Class: SwarmSDK::Tools::DocumentConverters::DocxConverter

Inherits:
BaseConverter
  • Object
show all
Defined in:
lib/swarm_sdk/tools/document_converters/docx_converter.rb

Overview

Converts DOCX documents to text with image extraction

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from BaseConverter

available?, gem_available?

Class Method Details

.extensions ⇒ Object 



17
18
19
 
# File 'lib/swarm_sdk/tools/document_converters/docx_converter.rb', line 17

def extensions
  [".docx", ".doc"]
end

.format_name ⇒ Object 



13
14
15
 
# File 'lib/swarm_sdk/tools/document_converters/docx_converter.rb', line 13

def format_name
  "DOCX"
end

.gem_name ⇒ Object 



9
10
11
 
# File 'lib/swarm_sdk/tools/document_converters/docx_converter.rb', line 9

def gem_name
  "docx"
end

Instance Method Details

#convert(file_path) ⇒ String, RubyLLM::Content

Convert a DOCX document to text/content

Parameters:

  • file_path (String) —

    Path to the DOCX file

Returns:

  • (String, RubyLLM::Content) —

    Converted content or error message



25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
 
# File 'lib/swarm_sdk/tools/document_converters/docx_converter.rb', line 25

def convert(file_path)
  unless self.class.available?
    return unsupported_format_reminder(self.class.format_name, self.class.gem_name)
  end

  # Check for legacy DOC format
  if File.extname(file_path).downcase == ".doc"
    return error("DOC format is not supported. Please convert to DOCX first.")
  end

  begin
    require "docx"
    require "tmpdir"

    doc = Docx::Document.open(file_path)

    # Extract images from the DOCX
    image_paths = ImageExtractors::DocxImageExtractor.extract_images(doc, file_path)

    output = []
    output << "Document: #{File.basename(file_path)}"
    output << "=" * 60
    output << ""

    # Extract paragraphs
    paragraphs = doc.paragraphs.map(&:text).reject(&:empty?)

    # Check for empty document
    if paragraphs.empty? && doc.tables.empty?
      output << "(Document is empty - no paragraphs or tables)"
    else
      output += paragraphs

      # Extract tables with enhanced formatting
      if doc.tables.any?
        output << ""
        output << "Tables:"
        output << "-" * 60

        doc.tables.each_with_index do |table, idx|
          output << ""
          output << "Table #{idx + 1} (#{table.row_count} rows × #{table.column_count} columns):"

          table.rows.each do |row|
            output << row.cells.map(&:text).join(" | ")
          end
        end
      end
    end

    text_content = output.join("\n")

    # If there are images, return Content with attachments
    if image_paths.any?
      content = RubyLLM::Content.new(text_content)
      image_paths.each do |image_path|
        content.add_attachment(image_path)
      end
      content
    else
      # No images, return just text
      text_content
    end
  rescue Zip::Error => e
    error("Invalid or corrupted DOCX file: #{e.message}")
  rescue Errno::ENOENT => e
    error("File not found or missing document.xml: #{e.message}")
  rescue StandardError => e
    error("Failed to parse DOCX file: #{e.message}")
  end
end