app.py

import streamlit as st
import pandas as pd
import requests
import base64

def get_backlinks(api_login, api_key, target_url, filters):
    # Encoding credentials
    encoded_credentials = base64.b64encode(f"{api_login}:{api_key}".encode()).decode()

    # Setting headers with Basic Authentication
    headers = {
        'Authorization': f'Basic {encoded_credentials}'
    }

    # Prepare post data
    post_data = {
        0: {
            "target": target_url,
            "limit": 1000,
            "mode": "as_is",
            "filters": filters
        }
    }

    # Making the API request
    response = requests.post("https://api.dataforseo.com/v3/backlinks/backlinks/live", json=post_data, headers=headers)

    response_json = response.json()

    # Log the full response for debugging
    st.text("API Response:")
    st.text(f"Response Status Code: {response.status_code}")
    st.text(f"Response Headers: {response.headers}")
    st.text(f"Response Body: {response_json}")  # Now response_json is defined

    # Check if the response contains 'results' key and handle the JSON structure appropriately
    if response.status_code == 200:
        response_data = response.json()
        
        # Debugging: Print out the keys of the response_data
        st.text(f"Keys in response JSON: {list(response_data.keys())}")

        if 'tasks' in response_data:
            # Assuming there is only one task and one result within each task
            task_result = response_data['tasks'][0]['result']
            if task_result and 'items' in task_result[0]:
                # The actual backlink items are nested within 'items'
                items = task_result[0]['items']
                st.text(f"First few items: {items[:5]}")  # Debugging line to show the items structure
                
                # Convert to DataFrame
                df = pd.json_normalize(items)
                return df
            else:
                st.error("Received empty 'result' from API or missing 'items'.")
                return None
        else:
            st.error(f"No 'tasks' key in response JSON. Full response: {response_data}")
            return None
    else:
        error_message = response.json().get('status_message', 'No specific error message provided')
        st.error(f"Error: Code: {response.status_code} Message: {error_message}")
        return None

def convert_df_to_csv(df):
    # Convert DataFrame to CSV
    return df.to_csv(index=False).encode('utf-8')

# Streamlit layout
st.sidebar.title("DataForSEO API Parameters")
api_login = st.sidebar.text_input("API Login", value="[email protected]")
api_key = st.sidebar.text_input("API Key", type="password")

# Filters input
st.sidebar.title("Filters")
url_from_not_contain = st.sidebar.text_input("URL from does not contain (comma-separated)")
backlink_spam_score = st.sidebar.number_input("Backlink Spam Score ≤", value=20)
page_from_rank_value = st.sidebar.number_input("Page From Rank Value ≥", value=0)
domain_from_rank_value = st.sidebar.number_input("Domain From Rank Value ≥", value=0)
page_from_language = st.sidebar.selectbox("Page From Language", ['en', 'other'])
is_lost = st.sidebar.checkbox("Is Lost", value=False)
dofollow = st.sidebar.checkbox("Dofollow", value=False)
is_broken = st.sidebar.checkbox("Is Broken", value=False)

filters = []

if url_from_not_contain:
    for url in url_from_not_contain.split(','):
        filters.append(["url_from", "not_like", f"%{url.strip()}%"])
        filters.append("and")
        
if is_lost:
    filters.append(["is_lost", "=", is_lost])
    filters.append("and")

if dofollow:
    filters.append(["dofollow", "=", dofollow])
    filters.append("and")

if is_broken:
    filters.append(["is_broken", "=", is_broken])
    filters.append("and")

if page_from_rank_value != 0:
    filters.append(["page_from_rank", ">=", page_from_rank_value])
    filters.append("and")

if domain_from_rank_value != 0:
    filters.append(["domain_from_rank", ">=", domain_from_rank_value])
    filters.append("and")

filters.append(["backlink_spam_score", "<=", backlink_spam_score])
filters.append("and")

filters.append(["page_from_language", "=", page_from_language])
filters.append("and")

filters.append(["anchor", "not_like", ""])  # Assuming empty string represents empty anchor
filters.append("and")

# Filter for domain from rank being greater than or equal to 100
filters.append(["domain_from_rank", ">=", 100])
filters.append("and")

# Remove the last "and" if it's the last element
if filters and filters[-1] == "and":
    filters.pop()

# Output the entire filters section
st.text("Current filters:")
for f in filters:
    st.text(f)

# Main app layout
col1, col2 = st.columns(2)

with col1:
    st.header("Input")
    target_url = st.text_input("Enter the target URL")

generate_button = st.sidebar.button("Generate All")
reset_button = st.sidebar.button("Reset")

df = None

# Generate CSV and download button
if generate_button and target_url:
    df = get_backlinks(api_login, api_key, target_url, filters)
    if df is not None:
        csv = convert_df_to_csv(df)
        st.download_button(
            label="Download data as CSV",
            data=csv,
            file_name='backlinks.csv',
            mime='text/csv',
        )
    else:
        st.error("Failed to generate CSV: No data returned from the API or data processing error.")

# Reset functionality
if reset_button:
    st.experimental_rerun()