utils.py

import os
import fitz  # PyMuPDF
import json
import re
import difflib
import base64  # ✅ added for base64 encoding
from simple_salesforce import Salesforce
from dotenv import load_dotenv
from datetime import datetime  # ✅ added for date conversion

# Load .env variables
load_dotenv()

def get_salesforce_client():
    try:
        sf = Salesforce(
            username=os.getenv("SF_USERNAME"),
            password=os.getenv("SF_PASSWORD"),
            security_token=os.getenv("SF_SECURITY_TOKEN"),
            domain=os.getenv("SF_DOMAIN").replace("https://", "").replace(".salesforce.com", "")
        )
        return sf, None
    except Exception as e:
        return None, str(e)

def get_salesforce_objects(sf):
    try:
        desc = sf.describe()
        object_names = []
        for obj in desc['sobjects']:
            if not obj['deprecatedAndHidden']:  # Skip hidden/deprecated
                object_names.append(obj['name'])
        object_names.sort()
        return object_names, None
    except Exception as e:
        return [], str(e)

def get_object_fields(sf, object_name):
    try:
        metadata = sf.__getattr__(object_name).describe()
        fields = [field['name'] for field in metadata['fields']]
        return fields, None
    except Exception as e:
        return [], str(e)

def extract_text_from_pdf(pdf_path):
    try:
        doc = fitz.open(pdf_path)
        text = ""
        for page in doc:
            text += page.get_text()
        return text.strip(), None
    except Exception as e:
        return None, str(e)

# ✅ Helper: Convert date to YYYY-MM-DD
def convert_to_salesforce_date(date_str):
    for fmt in ("%m/%d/%Y", "%m/%d/%y"):
        try:
            dt = datetime.strptime(date_str, fmt)
            return dt.strftime("%Y-%m-%d")
        except ValueError:
            continue
    return date_str  # Return original if no match

# ✅ Extract key-value pairs smartly from the PDF text
def extract_key_value_pairs(pdf_path):
    try:
        doc = fitz.open(pdf_path)
        text = ""
        for page in doc:
            text += page.get_text()

        keys = []
        values = []

        # Pattern 1: Direct "Key: Value"
        lines = text.splitlines()
        for line in lines:
            match = re.match(r"^([A-Z][A-Za-z0-9 ()/_\-]{3,50})\s*[:\-]\s*(.+)$", line.strip())
            if match:
                key = match.group(1).strip()
                value = match.group(2).strip()
                if len(value) > 1:
                    keys.append(key)
                    values.append(value)

        # Pattern 2: Special extractions

        # Total Agreement Value
        match_val = re.search(r"Total Agreement Value[^\$]*\$\s?([\d,]+(?:\.\d{2})?)", text, re.IGNORECASE)
        if match_val:
            keys.append("Total Agreement Value")
            numeric_value = match_val.group(1).replace(",", "")
            values.append(numeric_value)

        # Agreement Name
        match_name = re.search(r"Agreement\s+(MSA\s+[A-Za-z0-9 _\-]+)", text, re.IGNORECASE)
        if match_name:
            keys.append("Agreement Name")
            values.append(match_name.group(1).strip())

        # Agreement Start and End Dates
        match_dates = re.search(
            r"effective as of\s*([0-9]{1,2}/[0-9]{1,2}/[0-9]{2,4}).*?until\s*[<\(]?([0-9]{1,2}/[0-9]{1,2}/[0-9]{2,4})",
            text, re.IGNORECASE | re.DOTALL
        )
        if match_dates:
            keys.append("Agreement Start Date")
            values.append(convert_to_salesforce_date(match_dates.group(1).strip()))
            keys.append("Agreement End Date")
            values.append(convert_to_salesforce_date(match_dates.group(2).strip()))

        return [{"keys": keys, "values": values}], None
    except Exception as e:
        return None, str(e)

# ✅ FIXED: Use original key as mapping key (not value)
def map_fields(extracted_data, object_fields):
    try:
        mappings = {}
        confidence_scores = {}
        keys = extracted_data[0]["keys"]
        values = extracted_data[0]["values"]

        for key, value in zip(keys, values):
            if key.lower() in ["name", "email"]:
                continue  # Skip these

            best_match = difflib.get_close_matches(key, object_fields, n=1, cutoff=0.0)
            if best_match:
                matched_field = best_match[0]
                confidence = difflib.SequenceMatcher(None, key.lower(), matched_field.lower()).ratio()
                mappings[key] = matched_field
                confidence_scores[key] = round(confidence, 2)
            else:
                mappings[key] = object_fields[0]
                confidence_scores[key] = 0.0

        return mappings, confidence_scores, None
    except Exception as e:
        return None, None, str(e)

def create_record(sf, object_name, data):
    try:
        result = sf.__getattr__(object_name).create(data)
        return result.get("id", "Unknown ID"), None
    except Exception as e:
        return None, str(e)

# ✅ FIXED: Properly encode PDF as base64 string
def attach_pdf(sf, record_id, pdf_path):
    try:
        with open(pdf_path, "rb") as f:
            body = f.read()
        base64_body = base64.b64encode(body).decode("utf-8")

        content_version = sf.ContentVersion.create({
            "Title": os.path.basename(pdf_path),
            "PathOnClient": os.path.basename(pdf_path),
            "VersionData": base64_body
        })

        content_document_id = sf.query(
            f"SELECT ContentDocumentId FROM ContentVersion WHERE Id = '{content_version['id']}'"
        )["records"][0]["ContentDocumentId"]

        sf.ContentDocumentLink.create({
            "ContentDocumentId": content_document_id,
            "LinkedEntityId": record_id,
            "ShareType": "V"
        })

        return "PDF attached successfully", None
    except Exception as e:
        return None, str(e)

def log_failure(pdf_path, object_name, error):
    with open("failures.json", "a") as f:
        json.dump({"pdf": pdf_path, "object": object_name, "error": error}, f)
        f.write("\n")