hocr_turtletext 0.1.2 → 0.1.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 23ed7d4b8c32b028f82675d42d9a8b2625692276
-  data.tar.gz: c0e778642e9c0fe39a72116797ea39223cbf17bb
+  metadata.gz: 39a62a3746924d1fee867013bc7a03f361d805bf
+  data.tar.gz: ee01353babbe3f3d4fcb78b84f40fcd1d2fea4e7
 SHA512:
-  metadata.gz: ec2e4c2fc85a3b241afff470cb820c0b907301be78cd4f2d28e67c60509337bf75ba8c7cd1a17c592fb2e3234c96803b0c3db5247d50cbbefd2445fbb6068149
-  data.tar.gz: 53dd315bc5c91df9d2da50a18385f64b243bd852c4d77b7cf88d6dc0431f54e3f6a441be71ddf6906d324d76f9e61b13bb551a35fda0a017bcda97a304d97bda
+  metadata.gz: c62c8d2a3c6a1c7aa7a31185d92e5a9903471f345ec72aeb56b56a2c95953478aa713011a6ebf8483c20d26b940cbbdd0d248524c37c1aa467d79f7a827641d0
+  data.tar.gz: f2942de656083773a4132fc2a84ead1cb5b670b4cb4b5dc5ce1026eb7d1a472b0035d1e5afbfa635236b860aa91f620468851fde5bf0179347a5fd4f9acd1a2f

data/README.md

@@ -146,12 +146,6 @@ It returns a Hash of x/y co-ordinates that is the bottom-left corner of the text
 ```
 Note: in the case of multiple matches, only the first match is returned.
 
-## Development
-
-After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests. You can also run `bin/console` for an interactive prompt that will allow you to experiment.
-
-To install this gem onto your local machine, run `bundle exec rake install`. To release a new version, update the version number in `version.rb`, and then run `bundle exec rake release`, which will create a git tag for the version, push git commits and tags, and push the `.gem` file to [rubygems.org](https://rubygems.org).
-
 ## Contributing
 
 - Check issue tracker if someone is working on what you plan to work on

data/lib/hocr_turtletext/reader.rb

@@ -10,12 +10,14 @@ class HocrTurtletext::Reader
 
   def content
     hocr_content = File.read(@hocr_path)
-    lines = precise_content(hocr_content)
-    pos_hash = to_pos_hash(lines)
-    fuzzed_y(pos_hash)
+    html = Nokogiri::HTML(hocr_content)
+    pos_info_words = extract_words_from_html(html)
+    pos_hash = to_pos_hash pos_info_words
+    fuzzed_y = fuzzed_y(pos_hash)
+    concat_words_in_lines(fuzzed_y)
   end
 
-  def text_in_region(xmin,xmax,ymin,ymax,inclusive=false)
+  def text_in_region(xmin, xmax, ymin, ymax, inclusive=false)
     return [] unless xmin && xmax && ymin && ymax
     text_map = content
     box = []
@@ -37,12 +39,12 @@ class HocrTurtletext::Reader
   def text_position(text)
     item = if text.class <= Regexp
              content.map do |k,v|
-               if x = v.reduce(nil){|memo,vv|  memo = (vv[1] =~ text) ? vv[0] : memo  }
+               if x = v.reduce(nil){ |memo,vv|  memo = (vv[1] =~ text) ? vv[0] : memo }
                  [k,x]
                end
              end
            else
-             content.map {|k,v| if x = v.rassoc(text) ; [k,x] ; end }
+             content.map { |k,v| if x = v.rassoc(text) ; [k,x] ; end }
            end
     item = item.compact.flatten
     unless item.empty?
@@ -51,7 +53,7 @@ class HocrTurtletext::Reader
   end
 
   def bounding_box(&block)
-    HocrTurtletext::Textangle.new(self,&block)
+    HocrTurtletext::Textangle.new(self, &block)
   end
 
   private
@@ -64,10 +66,37 @@ class HocrTurtletext::Reader
     @options[:y_precision] ||= 3
   end
 
+  def extract_words_from_html(html)
+    pos_info_words = []
+
+    html.css('span.ocrx_word, span.ocr_word')
+        .reject { |word| word.text.strip.empty? }
+        .each do |word|
+      word_attributes = word.attributes['title'].value.to_s
+                            .delete(';').split(' ')
+      pos_info_word = word_info(word, word_attributes)
+      pos_info_words.push pos_info_word
+    end
+    pos_info_words
+  end
+
+  def to_pos_hash(lines)
+    lines.sort_by { |line| line[:y_start] }
+
+    pos_hash = {}
+    lines.each do |run|
+      pos_hash[run[:y_start]] ||= {}
+      pos_hash[run[:y_start]][run[:x_start]] = run
+    end
+    pos_hash
+  end
+
   def fuzzed_y(input)
     output = []
     input.keys.sort.each do |precise_y|
-      matching_y = output.map(&:first).select { |new_y| (new_y - precise_y).abs < y_precision }.first || precise_y
+      matching_y = output.map(&:first)
+                       .select { |new_y| (new_y - precise_y).abs < y_precision }
+                       .first || precise_y
       y_index = output.index{ |y| y.first == matching_y }
       new_row_content = input[precise_y].to_a
       if y_index
@@ -81,77 +110,34 @@ class HocrTurtletext::Reader
     output
   end
 
-  def precise_content(hocr_content)
-    html = Nokogiri::HTML(hocr_content)
-    lines = []
-    html.css('span.ocr_line').map do |line|
-      chunks = chunks_from_processed_ocr_line(line)
-      lines.concat(chunks)
-    end
-    lines
-  end
-
-  def chunks_from_processed_ocr_line(ocr_line)
-    pos_info_line = add_positional_info_to_line(ocr_line)
-    sorted_pos_info_line = sort_words_in_line(pos_info_line)
-    concat_words_in_line(sorted_pos_info_line)
-  end
-
-  def add_positional_info_to_line(ocr_line)
-    ocr_line.css('span.ocrx_word, span.ocr_word')
-            .reject { |word| word.text.strip.empty? }
-            .map do |word|
-      word_attributes = word.attributes['title'].value.to_s
-                            .delete(';').split(' ')
-      info(word, word_attributes)
-    end
-  end
-
-  def sort_words_in_line(pos_info_line)
-    # sort word by x value, concat if x2.x_start - x1.x_end < some_x_threshold
-    pos_info_line.sort_by { |word| word[:x_start] }
-    pos_info_line.slice_when do |x, y|
-      y[:x_start] - x[:x_end] > x_whitespace_threshold
-    end.to_a
-  end
-
-  def concat_words_in_line(sorted_pos_info_line)
-    chunks = []
-    # merge all words in each chunk
-    sorted_pos_info_line.each do |chunk|
-      sentence = nil
-      chunk.each do |word|
-        if sentence.nil?
-          sentence = word
+  def concat_words_in_lines(fuzzed_y)
+    fuzzed_y.map do |line|
+      x_pos_keyed_words = line[1]
+      concatenated_words = []
+      x_pos_keyed_words.each do |x_pos_keyed_word|
+        word_hash = x_pos_keyed_word[1]
+        if concatenated_words.empty? ||
+            word_hash[:x_start] - concatenated_words.last[:x_end] > x_whitespace_threshold
+          concatenated_words.push word_hash
         else
-          sentence[:word] = "#{sentence[:word]} #{word[:word]}"
-          sentence[:x_end] = word[:x_end]
+          concatenated_words.last[:word] = "#{concatenated_words.last[:word]} #{word_hash[:word]}"
+          concatenated_words.last[:x_end] = word_hash[:x_end]
         end
       end
-      chunks.push sentence
-    end
-    chunks
-  end
-
-  def to_pos_hash(lines)
-    lines.sort_by { |line| line[:y_start] }
-
-    pos_hash = {}
-    lines.each do |run|
-      pos_hash[run[:y_start]] ||= {}
-      pos_hash[run[:y_start]][run[:x_start]] ||= ''
-      pos_hash[run[:y_start]][run[:x_start]] << run[:word]
+      line[1] = concatenated_words.map! do |word_hash|
+        [word_hash[:x_start], word_hash[:word]]
+      end
+      line
     end
-    pos_hash
   end
 
-  def info(word, data)
+  def word_info(word, data)
     {
-        word: word.text,
-        x_start: data[1].to_i,
-        y_start: data[2].to_i,
-        x_end: data[3].to_i,
-        y_end: data[4].to_i
+      word: word.text,
+      x_start: data[1].to_i,
+      y_start: data[2].to_i,
+      x_end: data[3].to_i,
+      y_end: data[4].to_i
     }
   end
 end

data/lib/hocr_turtletext/version.rb

@@ -1,3 +1,3 @@
 module HocrTurtletext
-  VERSION = '0.1.2'.freeze
+  VERSION = '0.1.3'.freeze
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: hocr_turtletext
 version: !ruby/object:Gem::Version
-  version: 0.1.2
+  version: 0.1.3
 platform: ruby
 authors:
 - Sue Zheng Hao
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2020-01-24 00:00:00.000000000 Z
+date: 2020-01-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler