creek 2.5.3 → 2.6.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 0d90975898b7d260d37f6543755f8c2f4d5931d6e387a56c343bce23eee7e3b1
-  data.tar.gz: 0c381e483070a0e883f128e1a03147a5acf60c4a1095b42ef0d3106af7889f8b
+  metadata.gz: c1b4c7fc5370d7342b94d1093ee6c862720df9ba290513719eb726c91beb67d9
+  data.tar.gz: 54000d7369f0b38f19486d6d15355fcda2baf9b848173e5a7d9e50c537626959
 SHA512:
-  metadata.gz: 1f23c318b54b45c2d1732c874e3f6519e7d528175a395b888e3352f302fb1f0b8625cdde965f16540081163282751dd410a712281651b9b3d3b91c9e89bec800
-  data.tar.gz: e601060be6ba35b3bc3ded42865bce8b46fc3ca92feec1439dc2e719f132bb2359d06129f1110dc6676d07f7707364fead716f8f584a5851388a1d183c6e6825
+  metadata.gz: 34c2a023e06a47258b430595db38b3a2eb8b56df7e9b5b41c7b8e6878803e155adff60849b3d6568f6afe7f0ffb38dbd0cf7c333871e09bb7dcc1957e7ef3e33
+  data.tar.gz: bacf97b02bef5b6eeefb6acdd3a74e9af621aeeff2ea8688486c19f2e9860ada8dd7b65499d4b1c79eef981b92b4ed4b966c6c4390b3f38ec575a66055316c24

data/.github/workflows/test.yml

@@ -0,0 +1,35 @@
+# This workflow uses actions that are not certified by GitHub.
+# They are provided by a third-party and are governed by
+# separate terms of service, privacy policy, and support
+# documentation.
+# This workflow will download a prebuilt Ruby version, install dependencies and run tests with Rake
+# For more information see: https://github.com/marketplace/actions/setup-ruby-jruby-and-truffleruby
+
+name: Ruby
+
+on:
+  push:
+    branches: [ "*" ]
+  pull_request:
+    branches: [ "*" ]
+
+permissions:
+  contents: read
+
+jobs:
+  test:
+
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        ruby-version: ['2.6', '2.7', '3.0', '3.1', '3.2']
+
+    steps:
+    - uses: actions/checkout@v3
+    - name: Set up Ruby
+      uses: ruby/setup-ruby@v1
+      with:
+        ruby-version: ${{ matrix.ruby-version }}
+        bundler-cache: true # runs 'bundle install' and caches installed gems automatically
+    - name: Run tests
+      run: bundle exec rake

data/README.md

@@ -1,5 +1,5 @@
 [![version](https://badge.fury.io/rb/creek.svg)](https://badge.fury.io/rb/creek)
-[![downloads](https://ruby-gem-downloads-badge.herokuapp.com/creek?type=total&total_label=downloads)](https://ruby-gem-downloads-badge.herokuapp.com/creek?type=total&total_label=downloads)
+[![downloads](https://img.shields.io/gem/dt/creek)](https://rubygems.org/gems/creek)
 
 # Creek - Stream parser for large Excel (xlsx and xlsm) files.
 

data/creek.gemspec

@@ -20,7 +20,7 @@ Gem::Specification.new do |spec|
 
   spec.required_ruby_version = '>= 2.0.0'
 
-  spec.add_development_dependency "bundler", "~> 2.1.2"
+  spec.add_development_dependency "bundler"
   spec.add_development_dependency "rake"
   spec.add_development_dependency 'rspec', '~> 3.6.0'
   spec.add_development_dependency 'pry-byebug'

data/lib/creek/book.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 require 'zip/filesystem'
 require 'nokogiri'
 require 'date'

data/lib/creek/drawing.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 require 'pathname'
 
 module Creek
@@ -66,7 +68,7 @@ module Creek
     # Drawing xml contains relationships ID's and coordinates (row, col).
     # Drawing relationships xml contains images' locations.
     def load_drawings_and_rels
-      @drawings = parse_xml(@drawing_filepath).css('xdr|twoCellAnchor')
+      @drawings = parse_xml(@drawing_filepath).css('xdr|twoCellAnchor', 'xdr|oneCellAnchor' )
       drawing_rels_filepath = expand_to_rels_path(@drawing_filepath)
       @drawings_rels = parse_xml(drawing_rels_filepath).css('Relationships')
     end
@@ -82,7 +84,7 @@ module Creek
       col_from_selector = 'xdr:from/xdr:col'.freeze
       col_to_selector = 'xdr:to/xdr:col'.freeze
 
-      @drawings.xpath('//xdr:twoCellAnchor').each do |drawing|
+      @drawings.xpath('//xdr:twoCellAnchor', '//xdr:oneCellAnchor').each do |drawing|
         # embed = drawing.xpath(image_selector).first.attributes['embed']
         temp = drawing.xpath(image_selector).first
         embed = temp.attributes['embed'] if temp
@@ -93,12 +95,17 @@ module Creek
 
         row_from = drawing.xpath(row_from_selector).text.to_i
         col_from = drawing.xpath(col_from_selector).text.to_i
-        row_to = drawing.xpath(row_to_selector).text.to_i
-        col_to = drawing.xpath(col_to_selector).text.to_i
 
-        (col_from..col_to).each do |col|
-          (row_from..row_to).each do |row|
-            @images_pathnames[[row, col]].push(path)
+        if drawing.name == 'oneCellAnchor'
+          @images_pathnames[[row_from , col_from ]].push(path)
+        else
+          row_to = drawing.xpath(row_to_selector).text.to_i
+          col_to = drawing.xpath(col_to_selector).text.to_i
+
+          (col_from..col_to).each do |col|
+            (row_from..row_to).each do |row|
+              @images_pathnames[[row, col]].push(path)
+            end
           end
         end
       end

data/lib/creek/shared_strings.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 require 'zip/filesystem'
 require 'nokogiri'
 
@@ -5,6 +7,8 @@ module Creek
 
   class Creek::SharedStrings
 
+    SPREADSHEETML_URI = 'http://schemas.openxmlformats.org/spreadsheetml/2006/main'
+
     attr_reader :book, :dictionary
 
     def initialize book
@@ -27,9 +31,17 @@ module Creek
 
     def self.parse_shared_string_from_document(xml)
       dictionary = Hash.new
-
-      xml.css('si').each_with_index do |si, idx|
-        text_nodes = si.css('>t, r t')
+      namespace = xml.namespaces.detect{|_key, uri| uri == SPREADSHEETML_URI }
+      prefix = if namespace && namespace[0].start_with?('xmlns:')
+                 namespace[0].delete_prefix('xmlns:') + '|'
+               else
+                 ''
+               end
+      node_selector = "#{prefix}si"
+      text_selector = ">#{prefix}t, #{prefix}r #{prefix}t"
+
+      xml.css(node_selector).each_with_index do |si, idx|
+        text_nodes = si.css(text_selector)
         if text_nodes.count == 1 # plain text node
           dictionary[idx] = Creek::Styles::Converter.unescape_string(text_nodes.first.content)
         else # rich text nodes with text fragments

data/lib/creek/sheet.rb

@@ -8,6 +8,7 @@ module Creek
     include Creek::Utils
 
     HEADERS_ROW_NUMBER = '1'
+    SPREADSHEETML_URI = 'http://schemas.openxmlformats.org/spreadsheetml/2006/main'
 
     attr_accessor :with_headers
     attr_reader :book,
@@ -97,13 +98,30 @@ module Creek
           cell_type = nil
           cell_style_idx = nil
           @book.files.file.open(path) do |xml|
+            prefix = ''
+            name_row = "row"
+            name_c = "c"
+            name_v = "v"
+            name_t = "t"
             Nokogiri::XML::Reader.from_io(xml).each do |node|
-              if node.name == 'row' && node.node_type == opener
+              if prefix.empty? && node.namespaces.any?
+                namespace = node.namespaces.detect{|_key, uri| uri == SPREADSHEETML_URI }
+                prefix = if namespace && namespace[0].start_with?('xmlns:')
+                           namespace[0].delete_prefix('xmlns:') + ':'
+                         else
+                           ''
+                         end
+                name_row = "#{prefix}row"
+                name_c = "#{prefix}c"
+                name_v = "#{prefix}v"
+                name_t = "#{prefix}t"
+              end
+              if node.name == name_row && node.node_type == opener
                 row = node.attributes
                 row['cells'] = {}
                 cells = {}
                 y << (include_meta_data ? row : cells) if node.self_closing?
-              elsif node.name == 'row' && node.node_type == closer
+              elsif node.name == name_row && node.node_type == closer
                 processed_cells = fill_in_empty_cells(cells, row['r'], cell, use_simple_rows_format)
                 @headers = processed_cells if with_headers && row['r'] == HEADERS_ROW_NUMBER
 
@@ -117,11 +135,11 @@ module Creek
 
                 row['cells'] = processed_cells
                 y << (include_meta_data ? row : processed_cells)
-              elsif node.name == 'c' && node.node_type == opener
+              elsif node.name == name_c && node.node_type == opener
                 cell_type      = node.attributes['t']
                 cell_style_idx = node.attributes['s']
                 cell           = node.attributes['r']
-              elsif %w[v t].include?(node.name) && node.node_type == opener
+              elsif (node.name == name_v || node.name == name_t) && node.node_type == opener
                 unless cell.nil?
                   node.read
                   cells[cell] = convert(node.value, cell_type, cell_style_idx)

data/lib/creek/styles/converter.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 require 'set'
 
 module Creek

data/lib/creek/styles/style_types.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 # https://github.com/hmcgowan/roo/blob/master/lib/roo/excelx.rb
 # https://github.com/woahdae/simple_xlsx_reader/blob/master/lib/simple_xlsx_reader.rb#L231
 module Creek

data/lib/creek/styles.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 module Creek
   class Styles
     attr_accessor :book

data/lib/creek/utils.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 module Creek
   module Utils
     def expand_to_rels_path(filepath)

data/lib/creek/version.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 module Creek
-  VERSION = "2.5.3"
+  VERSION = "2.6.3"
 end

data/spec/drawing_spec.rb

@@ -3,9 +3,11 @@ require './spec/spec_helper'
 describe 'drawing' do
   let(:book) { Creek::Book.new('spec/fixtures/sample-with-images.xlsx') }
   let(:book_no_images) { Creek::Book.new('spec/fixtures/sample.xlsx') }
+  let(:book_with_one_cell_anchored_images) { Creek::Book.new('spec/fixtures/sample-with-one-cell-anchored-images.xlsx') }
   let(:drawingfile) { 'xl/drawings/drawing1.xml' }
   let(:drawing) { Creek::Drawing.new(book, drawingfile) }
   let(:drawing_without_images) { Creek::Drawing.new(book_no_images, drawingfile) }
+  let(:drawing_with_one_cell_anchored_images) { Creek::Drawing.new(book_with_one_cell_anchored_images, drawingfile) }
 
   describe '#has_images?' do
     it 'has' do
@@ -48,5 +50,21 @@ describe 'drawing' do
         expect(image1).to eq(image2)
       end
     end
+
+    context 'when one cell anchored images in cell' do
+      it 'returns image for anchored cell' do
+        image = drawing_with_one_cell_anchored_images.images_at('A2')[0]
+        expect(image.class).to eq(Pathname)
+        expect(image.exist?).to eq(true)
+      end
+
+      it 'returns nil for non-anchored cell' do
+        image = drawing_with_one_cell_anchored_images.images_at('A3')[0]
+        # Image can be seen present on cell A4 in `sample-with-one-cell-anchored-images.xlsx`
+        image_at_non_anchored_cell = drawing_with_one_cell_anchored_images.images_at('A4')
+        expect(image.class).to eq(Pathname)
+        expect(image_at_non_anchored_cell).to eq(nil)
+      end
+    end
   end
 end

data/spec/fixtures/sst_namespaced.xml

@@ -0,0 +1,91 @@
+<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<sst xmlns:x="http://schemas.openxmlformats.org/spreadsheetml/2006/main" count="6" uniqueCount="5">
+    <x:si>
+        <x:t>Cell A1</x:t>
+    </x:si>
+    <x:si>
+        <x:t>Cell B1</x:t>
+    </x:si>
+    <x:si>
+        <x:t>My Cell</x:t>
+    </x:si>
+    <x:si>
+        <x:r>
+            <x:rPr>
+                <x:sz val="11"/>
+                <x:color rgb="FFFF0000"/>
+                <x:rFont val="Calibri"/>
+                <x:family val="2"/>
+                <x:scheme val="minor"/>
+            </x:rPr>
+            <x:t>Cell</x:t>
+        </x:r>
+        <x:r>
+            <x:rPr>
+                <x:sz val="11"/>
+                <x:color theme="1"/>
+                <x:rFont val="Calibri"/>
+                <x:family val="2"/>
+                <x:scheme val="minor"/>
+            </x:rPr>
+            <x:t xml:space="preserve"> </x:t>
+        </x:r>
+        <x:r>
+            <x:rPr>
+                <x:b/>
+                <x:sz val="11"/>
+                <x:color theme="1"/>
+                <x:rFont val="Calibri"/>
+                <x:family val="2"/>
+                <x:scheme val="minor"/>
+            </x:rPr>
+            <x:t>A2</x:t>
+        </x:r>
+    </x:si>
+    <x:si>
+        <x:r>
+            <x:rPr>
+                <x:sz val="11"/>
+                <x:color rgb="FF00B0F0"/>
+                <x:rFont val="Calibri"/>
+                <x:family val="2"/>
+                <x:scheme val="minor"/>
+            </x:rPr>
+            <x:t>Cell</x:t>
+        </x:r>
+        <x:r>
+            <x:rPr>
+                <x:sz val="11"/>
+                <x:color theme="1"/>
+                <x:rFont val="Calibri"/>
+                <x:family val="2"/>
+                <x:scheme val="minor"/>
+            </x:rPr>
+            <x:t xml:space="preserve"> </x:t>
+        </x:r>
+        <x:r>
+            <x:rPr>
+                <x:i/>
+                <x:sz val="11"/>
+                <x:color theme="1"/>
+                <x:rFont val="Calibri"/>
+                <x:family val="2"/>
+                <x:scheme val="minor"/>
+            </x:rPr>
+            <x:t>B2</x:t>
+        </x:r>
+    </x:si>
+    <x:si>
+        <x:t>Cell with_x000D_escaped_x000D_characters</x:t>
+    </x:si>
+    <x:si>
+        <x:t>吉田兼好</x:t>
+        <x:rPh sb="0" eb="2">
+            <x:t xml:space="preserve">ヨシダ </x:t>
+        </x:rPh>
+        <x:rPh sb="2" eb="4">
+            <x:t xml:space="preserve">ケンコウ </x:t>
+        </x:rPh>
+        <x:phoneticPr fontId="1"/>
+    </x:si>
+</x:sst>

data/spec/shared_string_spec.rb

@@ -17,4 +17,21 @@ describe 'shared strings' do
     expect(dictionary[6]).to eq('吉田兼好')
   end
 
-end
+  context 'when the nodes are namespaced' do
+    it 'parses the dictionary correctly' do
+      shared_strings_xml_file = File.open('spec/fixtures/sst_namespaced.xml')
+      doc = Nokogiri::XML(shared_strings_xml_file)
+      dictionary = Creek::SharedStrings.parse_shared_string_from_document(doc)
+
+      expect(dictionary.keys.size).to eq(7)
+      expect(dictionary[0]).to eq('Cell A1')
+      expect(dictionary[1]).to eq('Cell B1')
+      expect(dictionary[2]).to eq('My Cell')
+      expect(dictionary[3]).to eq('Cell A2')
+      expect(dictionary[4]).to eq('Cell B2')
+      expect(dictionary[5]).to eq("Cell with\rescaped\rcharacters")
+      expect(dictionary[6]).to eq('吉田兼好')
+    end
+
+  end
+end

data/spec/sheet_spec.rb

@@ -65,6 +65,20 @@ describe 'sheet' do
       end
     end
 
+    context 'when one cell anchored images in cell' do
+      let(:book_with_one_cell_anchored_images) { Creek::Book.new('spec/fixtures/sample-with-one-cell-anchored-images.xlsx') }
+      let(:sheet_with_one_cell_anchored_images) { Creek::Sheet.new(book_with_one_cell_anchored_images, 'Sheet 1', 1, '', '', '1', sheetfile) }
+      let(:rows) { sheet_with_one_cell_anchored_images.with_images.rows.map { |r| r } }
+
+      it 'returns image for anchored cell' do
+        expect(load_cell(rows, 'A2').size).to eq(1)
+      end
+
+      it 'returns nil for non-anchored cell' do
+        expect(load_cell(rows, 'A4')).to eq(nil)
+      end
+    end
+
     context 'with excel without images' do
       let(:book_no_images) { Creek::Book.new('spec/fixtures/sample.xlsx') }
       let(:sheet_no_images) { Creek::Sheet.new(book_no_images, 'Sheet 1', 1, '', '', '1', sheetfile) }
@@ -74,6 +88,16 @@ describe 'sheet' do
         expect(load_cell(rows, 'A10')).to eq(0.15)
       end
     end
+
+    context 'when nodes are namespaced' do
+      let(:namespaced_book) { Creek::Book.new('spec/fixtures/sample_namespaced.xlsx') }
+      let(:namespaced_sheet) { Creek::Sheet.new(namespaced_book, 'Sheet 1', 1, '', '', '1', sheetfile) }
+
+      it 'parses rows correctly' do
+        rows = namespaced_sheet.rows.map { |r| r }
+        expect(load_cell(rows, 'A10')).to eq(0.15)
+      end
+    end
   end
 
   describe '#images_at' do
@@ -130,5 +154,37 @@ describe 'sheet' do
         expect(row['HeaderC']).to eq 'value3'
       end
     end
+
+    context 'when nodes are namespaced' do
+      let(:namespaced_book) { Creek::Book.new('spec/fixtures/sample-with-headers_namespaced.xlsx') }
+      let(:sheet) { Creek::Sheet.new(namespaced_book, 'Sheet 1', 1, '', '', '1', sheetfile) }
+
+      it 'returns values by letters' do
+        expect(subject['A']).to eq 'value1'
+        expect(subject['B']).to eq 'value2'
+      end
+
+      context 'when enable with_headers property' do
+        before { sheet.with_headers = true }
+
+        it 'returns values by headers name' do
+          expect(subject['HeaderA']).to eq 'value1'
+          expect(subject['HeaderB']).to eq 'value2'
+          expect(subject['HeaderC']).to eq 'value3'
+        end
+
+        it 'returns headers correctly when called multiple times' do
+          row = sheet.simple_rows.to_a[1]
+          expect(row['HeaderA']).to eq 'value1'
+          expect(row['HeaderB']).to eq 'value2'
+          expect(row['HeaderC']).to eq 'value3'
+
+          row = sheet.simple_rows.to_a[1]
+          expect(row['HeaderA']).to eq 'value1'
+          expect(row['HeaderB']).to eq 'value2'
+          expect(row['HeaderC']).to eq 'value3'
+        end
+      end
+    end
   end
 end

metadata

@@ -1,29 +1,29 @@
 --- !ruby/object:Gem::Specification
 name: creek
 version: !ruby/object:Gem::Version
-  version: 2.5.3
+  version: 2.6.3
 platform: ruby
 authors:
 - pythonicrubyist
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-07-12 00:00:00.000000000 Z
+date: 2023-05-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: 2.1.2
+        version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: 2.1.2
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
@@ -103,6 +103,7 @@ extensions: []
 extra_rdoc_files: []
 files:
 - ".DS_Store"
+- ".github/workflows/test.yml"
 - ".gitignore"
 - Gemfile
 - LICENSE.txt
@@ -128,13 +129,17 @@ files:
 - spec/fixtures/large_numbers.xlsx
 - spec/fixtures/sample-as-zip.zip
 - spec/fixtures/sample-with-headers.xlsx
+- spec/fixtures/sample-with-headers_namespaced.xlsx
 - spec/fixtures/sample-with-images.xlsx
+- spec/fixtures/sample-with-one-cell-anchored-images.xlsx
 - spec/fixtures/sample.xlsx
 - spec/fixtures/sample_dates.xlsx
+- spec/fixtures/sample_namespaced.xlsx
 - spec/fixtures/sheets/sample_dates.xlsx
 - spec/fixtures/sheets/sheet1.xml
 - spec/fixtures/sheets/single_data_programme.xlsx
 - spec/fixtures/sst.xml
+- spec/fixtures/sst_namespaced.xml
 - spec/fixtures/styles/first.xml
 - spec/fixtures/temp_string_io_file_path_with_no_extension
 - spec/shared_string_spec.rb
@@ -162,7 +167,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.0.3
+rubygems_version: 3.0.3.1
 signing_key: 
 specification_version: 4
 summary: A Ruby gem for parsing large Excel(xlsx and xlsm) files.
@@ -175,13 +180,17 @@ test_files:
 - spec/fixtures/large_numbers.xlsx
 - spec/fixtures/sample-as-zip.zip
 - spec/fixtures/sample-with-headers.xlsx
+- spec/fixtures/sample-with-headers_namespaced.xlsx
 - spec/fixtures/sample-with-images.xlsx
+- spec/fixtures/sample-with-one-cell-anchored-images.xlsx
 - spec/fixtures/sample.xlsx
 - spec/fixtures/sample_dates.xlsx
+- spec/fixtures/sample_namespaced.xlsx
 - spec/fixtures/sheets/sample_dates.xlsx
 - spec/fixtures/sheets/sheet1.xml
 - spec/fixtures/sheets/single_data_programme.xlsx
 - spec/fixtures/sst.xml
+- spec/fixtures/sst_namespaced.xml
 - spec/fixtures/styles/first.xml
 - spec/fixtures/temp_string_io_file_path_with_no_extension
 - spec/shared_string_spec.rb