Class: Telegem::Plugins::FileExtract

Inherits:
Object
  • Object
show all
Defined in:
lib/plugins/file_extract.rb

Instance Method Summary collapse

Constructor Details

#initialize(bot, file_id, **options) ⇒ FileExtract

Returns a new instance of FileExtract.



9
10
11
12
13
14
15
16
17
18
19
 
# File 'lib/plugins/file_extract.rb', line 9

def initialize(bot, file_id, **options)
  @bot = bot
  @file_id = file_id
  @options = {
    auto_delete: true,
    max_file_size: 50 * 1024 * 1024,
    timeout: 60
  }.merge(options)
@temp_file = nil
@detected_type = nil
end

Instance Method Details

#cleanup ⇒ Object 



176
177
178
179
 
# File 'lib/plugins/file_extract.rb', line 176

def cleanup
  @temp_file.unlink if @temp_file
  @temp_file = nil
end

#extract ⇒ Object 



20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
 
# File 'lib/plugins/file_extract.rb', line 20

def extract
  download_if_needed
  detect_type
  case @detected_type
  when :pdf then extract_pdf
  when :json then extract_json
  when :html then extract_html
  when :txt, :md, :csv then extract_text
  else 
    {
      success: false,
      error: "Unsupported file type: #{@detected_type}"
    }
  end 
end

#extract_html ⇒ Object 



122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
 
# File 'lib/plugins/file_extract.rb', line 122

def extract_html
  begin 
    content = File.read(@temp_file.path)
    if content.strip.empty?
      return {
        success: false,
        error: "HTML file is empty"
      }
    end
    {
      success: true,
      type: :html,
      content: content,
      metadata: {
        size: content.size,
        title: content[/<title>(.*?)<\/title>/i, 1]
      }
    }
  rescue => e
    {
      success: false,
      error: "Failed to extract HTML: #{e.message}"
    }
    ensure 
      cleanup if @options[:auto_delete]   
  end
end

#extract_text ⇒ Object 



149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
 
# File 'lib/plugins/file_extract.rb', line 149

def extract_text
  begin 
    content = File.read(@temp_file.path)
    if content.strip.empty?
      return {
        success: false,
        error: "Text file is empty"
      }
    end
    {
      success: true,
      type: :text,
      content: content,
      metadata: {
        size: content.size,
        line_count: content.lines.count
      }
    }
  rescue => e
    {
      success: false,
      error: "Failed to extract text: #{e.message}"
    }
    ensure 
      cleanup if @options[:auto_delete]   
  end
end