chapter-organizer/Core/Parsers/EventOccurrenceParser.cs

using System.Text.RegularExpressions;
using Core.Entities;
using Core.Models;
using EventOccurrenceParsers = Core.Parsers.EventOccurrence;
using Core.Utility;

namespace Core.Parsers;

/// <summary>
/// Result of parsing event occurrence file, containing both occurrences and parsing issues.
/// </summary>
public class EventOccurrenceParserResult
{
    public IDictionary<EventDefinition, List<Entities.EventOccurrence>> Occurrences { get; set; } = new Dictionary<EventDefinition, List<Entities.EventOccurrence>>();
    public List<ParsingIssue> Issues { get; set; } = new();
    public List<string> SkippedHSSectionHeaders { get; set; } = new();
}

public class EventOccurrenceParser 
{
	private FileSystemInfo _txtFile;
	private ICollection<EventDefinition> _events;

	public EventOccurrenceParser(FileSystemInfo txtFile, ICollection<EventDefinition> events)
	{
		_events = events;
		_txtFile = txtFile;
	}

	public EventOccurrenceParserResult Parse()
    {
        var result = new EventOccurrenceParserResult();
        var occurrences = result.Occurrences;
        var issues = result.Issues;
        EventDefinition? currentEventDefinition = null;
        bool inContinuationMode = false;
        bool inHSSection = false;

		var lines = File.ReadLines(_txtFile.FullName);
        foreach (var (line, index) in lines.Select((line, index) => (line, index + 1)))
        {
            // Normalize input: trim and normalize hyphens (en-dash, em-dash -> regular hyphen)
            // This allows the grammar parser to assume normalized input
            var normalizedLine = TextUtil.SanitizeInput(line.Trim());
            
            // Skip empty lines
            if (EventOccurrenceParsers.LineClassifier.IsEmptyLine(normalizedLine))
            {
                // Empty lines break continuation mode
                inContinuationMode = false;
                continue;
            }

            // Skip comment lines (starting with "#") - use grammar parser
            if (EventOccurrenceParsers.LineClassifier.IsCommentLine(normalizedLine))
            {
                // Comment lines break continuation mode
                inContinuationMode = false;
                continue;
            }

			// Try to parse occurrence line using grammar parser
			var occurrenceLine = EventOccurrenceGrammar.TryParseOccurrenceLine(normalizedLine);
			if (!occurrenceLine.HasValue)
			{
				// Not an occurrence line, try other line types
				// Try to parse section header using grammar parser
				var sectionHeader = EventOccurrenceGrammar.TryParseSectionHeader(normalizedLine);
				if (sectionHeader.HasValue)
				{
					var (eventNamePart, schoolLevel) = sectionHeader.Value;
					
					// Section headers break continuation mode
					inContinuationMode = false;
					
					// Check if this is an HS event - if so, skip gracefully regardless of whether it matches
					// This prevents HS events from being incorrectly associated with MS events (e.g., 
					// "Biotechnology Design – HS" matching "Biotechnology" MS event)
					if (schoolLevel.Equals("HS", StringComparison.OrdinalIgnoreCase))
					{
						result.SkippedHSSectionHeaders.Add(normalizedLine);
						currentEventDefinition = null; // Skip subsequent occurrences
						inHSSection = true; // Mark that we're in an HS section
						continue; // No issue created
					}
					
					// For MS events, use fuzzy matching to find the best matching event definition
					var evt = EventOccurrenceParsers.SectionHeaderMatcher.MatchEventDefinition(eventNamePart, _events);
					if (evt == null)
					{
						// For unmatched MS headers, create issue
						var bestRatio = EventOccurrenceParsers.SectionHeaderMatcher.GetBestMatchRatio(eventNamePart, _events);
						issues.Add(new ParsingIssue
						{
							LineNumber = index,
							LineContent = normalizedLine,
							IssueType = ParsingIssueType.UnmatchedLine,
							Message = $"Section header '{eventNamePart} - {schoolLevel}' found but no matching event definition (best match ratio: {bestRatio})"
						});
						continue;
					}
					currentEventDefinition = evt;
					inHSSection = false; // Reset HS section flag for MS events
					continue;
				}
				
				// Check for General Schedule/Session using grammar parser
				if (EventOccurrenceParsers.SectionHeaderMatcher.IsGeneralSchedule(normalizedLine))
				{
					// General schedule breaks continuation mode
					inContinuationMode = false;
					inHSSection = false; // Reset HS section flag
					currentEventDefinition = EventDefinition.GeneralSchedule;
					continue;
				}
				
				// Also check for simple "MS" or "HS" in line (backward compatibility)
				if (EventOccurrenceParsers.SectionHeaderMatcher.HasSchoolLevel(normalizedLine))
				{
					// Section headers break continuation mode
					inContinuationMode = false;
					
					// Check if this is an HS event - if so, skip gracefully regardless of whether it matches
					// This prevents HS events from being incorrectly associated with MS events
					if (normalizedLine.Contains("HS", StringComparison.OrdinalIgnoreCase))
					{
						result.SkippedHSSectionHeaders.Add(normalizedLine);
						currentEventDefinition = null; // Skip subsequent occurrences
						inHSSection = true; // Mark that we're in an HS section
						continue; // No issue created
					}
					
					// For MS events, use fuzzy matching to find the best matching event definition
					var evt = EventOccurrenceParsers.SectionHeaderMatcher.MatchEventDefinition(normalizedLine, _events);
					if (evt == null)
					{
						// For unmatched MS headers, create issue
						var bestRatio = EventOccurrenceParsers.SectionHeaderMatcher.GetBestMatchRatio(normalizedLine, _events);
						issues.Add(new ParsingIssue
						{
							LineNumber = index,
							LineContent = normalizedLine,
							IssueType = ParsingIssueType.UnmatchedLine,
							Message = $"Section header with 'MS' or 'HS' found but no matching event definition (best match ratio: {bestRatio})"
						});
						continue;
					}
					currentEventDefinition = evt;
					inHSSection = false; // Reset HS section flag for MS events
					continue;
				}

				// Check if line starts with "*" to enter continuation mode
				if (normalizedLine.TrimStart().StartsWith("*", StringComparison.Ordinal))
				{
					inContinuationMode = true;
				}

				// Skip continuation lines (in continuation mode OR line starts with "*" or is parenthetical)
				if (inContinuationMode || EventOccurrenceParsers.LineClassifier.IsContinuationLine(normalizedLine))
				{
					continue;
				}

				// "Voting Delegates" section header is no longer used - occurrences are categorized by name pattern
				// Track as unmatched line if it's not empty
				if (!string.IsNullOrWhiteSpace(normalizedLine))
				{
					issues.Add(new ParsingIssue
					{
						LineNumber = index,
						LineContent = normalizedLine,
						IssueType = ParsingIssueType.UnmatchedLine,
						Message = "Line does not match expected format (Name Month Day Time/Location)"
					});
				}
				continue;
			}

			// Occurrence lines break continuation mode
			inContinuationMode = false;

			// Skip occurrences under HS sections (they won't match any event definition)
			if (inHSSection)
			{
				continue;
			}

			var (occurrenceName, month, dayOfMonthStr, timeAndLocation) = occurrenceLine.Value;

			// Remove weekday suffix from occurrence name if present
			occurrenceName = Regex.Replace(occurrenceName,
				@"(?<Weekday>Monday|Tuesday|Wednesday|Thursday|Friday|Saturday|Sunday),\s?$", "").Trim();

			// Determine event definition based on occurrence name pattern or current section
			EventDefinition? eventDefinition = EventOccurrenceParsers.EventDefinitionResolver.Resolve(occurrenceName, currentEventDefinition);

			// Track issue if we can't determine the event definition
			if (eventDefinition == null)
			{
				issues.Add(new ParsingIssue
				{
					LineNumber = index,
					LineContent = normalizedLine,
					IssueType = ParsingIssueType.MissingEventDefinition,
					Message = $"Cannot determine event definition for occurrence: {occurrenceName}"
				});
				continue;
			}

			// timeAndLocation is already normalized (hyphens normalized) since normalizedLine was sanitized
			
			// Parse time and location - extract time using regex, then use everything after time as location
			EventOccurrenceParsers.TimeLocationParser.Parse(timeAndLocation, out string time, out string location);

			// Parse date
			DateOnly? startDate = null;
			try
			{
				startDate = TextUtil.ParseDate(month, dayOfMonthStr.ToString(), DateTime.Now.Year);
			}
			catch (Exception ex)
			{
				issues.Add(new ParsingIssue
				{
					LineNumber = index,
					LineContent = normalizedLine,
					IssueType = ParsingIssueType.DateParseFailure,
					Message = $"Failed to parse date: {ex.Message}"
				});
				continue;
			}

			// Parse time
			TimeOnly? startTime = null;
			try
			{
				startTime = EventOccurrenceParsers.TimeParser.Parse(time);
			}
			catch (Exception ex)
			{
				issues.Add(new ParsingIssue
				{
					LineNumber = index,
					LineContent = normalizedLine,
					IssueType = ParsingIssueType.TimeParseFailure,
					Message = $"Failed to parse time '{time}': {ex.Message}"
				});
				continue;
			}

			if (startDate == null || startTime == null)
				continue;

			var t = new DateTime(startDate.Value, startTime.Value);

			var eventOccurrence = new Core.Entities.EventOccurrence
			{
				Name = occurrenceName, 
				StartTime = t, 
				Time = $"{time}", 
				Date = $"{month} {dayOfMonthStr}",
				Location = location
			};

			if (!occurrences.ContainsKey(eventDefinition))
				occurrences.Add(eventDefinition, []);
			occurrences[eventDefinition].Add(eventOccurrence);
			
			// Reset HS section flag when we successfully parse an occurrence (means we're in a valid section)
			inHSSection = false;
        }

        return result;
    }
}