import os
from PIL import Image
import cv2
import anthropic
import requests
from dotenv import load_dotenv
import pytesseract
import tempfile
import ssl
import instaloader
import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
import numpy as np
from requests.auth import HTTPBasicAuth
import mimetypes
import os
from PIL import Image
import cv2
import anthropic
from wordpress_xmlrpc import Client, WordPressPost
from wordpress_xmlrpc.methods.posts import NewPost
from wordpress_xmlrpc.compat import xmlrpc_client
from dotenv import load_dotenv
import pytesseract
import tempfile
import base64
from instascrape import Profile, Post
import requests
from io import BytesIO
import time
import json
import requests
import traceback
from urllib.parse import urlparse
import os

import openpyxl
from openai import OpenAI
import ast
client = OpenAI(
  api_key = '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'
)

import os
import shutil

import ast

import openpyxl
from difflib import SequenceMatcher

import pandas as pd
from fuzzywuzzy import fuzz

import requests
from requests.auth import HTTPBasicAuth
import json

import requests
from requests.auth import HTTPBasicAuth
import json
import html
import re
from requests.auth import HTTPBasicAuth
def normalize_title(title):
    # Decode HTML entities
    decoded = html.unescape(title)
    # Remove any remaining HTML tags
    no_html = re.sub('<[^<]+?>', '', decoded)
    # Convert to lowercase and remove non-alphanumeric characters
    return re.sub(r'[^a-z0-9]', '', no_html.lower())

def publish_or_update_wordpress_post(site_url, username, password, post_data):
    # WordPress REST API endpoints
    posts_url = f"{site_url}/wp-json/wp/v2/posts"
    
    # Set up authentication
    auth = HTTPBasicAuth(username, password)
    
    # Check if the API is accessible
    try:
        response = requests.get(f"{site_url}/wp-json", auth=auth)
        response.raise_for_status()
    except requests.exceptions.RequestException as e:
        raise Exception(f"Failed to access WordPress API: {str(e)}")

    # Check if a post with the same title exists
    existing_post = check_existing_post_by_title(site_url, auth, post_data['title'])
    
    if existing_post:
        # Update existing post
        update_url = f"{posts_url}/{existing_post['id']}"
        post_data['id'] = existing_post['id']
        response = requests.post(update_url, json=post_data, auth=auth)
    else:
        # Create new post
        response = requests.post(posts_url, json=post_data, auth=auth)
    
    if response.status_code in [200, 201]:
        return response.json()
    else:
        raise Exception(f"Failed to publish/update post: {response.text}")

def check_existing_post_by_title(site_url, auth, title):
    # URL encode the title
    encoded_title = requests.utils.quote(title)
   # print(encoded_title)
    search_url = f"{site_url}/wp-json/wp/v2/posts?search={encoded_title}"
   # print(search_url)
    response = requests.get(search_url, auth=auth)
    title = normalize_title(title)
    
    if response.status_code == 200:
       # print('existing post found')
        posts = response.json()
        for post in posts:
            rendered_title = normalize_title(post['title']['rendered'].lower())
           # print(post['title']['rendered'].lower())
           # print(title.lower())
            print(rendered_title, title)
            if rendered_title == title:
                print('post found', rendered_title)
                return post
    return None

#try:
 #   result = publish_or_update_wordpress_post(site_url, username, password, post_data)
 #   print("Post published/updated successfully:", result['link'])
#except Exception as e:
#    print("Error:", str(e))

def replace_keywords(excel_file, sheet_name, keyword_array):
    # Read the Excel sheet
    df = pd.read_excel(excel_file, sheet_name=sheet_name)
    
    # Ensure the column with keywords is named 'Keywords'
    if 'Keywords' not in df.columns:
        raise ValueError("Excel sheet must have a 'Keywords' column")
    
    # Get the list of keywords from the Excel sheet
    excel_keywords = df['Keywords'].tolist()
    
    # Function to find the best match for a keyword
    def find_best_match(keyword):
        best_match = max(excel_keywords, key=lambda x: fuzz.ratio(keyword.lower(), str(x).lower()))
        return best_match if fuzz.ratio(keyword.lower(), str(best_match).lower()) > 70 else keyword
    
    # Replace keywords in the array with best matches
    replaced_keywords = [find_best_match(keyword) for keyword in keyword_array]
    
    return replaced_keywords

def match_keywords(excel_path, sheet_name, text, column_letter='A'):
    # Open the Excel workbook and select the specified sheet
    workbook = openpyxl.load_workbook(excel_path)
    sheet = workbook[sheet_name]

    # Read keywords/phrases from the Excel sheet
    keywords = [cell.value for cell in sheet[column_letter] if cell.value]

    # Function to calculate similarity ratio
    def similarity(a, b):
        return SequenceMatcher(None, a.lower(), b.lower()).ratio()

    # Calculate similarity scores for each keyword/phrase
    scores = [(keyword, max(similarity(keyword, word) for word in text.split())) for keyword in keywords]

    # Sort by similarity score in descending order and get top 5
    top_matches = sorted(scores, key=lambda x: x[1], reverse=True)[:5]

    # Return only the keywords/phrases, not the scores
    return [match[0] for match in top_matches]

def remove_common_elements(array1, array2):
    return [x for x in array1 if x not in set(array2)]
def remove_keyphrases_from_excel(file_path, keyphrases, output_path=None):
    # Load the workbook
    wb = openpyxl.load_workbook(file_path)
    
    # Iterate through all sheets
    for sheet in wb.worksheets:
        # Iterate through all cells in the sheet
        for row in sheet.iter_rows():
            for cell in row:
                if cell.value:
                    # Convert cell value to string
                    cell_value = str(cell.value)
                    # Check if any keyphrase is in the cell value
                    for phrase in keyphrases:
                        if phrase in cell_value:
                            # Remove the keyphrase
                            cell_value = cell_value.replace(phrase, '')
                    # Update the cell value
                    cell.value = cell_value if cell_value.strip() else None
    
    # Save the modified workbook
    if output_path:
        wb.save(output_path)
    else:
        wb.save(file_path)
    
    print("Keyphrases removed successfully.")


def read_array_from_file(file_path):
    # Open the text file in read mode
    with open(file_path, 'r') as file:
        # Read the content of the file
        content = file.read()
        # Use ast.literal_eval to safely evaluate the string representation of the array
        array = ast.literal_eval(content)
    return array

def remove_all_files_in_folder(folder_path):
    # Check if the folder exists
    if not os.path.exists(folder_path):
        #print(f"The folder {folder_path} does not exist.")
        return
    
    # Iterate over all files in the folder and remove them
    for filename in os.listdir(folder_path):
        file_path = os.path.join(folder_path, filename)
        
        try:
            # Check if it's a file (not a folder) and remove it
            if os.path.isfile(file_path):
                os.remove(file_path)
            elif os.path.isdir(file_path):
                # If it's a directory, remove the directory and its contents
                shutil.rmtree(file_path)
        except Exception as e:
            print(f"Error deleting {file_path}: {e}")
    
    print(f"All files and directories in {folder_path} have been removed.")




def call_openai(prompt, temperature, max_tokens):
    #prompt = "Farmonaut wants to classify this google search query into only one of the following categories: a. Precision Agriculture, b. API/ Development, c. Traceability, d. Plantation, e. Unrelated, f. pests, diseases and weeds, g. irrigation, h. yield forecast, i. area estimation and crop identification, j. geotagging, k. fertilizers and soil health, l. Satellite/GIS/Remote Sensing, m. agri-tech startup/ company, n. agriculture content, o. not worth doing SEO. Answer only one category (without category alphabet). Google search query to classify: " + google_query
    #prompt = "Classify whether this google search query is related or unrelated to what Farmonaut does. Strictly answer RELATED OR UNRELATED : " + google_query
    completion = client.chat.completions.create(
    model="gpt-4o-mini-2024-07-18",
    messages=[
        {"role": "system", "content": "You are a expert in SEO and a representative of Farmonaut."},
        {"role": "user", "content": prompt}
    ],
    max_tokens=max_tokens,
    temperature=temperature
    )
    return completion.choices[0].message.content
def save_to_file(file_name, content):
    with open(file_name, 'w') as file:
        file.write(content)
    print(f"Content saved to {file_name}")


def string_to_array(string):
    # Use ast.literal_eval to safely evaluate the string as a list
    try:
        array = ast.literal_eval(string)
        print(array)
    except:
        array = string
    return array
def get_first_column_values(file_path, sheet_name=None):
    # Load the workbook
    workbook = openpyxl.load_workbook(file_path, data_only=True)
    
    # If a sheet name is specified, load that sheet, otherwise use the active sheet
    sheet = workbook[sheet_name] if sheet_name else workbook.active

    # Get all the values from the first column (Column A)
    first_column_values = []
    for cell in sheet['A']:
        # Convert cell value to string and append to list, handle empty cells
        first_column_values.append(str(cell.value) if cell.value is not None else "")

    return first_column_values

# Example usage:
# file_path = 'your_file.xlsx'
# values = get_first_column_values(file_path, 'Sheet1')
# print(values)


def get_file_extension(url):
    # Parse the URL
    parsed_url = urlparse(url)
    
    # Get the path component of the URL
    path = parsed_url.path
    
    # Extract the file extension
    file_extension = os.path.splitext(path)[1]
    
    # Return the extension (without the dot) or an empty string if there's no extension
    return file_extension[1:] if file_extension else ""
# posts = list(posts)
ssl._create_default_https_context = ssl._create_unverified_context

# Load environment variables
load_dotenv()
# Instagram session id (you need to get this from your browser after logging in to Instagram)
SESSIONID = os.getenv("INSTAGRAM_SESSIONID")

# Headers for Instagram requests
headers = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/79.0.3945.74 Safari/537.36 Edg/79.0.309.43",
    "cookie": f'sessionid={SESSIONID};'
}

# Initialize Instaloader
L = instaloader.Instaloader()
#L.login('himanshujain4578', 'harish@4321')
L.post_metadata_txt_pattern = ""
L.download_geotags = False
L.save_metadata = False
L.save_metadata_json = False
L.download_comments = False

# Anthropic API key (replace with your actual key)
#anthropic_api_key = os.getenv("ANTHROPIC_API_KEY")

# WordPress credentials
wp_url = "https://www.farmonaut.com"
wp_username = "ankuromar296"
wp_password = "Tjat A2hz 9XMv pXJi YbV0 GR8o"


def remove_keywords(selected_keywords, excel_file, sheet_name, keyword_column, num_keywords=5):
    # Read the Excel sheet
    df = pd.read_excel(excel_file, sheet_name=sheet_name)
    
    # Ensure the keyword column exists
    if keyword_column not in df.columns:
        raise ValueError(f"Column '{keyword_column}' not found in the Excel sheet.")
    
   
    
    # Remove the selected keywords from the DataFrame
    df = df[~df[keyword_column].isin(selected_keywords)]
    
    # Save the updated DataFrame back to the Excel file
    df.to_excel(excel_file, sheet_name=sheet_name, index=False)
    

def select_and_remove_keywords(text, excel_file, sheet_name, keyword_column, num_keywords=5):
    # Read the Excel sheet
    df = pd.read_excel(excel_file, sheet_name=sheet_name)
    
    # Ensure the keyword column exists
    if keyword_column not in df.columns:
        raise ValueError(f"Column '{keyword_column}' not found in the Excel sheet.")
    
    # Get the list of keywords
    keywords = df[keyword_column].tolist()
    
    # Create a TF-IDF vectorizer
    vectorizer = TfidfVectorizer()
    # Ensure text and keywords are strings
    text = str(text) if isinstance(text, dict) else text
    keywords = [str(keyword) if isinstance(keyword, dict) else keyword for keyword in keywords]

    # Fit the vectorizer on the text and transform the keywords
    tfidf_matrix = vectorizer.fit_transform([text] + keywords)
    
    # Calculate cosine similarity between the text and each keyword
    cosine_similarities = (tfidf_matrix * tfidf_matrix.T).toarray()[0][1:]
    
    # Get the indices of the top num_keywords similar keywords
    top_indices = np.argsort(cosine_similarities)[-num_keywords:][::-1]
    
    # Select the top keywords
    selected_keywords = [keywords[i] for i in top_indices]
    
    # Remove the selected keywords from the DataFrame
    df = df[~df[keyword_column].isin(selected_keywords)]
    
    # Save the updated DataFrame back to the Excel file
    df.to_excel(excel_file, sheet_name=sheet_name, index=False)
    
    return selected_keywords

# Existing functions remain the same
# (select_and_remove_keywords, get_instagram_posts, extract_text_from_image, extract_text_from_video, generate_blog_content)
def call_genai(prompt, temperature, max_tokens):
    client = anthropic.Anthropic(
        # defaults to os.environ.get("ANTHROPIC_API_KEY")
        api_key="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",
    )
    message = client.messages.create(
        model="claude-3-5-sonnet-20240620",
        max_tokens=max_tokens,
       temperature=temperature,
       system = "You are an SEO expert, a gis/ remote sensing expert, an agriculture and horticulture scientist, and a representative of Farmonaut (farmonaut.com).",
    messages=[
        {"role": "user", "content": prompt}
    ]
    )
    #print(message)
    return message.content[0].text

def upload_media_to_wordpress(file_path, title):
    endpoint = f"{wp_url}/wp-json/wp/v2/media"
    auth = HTTPBasicAuth(wp_username, wp_password)
    
    mime_type, _ = mimetypes.guess_type(file_path)
    media_data = {
        'alt_text':title,
        'caption':title,
        'description':title
    }
    with open(file_path, 'rb') as file:
        files = {'file': (os.path.basename(file_path), file, mime_type)}
        response = requests.post(endpoint, files=files, auth=auth, json = media_data)
    
    if response.status_code == 201:
        return response.json()['id'], response.json()['source_url']
    else:
        print(f"Failed to upload media. Status code: {response.status_code}")
        print(f"Response: {response.text}")
        return None, None

def extract_text_from_video(video_path):
    video = cv2.VideoCapture(video_path)
    fps = int(video.get(cv2.CAP_PROP_FPS))
    frame_count = int(video.get(cv2.CAP_PROP_FRAME_COUNT))
    duration = frame_count / fps
    
    text = ""
    for i in range(0, int(duration), 1):
        video.set(cv2.CAP_PROP_POS_MSEC, i * 1000)
        success, frame = video.read()
        if not success:
            break
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        frame_text = pytesseract.image_to_string(gray)
        text += frame_text + "\n"
    
    video.release()
    return text
def process_media(post):
    media_info = []
    with tempfile.TemporaryDirectory() as tmpdir:
        L.download_post(post, target=tmpdir)
        media_files = os.listdir(tmpdir)
        
        for file in media_files:
            file_path = os.path.join(tmpdir, file)
            if file.endswith('.mp4'):
                text = extract_text_from_video(file_path)
                media_type = 'video'
            elif file.endswith(('.jpg', '.jpeg', '.png', 'heic')):
                text = extract_text_from_image(file_path)
                media_type = 'image'
                # Upload image to WordPress
                media_id, media_url = upload_media_to_wordpress(file_path)
                if media_id and media_url:
                    media_info.append({
                        'type': media_type,
                        'text': text,
                        'id': media_id,
                        'url': media_url
                    })
            else:
                continue  # Skip files that are neither videos nor images
    
    return media_info


def process_media2(media_url_arr, title):
    #print(media_url_arr)
    media_info = []
    media_num = 0
    folder_name = 'insta_files'
    for url in media_url_arr:
      #print(url)
      media_num = media_num + 1
      
      file_path = folder_name + '/' + str(media_num) + '.' + str(get_file_extension(url))
      try:
        response = requests.get(url, stream=True)
      except:
        response = None
      if response is None:
        continue
      if response.status_code == 200:
          with open(file_path, 'wb') as file:
              for chunk in response.iter_content(chunk_size=8192):
                  if chunk:
                      file.write(chunk)
          print(f"File downloaded successfully: {file_path}")
      else:
          print(f"Failed to download file. Status code: {response.status_code}")

    
    # with tempfile.TemporaryDirectory() as tmpdir:
    #     for media_url in media_url_arr:
            
    #     L.download_post(post, target=tmpdir)
    media_files = os.listdir(folder_name)
    
    for file in media_files:
        file_path = os.path.join(folder_name, file)

            
        media_id, media_url = upload_media_to_wordpress(file_path, title)
        if media_id and media_url:
            media_info.append({
                'type': 'image',
                'text': title,
                'id': media_id,
                'url': media_url
            })
        else:
            continue  # Skip files that are neither videos nor images

    return media_info


def extract_text_from_image(image_path):
    image = Image.open(image_path)
    text = pytesseract.image_to_string(image)
    return text

def publish_to_wordpress(title, content, media_info, excerpt, category_id):
    endpoint = f"{wp_url}/posts"
    auth = HTTPBasicAuth(wp_username, wp_password)
    
    # Add images to the content
    for media in media_info:
        if media['type'] == 'image':
            content += f'\n\n<img src="{media["url"]}" alt="Instagram image">'
    slug = title.replace(' ', '-')
    post_data = {
        'title': title,
        'content': content,
        'status': 'publish',
        'excerpt':excerpt,
        'slug':slug,
        'comment_status':'open',
        'categories':[category_id],
        'featured_media': media_info[0]['id'] if media_info else None  # Set the first image as featured image
    }
    
    response = requests.post(endpoint, json=post_data, auth=auth)
    
    if response.status_code == 201:
        print("Post published successfully!")
    else:
        print(f"Failed to publish post. Status code: {response.status_code}")
        print(f"Response: {response.text}")

def get_instagram_posts(username, limit=100):
    #profile = instaloader.Profile.from_username(L.context, username)
    #posts = list(posts)
    return posts[:limit]

import pandas as pd
import ast

def get_keyword_arrays(file_path):
    # Read the Excel file into a DataFrame
    df = pd.read_excel(file_path, usecols=['Keywords'])

    # Iterate through each row and convert the string representation of the list into an actual list
    for keywords in df['Keywords']:
        try:
            # Safely evaluate the string to convert it into a list
            keyword_array = ast.literal_eval(keywords)
            yield keyword_array
        except (ValueError, SyntaxError):
            # Handle cases where the string is not a valid list format
            print(f"Invalid format for keywords: {keywords}")
            yield []


def ai_image_url(prompt):
    print(prompt)
    response = client.images.generate(
    model="dall-e-3",
    prompt=prompt,
    size="1024x1024",
    quality="standard",
    n=1,
    )

    image_url = response.data[0].url
    return image_url
def main():
    total_count, good_count, bad_count = 0,0,0
   # posts_done = ['C_kYH2NJURr', 'C_aMTxcN0Vs', 'C_TJJ_MN2jD', 'C_P35JyyIVW', 'C-_8t2MvV--', 'C-mgOBuR_pU']
    posts_done = []
    posts_failed = []
    #posts = get_instagram_posts(instagram_username)
    #posts = posts['data']['xdt_api__v1__feed__user_timeline_graphql_connection']['edges']

    farmonaut_info = '''
    Farmonaut is a pioneering agricultural technology company that offers advanced, satellite-based farm management solutions. Its mission is to make precision agriculture affordable and accessible to farmers worldwide by integrating innovative technology and data-driven insights into traditional farming practices. Farmonaut's platform provides valuable services such as real-time crop health monitoring, AI-based advisory systems, blockchain-based traceability, and resource management tools. Here's a comprehensive overview of its technologies, business model, target audience, and value proposition.

1. Technologies
Farmonaut leverages a blend of satellite imagery, artificial intelligence (AI), blockchain, and machine learning to address various agricultural challenges.

Satellite-Based Crop Health Monitoring: Farmonaut uses multispectral satellite images to monitor crop health, providing farmers with insights into vegetation health (NDVI), soil moisture levels, and other critical metrics. This data helps farmers make informed decisions about irrigation, fertilizer usage, and pest management, ultimately optimizing crop yields and reducing resource wastage​(
Farmonaut
)​(
Farmonaut
).

Jeevn AI Advisory System: This AI-driven personalized farm advisory tool delivers real-time insights, weather forecasts, and expert crop management strategies to farmers. Jeevn AI analyzes satellite data and other inputs to generate customized advice, improving farm productivity and efficiency​(
Farmonaut
).

Blockchain-Based Product Traceability: By integrating blockchain technology, Farmonaut enables traceability solutions for various industries, particularly agriculture. Blockchain ensures that every stage of the product's journey, from farm to consumer, is transparent and secure, thereby enhancing trust and reducing fraud in supply chains​(
Farmonaut
).

Fleet and Resource Management: Farmonaut provides tools for fleet management, enabling agribusinesses to manage their logistics more efficiently. This helps reduce operational costs by optimizing vehicle usage, ensuring safety, and improving the overall management of agricultural machinery​(
Farmonaut
).

Carbon Footprinting: To help agribusinesses monitor and reduce their environmental impact, Farmonaut offers carbon footprint tracking. This feature provides real-time data on emissions, allowing businesses to take steps towards sustainability and compliance with environmental regulations​(
Farmonaut
)​(
Farmonaut
).

2. Business Model
Farmonaut operates on a subscription-based model, offering various packages that cater to individual farmers, cooperatives, agribusinesses, and government institutions. The company provides flexible pricing tiers depending on the number of hectares monitored and the frequency of satellite data updates. Their platform is accessible through web and mobile apps, making it easy for users to track and manage their farms from anywhere​(
Farmonaut
).

The business also extends its services through APIs, enabling developers and businesses to integrate Farmonaut’s satellite and weather data into their own systems. This API access broadens the company’s revenue streams and encourages collaboration with other agritech companies and agricultural research organizations​(
Farmonaut
).

Additionally, Farmonaut partners with financial institutions to provide satellite-based verification for crop loans and insurance, reducing the likelihood of fraud and improving access to financing for farmers​(
Farmonaut
)​(
Farmonaut
).

3. Target Audience
Farmonaut’s solutions are tailored for a wide range of users in the agricultural ecosystem:

Individual Farmers: The platform offers individual farmers access to precise, real-time data on their crops. This allows them to make better decisions regarding crop health, pest management, and resource allocation. Small and medium-sized farms can particularly benefit from these affordable precision farming solutions​(
Farmonaut
).

Agribusinesses: Large-scale agribusinesses use Farmonaut for plantation management, fleet tracking, and resource optimization. These businesses can manage vast farming operations more efficiently by leveraging satellite monitoring and AI-driven insights​(
Farmonaut
).

Governments and NGOs: Farmonaut works with government agencies and non-governmental organizations (NGOs) to improve agricultural productivity, implement large-scale farm monitoring programs, and support sustainable farming initiatives. Governments also use Farmonaut's tools for crop area and yield estimation, especially in policy and subsidy distribution​(
Farmonaut
).

Financial Institutions: By providing satellite-based verification of farms, Farmonaut helps banks and insurance companies streamline crop loan approvals and reduce fraudulent claims in agricultural insurance​(
Farmonaut
)​(
Farmonaut
).

Corporate Clients: Companies, especially in sectors like textile and food, use Farmonaut's blockchain-based traceability solutions to ensure the authenticity and transparency of their supply chains. This strengthens consumer trust and enhances the brand's reputation​(
Farmonaut
).

4. Value Proposition and Benefits
Farmonaut’s key value propositions include:

Cost-Effective Precision Agriculture: Farmonaut democratizes access to precision agriculture by offering affordable services for real-time crop monitoring and farm management. Unlike traditional precision farming tools that require expensive hardware, Farmonaut relies on satellite imagery, making it a more economical solution for farmers of all scales​(
Farmonaut
).

Increased Farm Productivity: By providing real-time data on crop health, soil moisture, and weather patterns, Farmonaut allows farmers to make informed decisions that optimize their resources. This leads to better crop yields, reduced input costs, and minimized crop losses​(
Farmonaut
).

Sustainability: Through features like carbon footprint tracking and efficient resource management, Farmonaut promotes sustainable farming practices. This is crucial in today’s agriculture, where there is growing pressure to reduce environmental impact while increasing food production​(
Farmonaut
).

Transparency and Trust: Farmonaut’s blockchain-based traceability solution ensures transparency in supply chains, particularly for corporate clients in agriculture and related sectors. By offering verifiable data on product origin and journey, the system helps build consumer trust​(
Farmonaut
).

Access to Financing: Farmonaut's partnerships with financial institutions provide farmers with satellite-based verification for loans and insurance. This improves access to financing while reducing the risks for lenders​(
Farmonaut
)​(
Farmonaut
).

Scalability: The platform is highly scalable, serving clients from smallholder farmers to large agribusinesses and government bodies. Its modular design allows users to choose the services they need and scale them up as their operations grow​(
Farmonaut
).

Conclusion
Farmonaut stands out in the agritech space by offering a comprehensive suite of tools that combine satellite technology, AI, and blockchain to meet the diverse needs of modern agriculture. Whether it's precision farming, supply chain transparency, or sustainability, Farmonaut is at the forefront of the agricultural revolution, making it easier for farmers and agribusinesses to thrive in an increasingly data-driven world. By lowering the cost barrier and providing advanced solutions, Farmonaut continues to empower farmers, improve productivity, and promote sustainable agricultural practices globally.
    '''
    try:
        done_posts = read_array_from_file('posts_done.txt')
    except:
        done_posts = []
    for temp_post in done_posts:
        is_this_a_festival_post = "no"
        if temp_post not in posts_done:
            posts_done.append(temp_post)
   
    post_num = 0
    # Example usage
    file_path = 'clustered_keywords.xlsx'
    for keyword_array in get_keyword_arrays(file_path):
        print(keyword_array)
   
    #for post in posts:
        post_num = post_num + 1
        print('post_num', post_num)
        if post_num > 10:
            try:

                keywords = keyword_array
                keywords_formatted = keywords
                #print(post_data['caption'])
                prompt = f"farmonaut want to publish a high ranking SEO blog comprising of the following keywords: {keyword_array}. Suggest a high ranking SEO optimized title for the blog. Don't output any other text."
                title = call_genai(prompt, 0, 50)
                print(title)
            
                prompt = f"farmonaut want to publish a high ranking SEO blog comprising of the following keywords: {keyword_array}. Suggest a highranking SEO optimized context/summary for this blog in 50 words. Don't output any other text."
                caption = call_genai(prompt, 0, 150)
                print(caption)
                post_data = {
                    'caption': caption,
                    'media': []
                }
                prompt = f"farmonaut want to publish a high ranking SEO blog comprising of the following a. keywords: {keyword_array}, b. title: {title}, c. context: {caption}. Suggest description of two DALL-E generated AI images that should be added to the blog for achieving high search engine ranking. Provide response in the following array format: [image1_description, image2_description] Don't output any other text."
                image_descriptions = call_genai(prompt, 0, 300)
                image_descriptions = string_to_array(image_descriptions)
                print(image_descriptions)
                image_url1 = ai_image_url(image_descriptions[0])
                image_url2 = ai_image_url(image_descriptions[1])
                media_url_arr = [image_url1, image_url2]
                print(media_url_arr)

                media_info = process_media2(media_url_arr, title)
                if media_info:
                    post_data['media'].extend(media_info)
                    post_data['featured_media'] =  media_info[0]['id'] if media_info else None
            
                stripe_html = '''
                <script async src="https://js.stripe.com/v3/pricing-table.js"></script>
        <stripe-pricing-table pricing-table-id="prctbl_1PiWeWK678BipZgWTcLKy62a"
        publishable-key="pk_live_51Oo0Y4K678BipZgWznS6Og83ucZuDEjQtEDEAyR4yoquiI5YaE2G2JQGMDsWm0CbZpajqbSrzqED8DHV4spwBlkO00xwAUs7Bq">
        </stripe-pricing-table>'''
                prompt = f"Generate a HTML formatted very detailed and comprehensive blog post of at least 5000 words with <strong>, <h1>, <h2>, <i>, <u>, <ol>, <ul>, <br>, <p>, <img> blocks wherever necessary in informational tone and as a first-person plural (we, us, our, ours) mandatorily including the following keywords: {keywords} \n\n The blog needs to be at least 5000 words in length. Please don't include any hallucinated information about Farmonaut in the blog. Please don't include any case study in the blog. To add any more details in the blog related to Farmonaut, use information from this text and further elaborate on it if necessary: {farmonaut_info} \n\n Strictly Incorporate these keywords into the blog: {keywords_formatted}. If any of the keywords look unrelated and out of context to the blog, then don't add them to the blog. Add Images (URLs) from this JSON object {post_data['media']} into the blog in <img> blocks wherever necessary including the absolute top of the blog. If it matches with the theme of the blog, add a table of how Farmonaut Satellite System is better than drone and IoT based farm monitoring. Add links to https://farmonaut.com/app_redirect, https://sat.farmonaut.com/api, https://play.google.com/store/apps/details?id=com.farmonaut.android, https://apps.apple.com/in/app/farmonaut/id1489095847, https://farmonaut.com/farmonaut-satellite-weather-api-developer-docs/ wherever necessary. Include this custom HTML code for subscribing to Farmonaut: {stripe_html} \n Add bullet points and subheadings wherever necessary.  Please include an FAQ section as well."
                blog_content = call_genai(prompt, 1, 8000)
                # print(blog_content)

                category_ids = "blogs: 5, case_study:546, area_estimation:542, remote_sensing:9, precision_farming:548, api_development:572, usa:561"
                prompt = f'Based upon this title: {title} , identify the best category id in which this title fits in: {category_ids}. Strictly only return the integer value as the response'
                category_id = call_genai(prompt, 0, 5)
                try:
                    category_id = int(category_id)
                except:
                    category_id = 5
                #publish_to_wordpress(title, blog_content, post_data['media'], post_data['caption'], category_id)
                post_data['title'] = title
                post_data['content'] = blog_content
                post_data['status'] = 'publish'
                post_data['excerpt'] = caption
                post_data['comment_status'] = 'open'
                post_data['categories'] = [category_id]
                
                publish_or_update_wordpress_post(wp_url, wp_username, wp_password, post_data)
                
            except:
                print(traceback.format_exc())
                
            
            remove_all_files_in_folder('insta_files')

            time.sleep(10)
main()