ipynbdiff 0.3.4 → 0.3.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d87c49e29a069b5064985b1b1097ac4d02c375161d9fe64df8e70f0155419220
-  data.tar.gz: 97931332962139f7f129a546bb7a7f9b670d779dcb032b27fbb1a3faa6f7dc85
+  metadata.gz: fbeaad80969c974720e75336301dc02822fbf183b278d8e5ddd5dd18c65ddde1
+  data.tar.gz: c186fa7fac873dff429cf3fd24c4a5cb61901670270850fd3d623c363db21182
 SHA512:
-  metadata.gz: 86c4be0fc53e0db2172221a17dd07e0c3f37492ffc699075b0de74dea3265feebee2207ef4e9170ead59be15c42b858918fbd2e56c7ed109cb674fbf7eb7dbbb
-  data.tar.gz: 10fdb124114a73425edc57f5c9fb80fbd546f6b8e7d58929a327f1be268774d8409238630286aa050e41384bf53c951d2b8c67aaeb76835bc3e0f37672775d48
+  metadata.gz: 23e6a0c192d671fdcb394334a89d40eef9a58e39d934d1138e0ba37943c928d035d2f471e868b2af3d7f24a02f77d927b1e72d8cd3dd8df1f6ef9616796a8ab0
+  data.tar.gz: 9632766d4dd4e6e57775d59d7a14d01fb664156bd3c41b1af894481f4bf33f807b54ebad0f664184f606c37a6b8f818326e7238e320ba06135f192753c659f20

data/.gitlab-ci.yml

@@ -0,0 +1,6 @@
+specs:
+  stage: test
+  image: ruby:2.7
+  script:
+    - bundle install
+    - bundle exec rspec

data/Gemfile

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+source "https://rubygems.org"
+
 gem 'diffy', '3.3.0'
 gem 'json', '2.5.1'
 gem 'rspec', '3.10.0'

data/Gemfile.lock

@@ -1,4 +1,5 @@
 GEM
+  remote: https://rubygems.org/
   specs:
     diff-lcs (1.4.4)
     diffy (3.3.0)
@@ -26,4 +27,4 @@ DEPENDENCIES
   rspec (= 3.10.0)
 
 BUNDLED WITH
-   2.2.28
+   2.2.29

data/README.md

@@ -28,15 +28,16 @@ Options:
 
 ```ruby
 @default_transform_options = {
-  preprocess_input: TRUE, # Whether the input should be transformed
+  preprocess_input: true, # Whether the input should be transformed
   write_output_to: nil, # Pass a path to save the output to a file
   format: :text, # These are the formats Diffy accepts https://github.com/samg/diffy
-  sources_are_files: FALSE, # Weather to use the from/to as string or path to a file 
+  sources_are_files: false, # Weather to use the from/to as string or path to a file
+  raise_if_invalid_notebook: false, # Raises an error if the notebooks are invalid, otherwise returns nil
   transform_options: @default_transform_options, # See below for transform options
   diff_opts: {
-    include_diff_info: FALSE # These are passed to Diffy https://github.com/samg/diffy
+    include_diff_info: false # These are passed to Diffy https://github.com/samg/diffy
   }
-} 
+}
 ```
 
 ### Transforming the notebooks
@@ -51,7 +52,7 @@ Options:
 
 ```ruby
 @default_transform_options = {
-    include_metadata: FALSE, # Whether to include or not the notebook metadata (kernel, language, etc)
+    include_metadata: false, # Whether to include or not the notebook metadata (kernel, language, etc)
     cell_decorator: :html # :html is useful to add styling with css, :percent is better for text format
 }
 ```

data/ipynbdiff.gemspec

@@ -2,7 +2,7 @@
 
 Gem::Specification.new do |s|
   s.name        = 'ipynbdiff'
-  s.version     = '0.3.4'
+  s.version     = ENV['LIB_VERSION']
   s.summary     = 'Human Readable diffs for Jupyter Notebooks'
   s.description = 'Better diff for Jupyter Notebooks by first preprocessing them and removing clutter'
   s.authors     = ['Eduardo Bonet']

data/lib/ipynbdiff.rb

@@ -6,48 +6,53 @@ module IpynbDiff
   require 'diffy'
 
   @default_transform_options = {
-    include_metadata: FALSE,
+    include_metadata: false,
     cell_decorator: :html
   }
 
   @default_diff_options = {
-    preprocess_input: TRUE,
+    preprocess_input: true,
     write_output_to: nil,
     format: :text,
-    sources_are_files: FALSE,
+    sources_are_files: false,
+    raise_if_invalid_notebook: false,
     transform_options: @default_transform_options,
     diff_opts: {
-      include_diff_info: FALSE
+      include_diff_info: false
     }
   }.freeze
 
   def self.prepare_input(to_prepare, options)
-    prepared = to_prepare
-    prepared = File.read(prepared) if options[:sources_are_files]
-    prepared = transform(prepared, options[:transform_options]) if options[:preprocess_input]
+    return '' unless to_prepare
 
-    prepared
+    prep = to_prepare
+    prep = File.read(prep) if options[:sources_are_files]
+    prep = transform(prep, raise_errors: true, options: options[:transform_options]) if options[:preprocess_input]
+    prep
   end
 
   def self.diff(
     from_notebook,
     to_notebook,
-    options = {}
+    options = @default_diff_options
   )
     options = @default_diff_options.merge(options)
 
-    from = from_notebook && prepare_input(from_notebook, options) || ''
-    to = to_notebook && prepare_input(to_notebook, options) || ''
+    from = prepare_input(from_notebook, options)
+    to = prepare_input(to_notebook, options)
 
     d = Diffy::Diff.new(from, to, **options[:diff_opts]).to_s(options[:format])
-
     File.write(options[:write_output_to], d) if options[:write_output_to]
-
     d
+  rescue InvalidNotebookError
+    raise if options[:raise_if_invalid_notebook]
   end
 
-  def self.transform(notebook, options)
+  def self.transform(notebook, raise_errors: false, options: @default_transform_options)
     options = @default_transform_options.merge(options)
+
     Transformer.new(**options).transform(notebook)
+  rescue InvalidNotebookError
+    raise if raise_errors
   end
 end

data/lib/output_transformer.rb

@@ -0,0 +1,65 @@
+# frozen_string_literal: true
+
+module IpynbDiff
+
+  # Transforms Jupyter output data into markdown
+  class OutputTransformer
+
+    ORDERED_KEYS = {
+      'execute_result' => %w[image/png image/svg+xml image/jpeg text/markdown text/latex text/plain],
+      'display_data' => %w[image/png image/svg+xml image/jpeg text/markdown text/latex]
+    }.freeze
+
+    def transform(output)
+      case (output_type = output['output_type'])
+      when 'error'
+        transform_error(output['traceback'])
+      when 'execute_result', 'display_data'
+        transform_non_error(ORDERED_KEYS[output_type], output['data'])
+      end
+    end
+
+    def transform_error(traceback)
+      traceback.map do |t|
+        t.split("\n").map do |line|
+          line.gsub(/\[[0-9][0-9;]*m/, '').sub("\u001B", '    ').gsub(/\u001B/, '').rstrip << "\n"
+        end
+      end
+    end
+
+    def transform_non_error(accepted_keys, elements)
+      accepted_keys.map do |key|
+        transform_element(key, elements[key]) if elements.key?(key)
+      end.flatten
+    end
+
+    def transform_element(output_type, output_element)
+      case output_type
+      when 'image/png', 'image/jpeg'
+        transform_image(output_type, output_element)
+      when 'image/svg+xml'
+        transform_svg(output_element)
+      when 'text/markdown', 'text/latex', 'text/plain'
+        transform_text(output_element)
+      end
+    end
+
+    def transform_image(image_type, image_content)
+      ["    ![](data:#{image_type};base64,#{image_content.gsub("\n", '')})", "\n"]
+    end
+
+    def transform_svg(image_content)
+      lines = image_content.is_a?(Array) ? image_content : [image_content]
+
+      single_line = lines.map(&:strip).join('').gsub(/\s+/, ' ')
+
+      ["    ![](data:image/svg+xml;utf8,#{single_line})", "\n"]
+    end
+
+    def transform_text(text_content)
+      lines = text_content.is_a?(Array) ? text_content : [text_content]
+
+      lines.map { |line| "    #{line}" }.append("\n")
+    end
+  end
+end

data/lib/transformer.rb

@@ -1,21 +1,37 @@
 # frozen_string_literal: true
 
 module IpynbDiff
+  class InvalidNotebookError < StandardError
+  end
+
   # Returns a markdown version of the Jupyter Notebook
   class Transformer
     require 'json'
     require 'yaml'
+    require 'output_transformer'
 
     @cell_decorator = :html
-    @include_metadata = TRUE
+    @include_metadata = true
+
 
-    def initialize(include_metadata: TRUE, cell_decorator: :html)
+    def initialize(include_metadata: true, cell_decorator: :html)
       @include_metadata = include_metadata
       @cell_decorator = cell_decorator
+      @output_transformer = OutputTransformer.new
     end
 
-    def transform(notebook)
+    def validate_notebook(notebook)
       notebook_json = JSON.parse(notebook)
+
+      return notebook_json if notebook_json.key?('cells')
+
+      raise InvalidNotebookError
+    rescue JSON::ParserError
+      raise InvalidNotebookError
+    end
+
+    def transform(notebook)
+      notebook_json = validate_notebook(notebook)
       transformed_blocks = notebook_json['cells'].map do |cell|
         decorate_cell(transform_cell(cell, notebook_json), cell)
       end
@@ -26,7 +42,7 @@ module IpynbDiff
 
     def decorate_cell(rows, cell)
       tags = cell['metadata']&.fetch('tags', [])
-      type = cell['cell_type']
+      type = cell['cell_type'] || 'raw'
 
       case @cell_decorator
       when :html
@@ -53,51 +69,22 @@ module IpynbDiff
 
     def transform_code_cell(cell, notebook)
       [
-        %(``` #{notebook['metadata']['kernelspec']['language']}\n),
+        %(``` #{notebook.dig('metadata', 'kernelspec', 'language') || ''}\n),
         *cell['source'],
         "\n```\n",
         *cell['outputs'].map { |output| transform_output(output) }
       ]
     end
 
-    def format_traceback(traceback)
-      traceback.map do |t|
-        t.split("\n").map do |line|
-          line.gsub(/\[[0-9][0-9;]*m/, '').sub("\u001B", '    ').gsub(/\u001B/, '').rstrip << "\n"
-        end
-      end
-    end
-
-    def transform_execute_result(output)
-      output['data']['text/plain'].map { |line| "    #{line}" }.append("\n")
-    end
-
-    def transform_image_result(output)
-      if output['data'].key?('image/png')
-        ["![](data:image/png;base64,#{output['data']['image/png'].gsub("\n", '')})", "\n"]
-      end
-    end
-
-    def transform_error_result(output)
-      format_traceback(output['traceback'])
-    end
-
     def transform_output(output)
-      transformed =
-        case output['output_type']
-        when 'execute_result'
-          transform_execute_result(output)
-        when 'display_data'
-          transform_image_result(output)
-        when 'error'
-          transform_error_result(output)
-        end
+      transformed = @output_transformer.transform(output)
 
       decorate_output(transformed, output).join('') if transformed
     end
 
     def transform_text_cell(cell)
-      cell['source'].append("\n")
+      source = cell['source']
+      (source.is_a?(Array) ? source : [source]).append("\n")
     end
 
     def transform_metadata(notebook_json)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ipynbdiff
 version: !ruby/object:Gem::Version
-  version: 0.3.4
+  version: 0.3.8
 platform: ruby
 authors:
 - Eduardo Bonet
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-10-18 00:00:00.000000000 Z
+date: 2021-11-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: diffy
@@ -116,11 +116,13 @@ extensions: []
 extra_rdoc_files: []
 files:
 - ".gitignore"
+- ".gitlab-ci.yml"
 - Gemfile
 - Gemfile.lock
 - README.md
 - ipynbdiff.gemspec
 - lib/ipynbdiff.rb
+- lib/output_transformer.rb
 - lib/transformer.rb
 homepage: https://gitlab.com/gitlab-org/incubation-engineering/mlops/rb-ipynbdiff
 licenses: