support PDF upload

Author: zalepa

app/controllers/admin/meetings_controller.rb

@@ -19,11 +19,13 @@ def new
 
     def create
       unless params[:agenda_file].present?
-        flash.now[:alert] = "Please select a JSON file to upload."
+        flash.now[:alert] = "Please select a file to upload."
         return render :new, status: :unprocessable_entity
       end
 
-      data = JSON.parse(params[:agenda_file].read)
+      data = parse_agenda_upload(params[:agenda_file])
+      return render :new, status: :unprocessable_entity unless data
+
       service = AgendaImportService.new(data).call
 
       if service.success?
@@ -37,20 +39,22 @@ def create
         flash.now[:alert] = service.errors.join(", ")
         render :new, status: :unprocessable_entity
       end
-    rescue JSON::ParserError
-      flash.now[:alert] = "Invalid JSON file."
-      render :new, status: :unprocessable_entity
     end
 
     def import_minutes
       @meeting = Meeting.find(params[:id])
 
       unless params[:minutes_file].present?
-        redirect_to admin_meeting_path(@meeting), alert: "Please select a JSON file to upload."
+        redirect_to admin_meeting_path(@meeting), alert: "Please select a file to upload."
+        return
+      end
+
+      data = parse_minutes_upload(params[:minutes_file])
+      unless data
+        redirect_to admin_meeting_path(@meeting), alert: @parse_error
         return
       end
 
-      data = JSON.parse(params[:minutes_file].read)
       service = MinutesImportService.new(data).call
 
       if service.success?
@@ -60,8 +64,6 @@ def import_minutes
       else
         redirect_to admin_meeting_path(@meeting), alert: service.errors.join(", ")
       end
-    rescue JSON::ParserError
-      redirect_to admin_meeting_path(@meeting), alert: "Invalid JSON file."
     end
 
     def delete_minutes
@@ -94,5 +96,46 @@ def destroy
       @meeting.destroy!
       redirect_to admin_meetings_path, notice: "Meeting deleted."
     end
+
+    private
+
+    def parse_agenda_upload(file)
+      if pdf_file?(file)
+        parser = AgendaPdfParserService.new(file.tempfile)
+        data = parser.call
+        unless parser.success?
+          flash.now[:alert] = parser.errors.join(", ")
+          return nil
+        end
+        data
+      else
+        JSON.parse(file.read)
+      end
+    rescue JSON::ParserError
+      flash.now[:alert] = "Invalid JSON file."
+      nil
+    end
+
+    def parse_minutes_upload(file)
+      if pdf_file?(file)
+        parser = MinutesPdfParserService.new(file.tempfile)
+        data = parser.call
+        unless parser.success?
+          @parse_error = parser.errors.join(", ")
+          return nil
+        end
+        data
+      else
+        JSON.parse(file.read)
+      end
+    rescue JSON::ParserError
+      @parse_error = "Invalid JSON file."
+      nil
+    end
+
+    def pdf_file?(file)
+      file.content_type == "application/pdf" ||
+        file.original_filename&.end_with?(".pdf")
+    end
   end
 end

app/services/agenda_pdf_parser_service.rb

@@ -0,0 +1,51 @@
+class AgendaPdfParserService
+  attr_reader :errors
+
+  PARSER_SCRIPT = Rails.root.join("lib", "parsers", "parse_agenda.py").to_s
+
+  def initialize(pdf_io)
+    @pdf_io = pdf_io
+    @errors = []
+  end
+
+  def call
+    tempfile = write_to_tempfile
+    output_file = Tempfile.new([ "agenda_parsed", ".json" ])
+
+    stdout, stderr, status = Open3.capture3(
+      "python3", PARSER_SCRIPT, tempfile.path, output_file.path
+    )
+
+    unless status.success?
+      @errors << "PDF parsing failed: #{stderr.presence || stdout}"
+      return nil
+    end
+
+    JSON.parse(File.read(output_file.path))
+  rescue JSON::ParserError => e
+    @errors << "Parser produced invalid JSON: #{e.message}"
+    nil
+  ensure
+    tempfile&.close!
+    output_file&.close!
+  end
+
+  def success?
+    @errors.empty?
+  end
+
+  private
+
+  def write_to_tempfile
+    temp = Tempfile.new([ "agenda_upload", ".pdf" ])
+    temp.binmode
+    if @pdf_io.respond_to?(:read)
+      temp.write(@pdf_io.read)
+      @pdf_io.rewind if @pdf_io.respond_to?(:rewind)
+    else
+      temp.write(File.binread(@pdf_io))
+    end
+    temp.flush
+    temp
+  end
+end

app/services/minutes_pdf_parser_service.rb

@@ -0,0 +1,51 @@
+class MinutesPdfParserService
+  attr_reader :errors
+
+  PARSER_SCRIPT = Rails.root.join("lib", "parsers", "parse_minutes.py").to_s
+
+  def initialize(pdf_io)
+    @pdf_io = pdf_io
+    @errors = []
+  end
+
+  def call
+    tempfile = write_to_tempfile
+    output_file = Tempfile.new([ "minutes_parsed", ".json" ])
+
+    stdout, stderr, status = Open3.capture3(
+      "python3", PARSER_SCRIPT, tempfile.path, output_file.path
+    )
+
+    unless status.success?
+      @errors << "PDF parsing failed: #{stderr.presence || stdout}"
+      return nil
+    end
+
+    JSON.parse(File.read(output_file.path))
+  rescue JSON::ParserError => e
+    @errors << "Parser produced invalid JSON: #{e.message}"
+    nil
+  ensure
+    tempfile&.close!
+    output_file&.close!
+  end
+
+  def success?
+    @errors.empty?
+  end
+
+  private
+
+  def write_to_tempfile
+    temp = Tempfile.new([ "minutes_upload", ".pdf" ])
+    temp.binmode
+    if @pdf_io.respond_to?(:read)
+      temp.write(@pdf_io.read)
+      @pdf_io.rewind if @pdf_io.respond_to?(:rewind)
+    else
+      temp.write(File.binread(@pdf_io))
+    end
+    temp.flush
+    temp
+  end
+end

app/views/admin/meetings/new.html.erb

@@ -3,9 +3,9 @@
 <div class="max-w-lg bg-white rounded-lg border border-gray-200 p-6">
   <%= form_with url: admin_meetings_path, multipart: true do |f| %>
     <div class="mb-4">
-      <label class="block text-sm font-medium text-gray-700 mb-2">Agenda JSON file</label>
-      <input type="file" name="agenda_file" accept=".json" class="block w-full text-sm text-gray-500 file:mr-4 file:py-2 file:px-4 file:rounded-lg file:border-0 file:text-sm file:font-medium file:bg-blue-50 file:text-blue-700 hover:file:bg-blue-100 file:cursor-pointer" />
-      <p class="mt-1 text-xs text-gray-500">Select a structured JSON file containing the meeting agenda.</p>
+      <label class="block text-sm font-medium text-gray-700 mb-2">Agenda file</label>
+      <input type="file" name="agenda_file" accept=".json,.pdf" class="block w-full text-sm text-gray-500 file:mr-4 file:py-2 file:px-4 file:rounded-lg file:border-0 file:text-sm file:font-medium file:bg-blue-50 file:text-blue-700 hover:file:bg-blue-100 file:cursor-pointer" />
+      <p class="mt-1 text-xs text-gray-500">Upload a PDF agenda or a structured JSON file.</p>
     </div>
 
     <div class="flex items-center gap-3">

app/views/admin/meetings/show.html.erb

@@ -55,7 +55,7 @@
   <% end %>
 
   <%= form_with url: import_minutes_admin_meeting_path(@meeting), method: :post, multipart: true, class: "flex items-center gap-3" do %>
-    <input type="file" name="minutes_file" accept=".json" class="text-sm text-gray-500 file:mr-2 file:py-2 file:px-4 file:rounded-lg file:border-0 file:text-sm file:font-medium file:bg-gray-100 file:text-gray-700 hover:file:bg-gray-200 file:cursor-pointer file:transition-colors" />
+    <input type="file" name="minutes_file" accept=".json,.pdf" class="text-sm text-gray-500 file:mr-2 file:py-2 file:px-4 file:rounded-lg file:border-0 file:text-sm file:font-medium file:bg-gray-100 file:text-gray-700 hover:file:bg-gray-200 file:cursor-pointer file:transition-colors" />
     <button type="submit" class="px-4 py-2 bg-blue-600 text-white text-sm font-medium rounded-lg hover:bg-blue-700 transition-colors cursor-pointer">Import Minutes</button>
   <% end %>
 </div>