Roadbound-BRR/lib/parsers/arriva_schedule_parser.dart

471 lines
14 KiB
Dart

import "dart:typed_data";
import "package:docx_to_text/docx_to_text.dart";
import "../models/operations/trip.dart";
import "../exceptions/schedule_parse_exception.dart";
import "schedule_parser.dart";
class ArrivaScheduleParser implements ScheduleParser {
// OUTBOUND FORMAT: HHMM_HHMM times... EC### duty running trip
// Times appear on the LEFT, trip info on the RIGHT
static final _outboundPattern = RegExp(
r"^(\d{4})_(\d{4})\s+(.*?)\s+(EC\d+)\s+(\d+)\s+([NRF]?)(\d+)\s+(\d+)",
);
static final _outboundFinishingPattern = RegExp(
r"^(\d{4})_(\d{4})\s+(.*?)\s+(EC\d+)\s+(\d+)\s+F",
);
// INBOUND FORMAT: trip duty running EC### ... HHMM_HHMM times...
// Trip info appears on the LEFT, times on the RIGHT
// Note: running number may have N/R/F prefix like "N503" or be separate like "N 503"
static final _inboundPattern = RegExp(
r"^(\d+)\s+(\d+)\s+([NRF]?)(\d+)\s+(EC\d+)\s+.*?(\d{4})_(\d{4})\s+([\d\s]+)",
);
static final _inboundFinishingPattern = RegExp(
r"^(\d+)\s+(\d+)\s+F\s+(EC\d+)\s+.*?(\d{4})_(\d{4})(.*)",
);
@override
bool canParse(String content) {
// Check if content looks like Arriva format
return content.contains("EC5") &&
content.contains("FINISH AT POINT") &&
RegExp(r"\d{4}_\d{4}").hasMatch(content);
}
@override
Future<List<Trip>> parseBytes(Uint8List bytes) async {
// Step 1: Extract text
final text = _extractTextFromDocx(bytes);
// DEBUG: Print extracted text
print("=== EXTRACTED TEXT START ===");
print(text);
print("=== EXTRACTED TEXT END ===");
// Step 2: Parse document structure (headers and trips)
final lines = text.split("\n");
final documentSections = _parseDocumentSections(lines);
print("=== FOUND ${documentSections.length} SECTIONS ===");
for (var section in documentSections) {
print(
"Section: ${section.direction}, ${section.stations.length} stations, ${section.tripLines.length} trips",
);
}
if (documentSections.isEmpty) {
throw ScheduleParseException("No trip data found in schedule");
}
// Step 3: Parse trips from all sections
final trips = <Trip>[];
for (var section in documentSections) {
final sectionTrips = _parseSectionTrips(section);
trips.addAll(sectionTrips);
print(
"✓ Parsed ${sectionTrips.length} trips from ${section.direction} section",
);
}
// Step 4: Sort by scheduled time
trips.sort((a, b) => a.scheduledTime.compareTo(b.scheduledTime));
return trips;
}
String _extractTextFromDocx(Uint8List bytes) {
try {
return docxToText(bytes);
} catch (e, stackTrace) {
print("Arriva parser document read failed: $e");
print(stackTrace);
throw ScheduleParseException("Failed to read document: $e");
}
}
String _formatTime(String rawTime) {
if (rawTime.length != 4) {
throw FormatException("Invalid time format: $rawTime");
}
return "${rawTime.substring(0, 2)}:${rawTime.substring(2, 4)}";
}
List<_DocumentSection> _parseDocumentSections(List<String> lines) {
final sections = <_DocumentSection>[];
_DocumentSection? currentSection;
for (var i = 0; i < lines.length; i++) {
final line = lines[i].trim();
// Check if this is a station header line
final stations = _extractStationHeader(line);
if (stations != null && stations.isNotEmpty) {
// Save previous section if it exists
if (currentSection != null && currentSection.tripLines.isNotEmpty) {
sections.add(currentSection);
}
// Direction is determined later from the first trip line
currentSection = _DocumentSection(
stations: stations,
direction: "unknown",
tripLines: [],
);
print(
"Found station header at line $i with ${stations.length} stations",
);
print(
" Stations: ${stations.take(3).join(", ")} ... ${stations.skip(stations.length - 2).join(", ")}",
);
continue;
}
// Check if this is a trip line
if (currentSection != null && _isTripLine(line)) {
// Infer direction from first trip line if not yet set
if (currentSection.direction == "unknown") {
currentSection.direction = _inferDirectionFromTripLine(line);
print(" Direction inferred: ${currentSection.direction}");
}
currentSection.tripLines.add(line);
}
}
// Add final section
if (currentSection != null && currentSection.tripLines.isNotEmpty) {
sections.add(currentSection);
}
return sections;
}
List<String>? _extractStationHeader(String line) {
// Station headers have multiple short uppercase codes, no digits, no underscores
if (line.contains(RegExp(r"\d")) ||
line.contains("EC") ||
line.contains("_") ||
line.length < 10) {
return null;
}
// Split by whitespace and filter for potential station codes (3-8 uppercase letters)
final parts = line.split(RegExp(r"\s+"));
final potentialStations = parts
.where(
(part) =>
part.length >= 3 &&
part.length <= 8 &&
RegExp(r"^[A-Z]+$").hasMatch(part),
)
.toList();
if (potentialStations.length < 8) return null;
// Filter out common metadata words that appear in headers
const nonStationWords = {
"TRP",
"DUTY",
"BUS",
"START",
"END",
"GAR",
"DEP",
"ARR",
"DENOTES",
"FINISHES",
"RELIEF",
"TRIP",
"NEXT",
"NO",
"AT",
"SPELL",
"HOURS",
"TOTAL",
"LAYOVER",
"MILES",
"LIVE",
"DEAD",
"MILEAGE",
"TIME",
"SIGN",
"FORM",
"NXT",
"THIS",
"HAS",
"OR",
"FOR",
"CHANGE",
"SERVICE",
"POINT",
"LSN",
"MAN",
"RUI",
"SN",
"ROUTE",
"RUNNING",
"PREV",
"FIN",
"ENTOD",
"SOALL",
"USHRS",
"ADDTL",
"CASH",
"TODAYS",
"REL",
"IEF",
};
final stations = potentialStations
.where((s) => !nonStationWords.contains(s))
.toList();
// Need at least 5 actual station-like codes remaining - the structural
// density of codes is what marks this as a station header, not a known list
return stations.length >= 5 ? stations : null;
}
// Determine direction from the format of the first trip line.
// Lines starting with digits are outbound (trip number comes first).
// Lines starting with underscores or a bare time are inbound (times come first).
String _inferDirectionFromTripLine(String line) {
if (RegExp(r"^\d{4}_\d{4}").hasMatch(line)) return "inbound";
if (RegExp(r"^_+\d{4}").hasMatch(line)) return "inbound";
if (RegExp(r"^\d+\s+\d+").hasMatch(line)) return "outbound";
return "outbound";
}
bool _isTripLine(String line) {
return RegExp(r"\d{4}_\d{4}").hasMatch(line) && line.contains("EC");
}
List<Trip> _parseSectionTrips(_DocumentSection section) {
final trips = <Trip>[];
for (final line in section.tripLines) {
Trip? trip;
// Detect actual line format by looking at structure
// Inbound: starts with numbers (trip duty running EC###) or (trip duty F EC###)
// Note: running number might be "N503 EC" (with spaces) or "N 503 EC" or just "503 EC"
// Outbound: starts with HHMM_HHMM
final isOutboundFormat = RegExp(r"^\d{4}_\d{4}").hasMatch(line);
final isInboundFormat = RegExp(
r"^\d+\s+\d+\s+(?:[NRF]\d+\s+|[NRF]\s+\d+\s+|F\s+|\d+\s+)EC",
).hasMatch(line);
if (isOutboundFormat) {
trip = _parseOutboundTrip(line, section.stations);
} else if (isInboundFormat) {
trip = _parseInboundTrip(line, section.stations);
}
if (trip != null) {
trips.add(trip);
} else {
final format = isOutboundFormat
? "outbound"
: isInboundFormat
? "inbound"
: "unknown";
print(
"Failed to parse $format line: ${line.substring(0, line.length > 80 ? 80 : line.length)}...",
);
}
}
return trips;
}
Trip? _parseInboundTrip(String line, List<String> stations) {
// INBOUND: trip duty running EC### ... HHMM_HHMM times...
var match = _inboundPattern.firstMatch(line);
if (match != null) {
final tripNumber = match.group(1)!;
final dutyNumber = match.group(2)!;
final tripType = match.group(3) ?? "";
final busWorkNumber = match.group(4)!;
final firstTime = match.group(6)!;
final secondTime = match.group(7)!;
final timesString = match.group(8) ?? "";
// Build complete time array: first_time, second_time, then remaining times
final times = [firstTime, secondTime];
final additionalTimes = _extractTimesFromString(timesString);
times.addAll(additionalTimes);
final stationTimes = _mapStationsToTimes(stations, times);
final scheduledTime = _formatTime(firstTime);
return Trip(
tripNumber: tripNumber,
dutyNumber: dutyNumber,
busWorkNumber: busWorkNumber,
scheduledTime: scheduledTime,
tripType: tripType,
isFinishing: false,
stationTimes: stationTimes,
stationOrder: stations,
direction: (int.tryParse(tripNumber) ?? 0).isOdd
? "inbound"
: "outbound",
);
}
// Try finishing pattern
match = _inboundFinishingPattern.firstMatch(line);
if (match != null) {
final tripNumber = match.group(1)!;
final dutyNumber = match.group(2)!;
final firstTime = match.group(4)!;
final secondTime = match.group(5)!;
final timesString = match.group(6) ?? "";
final times = [firstTime, secondTime];
final additionalTimes = _extractTimesFromString(timesString);
times.addAll(additionalTimes);
final stationTimes = _mapStationsToTimes(stations, times);
final scheduledTime = _formatTime(firstTime);
return Trip(
tripNumber: tripNumber,
dutyNumber: dutyNumber,
busWorkNumber: dutyNumber,
scheduledTime: scheduledTime,
tripType: "F",
isFinishing: true,
stationTimes: stationTimes,
stationOrder: stations,
direction: (int.tryParse(tripNumber) ?? 0).isOdd
? "inbound"
: "outbound",
);
}
return null;
}
Trip? _parseOutboundTrip(String line, List<String> stations) {
// OUTBOUND: HHMM_HHMM times... EC### duty running trip
var match = _outboundPattern.firstMatch(line);
if (match != null) {
final firstTime = match.group(1)!;
final secondTime = match.group(2)!;
final timesString = match.group(3) ?? "";
final dutyNumber = match.group(5)!;
final tripType = match.group(6) ?? "";
final busWorkNumber = match.group(7)!;
final tripNumber = match.group(8)!;
// Build complete time array: first_time, second_time, then remaining times
final times = [firstTime, secondTime];
times.addAll(_extractTimesFromString(timesString));
final stationTimes = _mapStationsToTimes(stations, times);
final scheduledTime = _formatTime(firstTime);
return Trip(
tripNumber: tripNumber,
dutyNumber: dutyNumber,
busWorkNumber: busWorkNumber,
scheduledTime: scheduledTime,
tripType: tripType,
isFinishing: false,
stationTimes: stationTimes,
stationOrder: stations,
direction: (int.tryParse(tripNumber) ?? 0).isOdd
? "inbound"
: "outbound",
);
}
// Try finishing pattern
match = _outboundFinishingPattern.firstMatch(line);
if (match != null) {
final firstTime = match.group(1)!;
final secondTime = match.group(2)!;
final timesString = match.group(3) ?? "";
final dutyNumber = match.group(5)!;
final times = [firstTime, secondTime];
times.addAll(_extractTimesFromString(timesString));
final stationTimes = _mapStationsToTimes(stations, times);
final scheduledTime = _formatTime(firstTime);
return Trip(
tripNumber:
dutyNumber, // Finishing trips may not have separate trip number
dutyNumber: dutyNumber,
busWorkNumber: dutyNumber,
scheduledTime: scheduledTime,
tripType: "F",
isFinishing: true,
stationTimes: stationTimes,
stationOrder: stations,
direction: (int.tryParse(dutyNumber) ?? 0).isOdd
? "inbound"
: "outbound",
);
}
return null;
}
List<String> _extractTimesFromString(String timesString) {
// Extract all 4-digit times from the string
final pattern = RegExp(r"\b(\d{4})\b");
return pattern.allMatches(timesString).map((m) => m.group(1)!).toList();
}
List<String> _extractAllTimes(String line) {
// Extract all 4-digit times, including those in HHMM_HHMM format
final timePattern = RegExp(r"\b(\d{4})(?:_(\d{4}))?\b");
final matches = timePattern.allMatches(line);
final times = <String>[];
for (final match in matches) {
// Add first time
times.add(match.group(1)!);
// Add second time if it exists (from HHMM_HHMM)
if (match.group(2) != null) {
times.add(match.group(2)!);
}
}
return times;
}
Map<String, String> _mapStationsToTimes(
List<String> stations,
List<String> times,
) {
final stationTimes = <String, String>{};
for (var i = 0; i < stations.length && i < times.length; i++) {
final time = times[i];
// Only add non-empty times (not "____" or similar)
if (RegExp(r"^\d{4}$").hasMatch(time)) {
stationTimes[stations[i]] = _formatTime(time);
}
}
return stationTimes;
}
}
class _DocumentSection {
final List<String> stations;
String direction;
final List<String> tripLines;
_DocumentSection({
required this.stations,
required this.direction,
required this.tripLines,
});
}