import language_tool_python  
import re  

class ResumeReviewer:
    def __init__(self):
        self.tool = language_tool_python.LanguageTool('en-US')  

    def grammar_check(self, parsed_resume):
        if not parsed_resume:
            return {}, 3  # ✅ Return empty dictionary and high grammar score

        section_wise_issues = {}
        total_grammar_errors = 0  
        total_words = 0  
        grammar_score = 0  

        ignore_keywords = {"imarticus", "github", "linkedin", "microsoft office", "google drive", "amazon web services"}
        ignore_rule_ids = {"WHITESPACE_RULE", "EN_QUOTES", "CONSECUTIVE_SPACES", "UPPERCASE_SENTENCE_START"}  # ✅ Ignore false uppercase errors

        ignore_patterns = [
            r'\d+',  # Ignore standalone numbers
            r'https?://\S+',  # Ignore URLs
            r'\S+@\S+',  # Ignore emails
            r'^\s*[-•]+\s*',  # ✅ Ignore bullet points
        ]

        name_to_exclude = parsed_resume.get("name", "").strip().lower()

        for section, text in parsed_resume.items():
            if isinstance(text, str) and text.strip():
                words_in_section = len(text.split())  
                total_words += words_in_section

                matches = self.tool.check(text)  
                grammar_issues = []
                spelling_errors = []

                for match in matches:
                    context_text = match.context.lower()

                    # ✅ Ignore rule-based exclusions
                    if match.ruleId in ignore_rule_ids or any(keyword in context_text for keyword in ignore_keywords):
                        continue  
                    if name_to_exclude and name_to_exclude in context_text:
                        continue  
                    if any(re.search(pattern, match.context) for pattern in ignore_patterns):
                        continue  
                    
                    # ✅ Ignore false uppercase errors caused by bullet points
                    if "sentence does not start with an uppercase letter" in match.message.lower() and re.search(r'^\s*[-•]+\s*', match.context):
                        continue  

                    issue = {
                        "context": match.context,
                        "error": match.message,
                        "rule_id": match.ruleId,
                        "suggested_correction": match.replacements
                    }

                    # ✅ Separate grammar and spelling errors
                    if "SPELL" in issue.get("rule_id", ""):
                        spelling_errors.append(issue)
                    else:
                        grammar_issues.append(issue)
                        total_grammar_errors += 1  

                if grammar_issues or spelling_errors:
                    section_wise_issues[section] = {
                        "grammar_issues": grammar_issues,
                        "spelling_errors": spelling_errors
                    }

        # ✅ Calculate mistake percentage
        mistake_percentage = (total_grammar_errors / total_words) * 100 if total_words > 0 else 0

        # ✅ Apply scoring logic from Excel
        if total_grammar_errors == 0:
            grammar_score = 3  
        elif mistake_percentage < 50:
            grammar_score = 2  
        else:
            grammar_score = 1  

        return section_wise_issues, grammar_score  # ✅ Always returns 2 values