf4f00d9ba730eb2529a

data/README.md

@@ -89,6 +89,7 @@ iguvium filename.pdf [options]
     -p, --pages     page numbers, comma-separated, no spaces
     -i, --images    use pictures in pdf (usually a bad idea)
     -n, --newlines  keep newlines
+    --phrases       keep phrases unsplit, could fix some merged cells
     -t, --text      extract full page text instead of tables
     --verbose       verbose output
 ```

data/exe/iguvium

@@ -9,6 +9,7 @@ opts = Slop.parse { |o|
   o.array '-p', '--pages', 'page numbers, comma-separated, no spaces'
   o.bool '-i', '--images', 'use pictures in pdf (usually a bad idea)'
   o.bool '-n', '--newlines', 'keep newlines'
+  o.bool '--phrases', 'keep phrases unsplit, could fix some merged cells'
   o.bool '-t', '--text', 'extract full page text instead of tables'
   o.bool '--verbose', 'verbose output'
   o.on '--version', 'print the version' do
@@ -54,7 +55,7 @@ page_numbers.each do |number|
   tables.empty? ? puts('no tables found') : puts
 
   tables.each_with_index do |table, i|
-    csv = table.to_a(newlines: opts[:newlines]).map(&:to_csv).join
+    csv = table.to_a(newlines: opts[:newlines], phrases: opts[:phrases]).map(&:to_csv).join
     next if csv.empty?
     csv_file = "#{path.gsub(/\.pdf$/, '')}_page_#{number + 1}_table_#{i}.csv"
     puts "Saving #{File.expand_path csv_file}"

data/lib/iguvium/table.rb

@@ -24,12 +24,22 @@ module Iguvium
     # Sometimes you may need to keep them; in this case use `newlines: true` option.
     #
     # @param  [Boolean] newlines keep newlines inside table cells, false by default
+    # @param  [Boolean] phrases keep phrases unsplit, false by default.
+    #   Poor man's merged cells workaround. Could break some tables, could fix some.
+    #
     # @return [Array] 2D array of strings (content of table's cells)
     #
-    def to_a(newlines: false)
+    def to_a(newlines: false, phrases: false)
       grid[:rows]
         .reverse
-        .map { |row| grid[:columns].map { |column| render(chars_inside(column, row), newlines: newlines) } }
+        .map { |row|
+        grid[:columns].map do |column|
+          render(
+            phrases ? words_inside(column, row) : chars_inside(column, row),
+            newlines: newlines
+          )
+        end
+      }
     end
 
     private
@@ -49,6 +59,20 @@ module Iguvium
         .select { |character| xrange.cover?(character.x) && yrange.cover?(character.y) }
     end
 
+    def words
+      @words ||=
+        characters
+        .sort
+        .chunk_while { |a, b| a.mergable?(b) }
+        .map { |chunk| chunk.inject(:+) }
+    end
+
+    def words_inside(xrange, yrange)
+      words.select { |character|
+        xrange.cover?(character.x) && yrange.cover?(character.y)
+      }
+    end
+
     def grid
       @grid ||=
         {

data/lib/iguvium/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Iguvium
-  VERSION = '0.8.2'
+  VERSION = '0.8.3'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: iguvium
 version: !ruby/object:Gem::Version
-  version: 0.8.2
+  version: 0.8.3
 platform: ruby
 authors:
 - Dima Ermilov
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2018-11-21 00:00:00.000000000 Z
+date: 2018-11-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: pdf-reader