From 556072ec76d3bc26cbed9a4a012b638fcba5700a Mon Sep 17 00:00:00 2001
From: Abhijeet <129729795+luciferlinx101@users.noreply.github.com>
Date: Fri, 25 Aug 2023 17:27:03 +0530
Subject: [PATCH] Revert "PDF and DOCX support in Write File - Feature
 Improvement, close #548 (#928)" (#1124)

This reverts commit 8b0135743860cc0942c32eb64ccfe23bcee531cc.
---
 Dockerfile                                    |   4 +-
 DockerfileCelery                              |   2 -
 requirements.txt                              |   3 -
 superagi/models/agent_execution_feed.py       |   7 +-
 superagi/resource_manager/file_manager.py     | 112 +++-----------
 .../tools/file/prompts/add_images_to_html.txt |   4 -
 .../file/prompts/content_to_html_prompt.txt   |   7 -
 superagi/tools/file/write_file.py             | 143 ++----------------
 .../stable_diffusion_image_gen.py             |   6 +-
 superagi/tools/tool_response_query_manager.py |   2 +-
 10 files changed, 36 insertions(+), 254 deletions(-)
 delete mode 100644 superagi/tools/file/prompts/add_images_to_html.txt
 delete mode 100644 superagi/tools/file/prompts/content_to_html_prompt.txt

diff --git a/Dockerfile b/Dockerfile
index 10a7d7f17..250f94a12 100644
--- a/Dockerfile
+++ b/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 wkhtmltopdf && \
+    apt-get install --no-install-recommends -y wget libpq-dev gcc g++ python3-dev && \
     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 wkhtmltopdf && \
+    apt-get install --no-install-recommends -y libpq-dev && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
 
diff --git a/DockerfileCelery b/DockerfileCelery
index ba608dcbd..682e50824 100644
--- a/DockerfileCelery
+++ b/DockerfileCelery
@@ -3,8 +3,6 @@ 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 .
diff --git a/requirements.txt b/requirements.txt
index be008f6fb..65e19c2c1 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -155,6 +155,3 @@ 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
diff --git a/superagi/models/agent_execution_feed.py b/superagi/models/agent_execution_feed.py
index aa7c3a148..61f449585 100644
--- a/superagi/models/agent_execution_feed.py
+++ b/superagi/models/agent_execution_feed.py
@@ -47,14 +47,12 @@ def get_last_tool_response(cls, session: Session, agent_execution_id: int, tool_
             AgentExecutionFeed.role == "system").order_by(AgentExecutionFeed.created_at.desc()).all()
 
         for agent_execution_feed in agent_execution_feeds:
-            if tool_name and not agent_execution_feed.feed.startswith(
-                f"Tool {tool_name}"
-            ):
+            if tool_name and not agent_execution_feed.feed.startswith("Tool " + tool_name):
                 continue
             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)
@@ -68,4 +66,3 @@ def fetch_agent_execution_feeds(cls, session, agent_execution_id: int):
             return agent_feeds
         else:
             return agent_feeds[2:]
-        
diff --git a/superagi/resource_manager/file_manager.py b/superagi/resource_manager/file_manager.py
index dbd0e4d16..4c20ba16d 100644
--- a/superagi/resource_manager/file_manager.py
+++ b/superagi/resource_manager/file_manager.py
@@ -8,21 +8,11 @@
 from superagi.models.agent import Agent
 from superagi.models.agent_execution import AgentExecution
 from superagi.types.storage_types import StorageType
-import pdfkit
-from htmldocx import HtmlToDocx
-
-class UnsupportedFileTypeError(Exception):
-    pass
-
-class FileNotCreatedError(Exception):
-    pass
-
 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,
@@ -42,7 +32,6 @@ 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,
@@ -56,7 +45,7 @@ def write_to_s3(self, file_name, final_path):
                 s3_helper = S3Helper()
                 s3_helper.upload_file(img, path=resource.path)
 
-    def write_file(self, file_name: str, content, return_file_path: bool = False):
+    def write_file(self, file_name: str, content):
         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,
@@ -66,19 +55,25 @@ def write_file(self, file_name: str, content, return_file_path: bool = False):
                                                                                                    self.agent_execution_id))
         else:
             final_path = ResourceHelper.get_resource_path(file_name)
-
         try:
-            self.save_file_by_type(file_name=file_name, file_path=final_path, content=content)
+            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"
         except Exception as err:
             return f"Error write_file: {err}"
-
-        logger.info(f"{file_name} - File written successfully")
-        
-        if return_file_path:
-            return final_path
-        return f"{file_name} - File written successfully"
-        
-    def write_csv_file(self, file_name: str, final_path: str, csv_data) -> str:
+    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)
         try:
             with open(final_path, mode="w", newline="") as file:
                 writer = csv.writer(file, lineterminator="\n")
@@ -87,63 +82,15 @@ def write_csv_file(self, file_name: str, final_path: str, csv_data) -> str:
             logger.info(f"{file_name} - File written successfully")
             return f"{file_name} - File written successfully"
         except Exception as err:
-            raise FileNotCreatedError(f"Error write_file: {err}") 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
+            return f"Error write_csv_file: {err}"
         
-        except Exception as err:
-            raise FileNotCreatedError(f"Error write_file: {err}") 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(f"Error write_file: {err}") 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(f"Error write_file: {err}") 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)
@@ -157,7 +104,6 @@ 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.
@@ -176,25 +122,3 @@ 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 in file_type_handlers:
-            return file_type_handlers[file_type](file_name, file_path, content)
-        else:
-            raise UnsupportedFileTypeError(f"Unsupported file type: {file_type}. Cannot save the file.")
-        
\ No newline at end of file
diff --git a/superagi/tools/file/prompts/add_images_to_html.txt b/superagi/tools/file/prompts/add_images_to_html.txt
deleted file mode 100644
index 57dda0a49..000000000
--- a/superagi/tools/file/prompts/add_images_to_html.txt
+++ /dev/null
@@ -1,4 +0,0 @@
-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:
diff --git a/superagi/tools/file/prompts/content_to_html_prompt.txt b/superagi/tools/file/prompts/content_to_html_prompt.txt
deleted file mode 100644
index e450f46e4..000000000
--- a/superagi/tools/file/prompts/content_to_html_prompt.txt
+++ /dev/null
@@ -1,7 +0,0 @@
-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}`
\ No newline at end of file
diff --git a/superagi/tools/file/write_file.py b/superagi/tools/file/write_file.py
index 4fcc3d75f..f8b61aeaf 100644
--- a/superagi/tools/file/write_file.py
+++ b/superagi/tools/file/write_file.py
@@ -1,17 +1,13 @@
 from typing import Type, Optional
-import base64
-import os
 
 from pydantic import BaseModel, Field
 
-from superagi.helper.prompt_reader import PromptReader
-from superagi.helper.resource_helper import ResourceHelper
+# from superagi.helper.s3_helper import upload_to_s3
 from superagi.resource_manager.file_manager import FileManager
 from superagi.tools.base_tool import BaseTool
-from superagi.tools.tool_response_query_manager import ToolResponseQueryManager
-from superagi.llms.base_llm import BaseLlm
-from superagi.models.agent import Agent
-from superagi.models.agent_execution import AgentExecution
+
+
+# from superagi.helper.s3_helper import upload_to_s3
 
 
 class WriteFileInput(BaseModel):
@@ -19,6 +15,7 @@ class WriteFileInput(BaseModel):
     file_name: str = Field(..., description="Name of the file to write. Only include the file name. Don't include path.")
     content: str = Field(..., description="File content to write")
 
+
 class WriteFileTool(BaseTool):
     """
     Write File tool
@@ -30,13 +27,12 @@ class WriteFileTool(BaseTool):
         args_schema : The args schema.
         resource_manager: File resource manager.
     """
-    llm: Optional[BaseLlm] = None
     name: str = "Write File"
     args_schema: Type[BaseModel] = WriteFileInput
-    description: str = "Writes content in a file. The content can carry text and images."
+    description: str = "Writes text to a file"
+    agent_id: int = None
     resource_manager: Optional[FileManager] = None
-    agent_id:int =None
-    agent_execution_id:int =None
+
     class Config:
         arbitrary_types_allowed = True
 
@@ -51,125 +47,4 @@ def _execute(self, file_name: str, content: str):
         Returns:
             success message if message is file written successfully or failure message if writing file fails.
         """
-        
-        attached_files = self._get_attached_files()
-        file_type = file_name.split('.')[-1].lower()
-        
-        if file_type not in ['pdf', 'docx', 'doc']:
-            return self.resource_manager.write_file(file_name, content)
-        
-        html_code_content = self._convert_content_into_html(content=content, attached_files = attached_files,formated_for=file_type)
-        
-        return self.resource_manager.write_file(file_name, html_code_content)
-
-    def _convert_content_into_html(self, content: str, attached_files: list, formated_for: str) -> str:
-        """
-        Converts the content into an HTML file
-        Args:
-            content (str): Content to be beautified and formatted
-            formated_for (str): HTML content to be specifically formatted for a specific document type
-        Returns:
-            HTML Code (str): HTML code of the formated content 
-        """
-        prompt = PromptReader.read_tools_prompt(__file__, "content_to_html_prompt.txt")
-        prompt = prompt.replace("{content}", content)
-
-        if image_file_paths := self._get_file_path_of_images(attached_files=attached_files):
-            embedding_image_prompt = PromptReader.read_tools_prompt(__file__, "add_images_to_html.txt")
-            for idx, image_path in enumerate(image_file_paths):
-                embedding_image_prompt += f"\n{idx+1}. {image_path}"
-            prompt = prompt.replace("{embedding_image}", embedding_image_prompt)
-        else:
-            prompt = prompt.replace("{embedding_image}", "")
-            
-        messages = [{"role": "system", "content": prompt}]
-        result = self.llm.chat_completion(messages, max_tokens=self.max_token_limit)
-        
-        if formated_for == 'pdf':
-            return self._html_formatting_for_pdf(content=result["content"], image_list=image_file_paths)
-        return result['content']
-    
-    def _get_file_path_of_images(self, attached_files: list):
-        """
-        Filters Images from the attached files list and finds out the corresponding paths. 
-
-        Args:
-            attached_files: List of names of files generated
-
-        Returns:
-            Full paths of the image files
-        """
-        image_extensions, image_paths = ['.jpg', '.jpeg', '.png', '.gif', '.bmp', '.tiff', '.webp'],[]
-        for file_name in attached_files:
-            path = self._get_file_path(file_name=file_name)
-            if not os.path.exists(path):
-                continue
-            _, file_extension = os.path.splitext(path)
-            # Check if the file extension is in the list of image extensions
-            if file_extension.lower() in image_extensions:
-                image_paths.append(path)
-        return image_paths
-    
-    def _get_file_path(self, file_name: str) -> list:
-        """
-        Gets the path of the file
-
-        Args:
-            file_name: Name of file 
-
-        Returns:
-            Full paths of the file
-        """
-        session = self.toolkit_config.session
-
-        return ResourceHelper().get_agent_read_resource_path(
-            file_name,
-            agent=Agent.get_agent_from_id(session, self.agent_id),
-            agent_execution=AgentExecution.get_agent_execution_from_id(
-                session, self.agent_execution_id
-            ),
-        )
-    
-    def _image_to_base64(self, image_path):
-        with open(image_path, "rb") as image_file:
-            encoded_image = base64.b64encode(image_file.read()).decode("utf-8")
-        return encoded_image
-    
-    def _html_formatting_for_pdf(self, content: str, image_list: list) -> str:
-        """
-        Converts image paths into base64 inputs in an HTML file
-        Args:
-            content (str): HTML code
-            image_list (list): List of images to be converted and replaced
-        Returns:
-            HTML Code (str): Formatted HTML code 
-        """
-        
-        for image_path in image_list:
-            content = content.replace(f"{image_path}", f"data:image/png;base64,{self._image_to_base64(image_path=image_path)}")
-        return content
-
-    def _get_attached_files(self):
-        output_directory = ResourceHelper.get_root_output_dir()
-        if "{agent_id}" not in output_directory:
-            return []
-        output_directory = ResourceHelper.get_formatted_agent_level_path(agent=Agent
-                                                                        .get_agent_from_id(session=self
-                                                                                           .toolkit_config.session,
-                                                                                           agent_id=self.agent_id),
-                                                                        path=output_directory)
-        agent_execution=AgentExecution.get_agent_execution_from_id(session=self.toolkit_config.session, agent_execution_id=self.agent_execution_id)
-        if agent_execution is not None and "{agent_execution_id}" in output_directory:
-            output_directory = ResourceHelper.get_formatted_agent_execution_level_path(agent_execution=agent_execution, path=output_directory)
-
-        return self._list_files(output_directory)
-    
-    def _list_files(self, directory):
-        found_files = []
-        for root, dirs, files in os.walk(directory):
-            found_files.extend(
-                file
-                for file in files
-                if not file.startswith(".") and "__pycache__" not in root
-            )
-        return found_files
\ No newline at end of file
+        return self.resource_manager.write_file(file_name, content)
diff --git a/superagi/tools/image_generation/stable_diffusion_image_gen.py b/superagi/tools/image_generation/stable_diffusion_image_gen.py
index 8a71f7c42..b3c5d2c96 100644
--- a/superagi/tools/image_generation/stable_diffusion_image_gen.py
+++ b/superagi/tools/image_generation/stable_diffusion_image_gen.py
@@ -57,8 +57,10 @@ def _execute(self, prompt: str, image_names: list, width: int = 512, height: int
         data = response.json()
 
         artifacts = data['artifacts']
-        base64_strings = [artifact['base64'] for artifact in artifacts]
-        image_paths=[]
+        base64_strings = []
+        for artifact in artifacts:
+            base64_strings.append(artifact['base64'])
+
         for i in range(num):
             image_base64 = base64_strings[i]
             img_data = base64.b64decode(image_base64)
diff --git a/superagi/tools/tool_response_query_manager.py b/superagi/tools/tool_response_query_manager.py
index 3aebc6eab..d2caf5e23 100644
--- a/superagi/tools/tool_response_query_manager.py
+++ b/superagi/tools/tool_response_query_manager.py
@@ -11,7 +11,7 @@ def __init__(self, session: Session, agent_execution_id: int,memory:VectorStore)
 
     def get_last_response(self, tool_name: str = None):
         return AgentExecutionFeed.get_last_tool_response(self.session, self.agent_execution_id, tool_name)
-      
+    
     def get_relevant_response(self, query: str,metadata:dict, top_k: int = 5):
         documents = self.memory.get_matching_text(query, metadata=metadata)
         relevant_responses = ""