PDF and DOCX support in Write File - Feature Improvement, close #548

Dockerfile

@@ -3,7 +3,7 @@ FROM python:3.10-slim-bullseye AS compile-image
 WORKDIR /app
 
 RUN apt-get update && \
-    apt-get install --no-install-recommends -y wget libpq-dev gcc g++ python3-dev && \
+    apt-get install --no-install-recommends -y wget libpq-dev gcc g++ python3-dev wkhtmltopdf && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
 
@@ -24,7 +24,7 @@ FROM python:3.10-slim-bullseye AS build-image
 WORKDIR /app
 
 RUN apt-get update && \
-    apt-get install --no-install-recommends -y libpq-dev && \
+    apt-get install --no-install-recommends -y libpq-dev wkhtmltopdf && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
 

DockerfileCelery

@@ -3,6 +3,8 @@ FROM python:3.9
 WORKDIR /app
 
 #RUN apt-get update && apt-get install --no-install-recommends -y git wget libpq-dev gcc python3-dev && pip install psycopg2
+RUN apt-get update && apt-get install -y wkhtmltopdf
+
 RUN pip install --upgrade pip
 
 COPY requirements.txt .

requirements.txt

@@ -154,3 +154,6 @@ html2text==2020.1.16
 duckduckgo-search==3.8.3 
 google-generativeai==0.1.0
 unstructured==0.8.1
+beautifulsoup4==4.12.2
+pdfkit==1.0.0
+htmldocx==0.0.6

superagi/exceptions/file_exceptions.py

@@ -0,0 +1,10 @@
+
+class UnsupportedFileTypeError(Exception):
+    def __init__(self, file_name: str, supported_types: list):
+        message = f"Unsupported file type for '{file_name}'. Supported types are: {', '.join(supported_types)}"
+        super().__init__(message)
+
+class FileNotCreatedError(Exception):
+    def __init__(self, file_name: str):
+        message = f"Failed to create the file '{file_name}'."
+        super().__init__(message)

superagi/models/agent_execution_feed.py

@@ -52,7 +52,7 @@ def get_last_tool_response(cls, session: Session, agent_execution_id: int, tool_
             if agent_execution_feed.feed.startswith("Tool"):
                 return agent_execution_feed.feed
         return ""
-
+    
     @classmethod
     def fetch_agent_execution_feeds(cls, session, agent_execution_id: int):
         agent_execution = AgentExecution.find_by_id(session, agent_execution_id)
@@ -66,3 +66,4 @@ def fetch_agent_execution_feeds(cls, session, agent_execution_id: int):
             return agent_feeds
         else:
             return agent_feeds[2:]
+

superagi/resource_manager/file_manager.py

@@ -1,18 +1,25 @@
 import csv
 from sqlalchemy.orm import Session
-from superagi.config.config import get_config
 import os
+
+from superagi.config.config import get_config
 from superagi.helper.resource_helper import ResourceHelper
 from superagi.helper.s3_helper import S3Helper
 from superagi.lib.logger import logger
 from superagi.models.agent import Agent
 from superagi.models.agent_execution import AgentExecution
 from superagi.types.storage_types import StorageType
+from superagi.exceptions.file_exceptions import UnsupportedFileTypeError, FileNotCreatedError
+
+import pdfkit
+from htmldocx import HtmlToDocx
+
 class FileManager:
     def __init__(self, session: Session, agent_id: int = None, agent_execution_id: int = None):
         self.session = session
         self.agent_id = agent_id
         self.agent_execution_id = agent_execution_id
+
     def write_binary_file(self, file_name: str, data):
         if self.agent_id is not None:
             final_path = ResourceHelper.get_agent_write_resource_path(file_name,
@@ -32,6 +39,7 @@ def write_binary_file(self, file_name: str, data):
             return f"Binary {file_name} saved successfully"
         except Exception as err:
             return f"Error write_binary_file: {err}"
+
     def write_to_s3(self, file_name, final_path):
         with open(final_path, 'rb') as img:
             resource = ResourceHelper.make_written_file_resource(file_name=file_name,
@@ -55,25 +63,16 @@ def write_file(self, file_name: str, content):
                                                                                                    self.agent_execution_id))
         else:
             final_path = ResourceHelper.get_resource_path(file_name)
+
         try:
-            with open(final_path, mode="w") as file:
-                file.write(content)
-                file.close()
-            self.write_to_s3(file_name, final_path)
-            logger.info(f"{file_name} - File written successfully")
-            return f"{file_name} - File written successfully"
+            self.save_file_by_type(file_name=file_name, file_path=final_path, content=content)
         except Exception as err:
             return f"Error write_file: {err}"
-    def write_csv_file(self, file_name: str, csv_data):
-        if self.agent_id is not None:
-            final_path = ResourceHelper.get_agent_write_resource_path(file_name,
-                                                                      agent=Agent.get_agent_from_id(self.session,
-                                                                                                    self.agent_id),
-                                                                      agent_execution=AgentExecution
-                                                                      .get_agent_execution_from_id(self.session,
-                                                                                                   self.agent_execution_id))
-        else:
-            final_path = ResourceHelper.get_resource_path(file_name)
+
+        logger.info(f"{file_name} - File written successfully")
+        return f"{file_name} - File written successfully"
+
+    def write_csv_file(self, file_name: str, final_path: str, csv_data) -> str:
         try:
             with open(final_path, mode="w", newline="") as file:
                 writer = csv.writer(file, lineterminator="\n")
@@ -82,15 +81,63 @@ def write_csv_file(self, file_name: str, csv_data):
             logger.info(f"{file_name} - File written successfully")
             return f"{file_name} - File written successfully"
         except Exception as err:
-            return f"Error write_csv_file: {err}"
+            raise FileNotCreatedError(file_name=file_name) from err
+
+    def write_pdf_file(self, file_name: str ,file_path: str, content):
+        # Saving the HTML file
+        html_file_path = f"{file_path[:-4]}.html"
+        self.write_txt_file(file_name=html_file_path.split('/')[-1], file_path=html_file_path, content=content)
+
+        # Convert HTML file to a PDF file
+        try:
+            options = {
+                'quiet': '',
+                'page-size': 'Letter',
+                'margin-top': '0.75in',
+                'margin-right': '0.75in',
+                'margin-bottom': '0.75in',
+                'margin-left': '0.75in',
+                'enable-local-file-access': ''
+            }
+            config = pdfkit.configuration(wkhtmltopdf = "/usr/bin/wkhtmltopdf")
+            pdfkit.from_file(html_file_path, file_path, options = options, configuration = config)
+            self.write_to_s3(file_name, file_path)
+            return file_path
 
+        except Exception as err:
+            raise FileNotCreatedError(file_name=file_name) from err
+
+    def write_docx_file(self, file_name: str ,file_path: str, content):
+        # Saving the HTML file
+        html_file_path = f"{file_path[:-4]}.html"
+        self.write_txt_file(file_name=html_file_path.split('/')[-1], file_path=html_file_path, content=content)
 
+        # Convert HTML file to a DOCx file
+        try:
+            new_parser = HtmlToDocx()
+            new_parser.parse_html_file(html_file_path, file_path)
+            self.write_to_s3(file_name, file_path)
+            return file_path
+        except Exception as err:
+            raise FileNotCreatedError(file_name=file_name) from err
+
+    def write_txt_file(self, file_name: str ,file_path: str, content) -> str:
+        try:
+            with open(file_path, mode="w") as file:
+                file.write(content)
+                file.close()
+            self.write_to_s3(file_name, file_path)
+            return file_path
+        except Exception as err:
+            raise FileNotCreatedError(file_name=file_name) from err
+
     def get_agent_resource_path(self, file_name: str):
         return ResourceHelper.get_agent_write_resource_path(file_name, agent=Agent.get_agent_from_id(self.session,
                                                                                                      self.agent_id),
                                                             agent_execution=AgentExecution
                                                             .get_agent_execution_from_id(self.session,
                                                                                          self.agent_execution_id))
+
     def read_file(self, file_name: str):
         if self.agent_id is not None:
             final_path = self.get_agent_resource_path(file_name)
@@ -104,6 +151,7 @@ def read_file(self, file_name: str):
             return content
         except Exception as err:
             return f"Error while reading file {file_name}: {err}"
+
     def get_files(self):
         """
         Gets all file names generated by the CodingTool.
@@ -122,3 +170,23 @@ def get_files(self):
             logger.error(f"Error while accessing files in {final_path}: {err}")
             files = []
         return files
+
+    def save_file_by_type(self, file_name: str, file_path: str, content):
+
+        # Extract the file type from the file_name
+        file_type = file_name.split('.')[-1].lower()
+
+        # Dictionary to map file types to corresponding functions
+        file_type_handlers = {
+            'txt': self.write_txt_file,
+            'pdf': self.write_pdf_file,
+            'docx': self.write_docx_file, 
+            'doc': self.write_docx_file,
+            'csv': self.write_csv_file,
+            'html': self.write_txt_file
+            # NOTE: Add more file types and corresponding functions as needed, These functions should be defined 
+        }
+
+        if file_type not in file_type_handlers:
+            raise UnsupportedFileTypeError(file_name=file_name, supported_types=list(file_type_handlers))
+

superagi/tools/file/prompts/add_images_to_html.txt

@@ -0,0 +1,4 @@
+Now, you will be provided with few image path locations. You will have to attach the following images in appropriate locations inside the html code.  
+Remember to maintain the elegancy and styling of the User Interface generated. Make sure you attach all the images provided to you.
+
+The relevant paths of the images are provided below:

superagi/tools/file/prompts/content_to_html_prompt.txt

@@ -0,0 +1,7 @@
+You are an HTML code generating AI Agent. Your task is to generate a well formatted and well styled HTML file for a given content. 
+Remember to style the HTML beautifully, for which you can add the <style> element in the html itself. Make sure that the User Interface looks elegant to the user. 
+Make sure you have covered all of the content which is provided to you. 
+
+{embedding_image}.
+
+Content to be used:`{content}`