androidlover5842/TrisolarisServer
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Normalize extracted address fields
All checks were successful
build-and-deploy / build-deploy (push) Successful in 32s
Details

Browse Source
This commit is contained in:
androidlover5842
2026-01-31 12:46:48 +05:30
parent 7469d8824b
commit db6ea5d529
1 changed files with 59 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

59
src/main/kotlin/com/android/trisolarisserver/component/DocumentExtractionService.kt
View File

@@ -244,6 +244,7 @@ class DocumentExtractionService(
normalizePinCode(results) normalizePinCode(results)
normalizeIdNumber(results) normalizeIdNumber(results)
markAadhaarIfValid(results) markAadhaarIfValid(results)
normalizeAddress(results)
applyBookingCityUpdates(document, results) applyBookingCityUpdates(document, results)
results["docType"] = computeDocType(results, handled) results["docType"] = computeDocType(results, handled)
applyGuestUpdates(document, propertyId, results) applyGuestUpdates(document, propertyId, results)
@@ -282,6 +283,13 @@ class DocumentExtractionService(
} }
} }
private fun normalizeAddress(results: MutableMap<String, String>) {
val key = DocumentPrompts.ADDRESS.first
val raw = cleanedValue(results[key]) ?: return
val normalized = cleanAddress(raw) ?: return
results[key] = normalized
}
private fun markAadhaarIfValid(results: MutableMap<String, String>) { private fun markAadhaarIfValid(results: MutableMap<String, String>) {
val idKey = DocumentPrompts.ID_NUMBER.first val idKey = DocumentPrompts.ID_NUMBER.first
val digits = normalizeDigits(cleanedValue(results[idKey])) val digits = normalizeDigits(cleanedValue(results[idKey]))
@@ -519,3 +527,54 @@ private fun isValidPin(value: String?): Boolean {
if (value.isNullOrBlank()) return false if (value.isNullOrBlank()) return false
return pinCodeRegex.matches(value) return pinCodeRegex.matches(value)
} }
private fun cleanAddress(raw: String): String? {
val relationRegex = Regex("^\\s*(S/O|D/O|W/O|C/O|H/O|F/O)\\b", RegexOption.IGNORE_CASE)
val prefixRegexes = listOf(
Regex("^\\s*ADDRESS\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*HOUSE\\s*/\\s*BLDG\\.?\\s*/\\s*APT\\.?\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*HOUSE-?BLDG\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*HOUSE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*H\\s*NO\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*HOUSE\\s*NO\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*STREET/ROAD/LANE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*STREET\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*ROAD\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*LANE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*AREA\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*SECTOR\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*COLONY\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*VILLAGE/TOWN/CITY\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*VILLAGE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*TOWN\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*CITY\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*P\\.?\\s*O\\.?\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*POST\\s*OFFICE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*P\\.?\\s*DIST\\.?\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*DISTRICT\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*DIST\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*STATE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*PIN\\s*CODE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*PINCODE\\b[:\\-\\s]*", RegexOption.IGNORE_CASE),
Regex("^\\s*PIN\\b[:\\-\\s]*", RegexOption.IGNORE_CASE)
)
val dropPhrases = setOf(
"address", "addr", "area", "area was", "street/road/lane", "village/town/city", "colony"
)
val parts = raw.replace("\n", ",").split(",")
val cleanedParts = mutableListOf<String>()
for (part in parts) {
var value = part.trim()
if (value.isBlank()) continue
if (relationRegex.containsMatchIn(value)) continue
for (regex in prefixRegexes) {
value = regex.replace(value, "").trim()
}
value = value.replace(Regex("\\s+"), " ").trim()
if (value.isBlank()) continue
if (value.length < 3) continue
if (dropPhrases.contains(value.lowercase())) continue
cleanedParts.add(value)
}
return if (cleanedParts.isEmpty()) null else cleanedParts.joinToString(", ")
}