ocr extract
This commit is contained in:
Claude Agent
@@ -50,6 +50,9 @@ class BaseStoreProfile(ABC):
|
||||
# Store display name
|
||||
STORE_NAME: str = "Unknown Store"
|
||||
|
||||
# Flag for known non-VAT payer stores (skips TVA extraction)
|
||||
IS_NON_VAT_PAYER: bool = False
|
||||
|
||||
# -------------------------------------------------------------------------
|
||||
# Generic patterns - can be overridden in subclasses
|
||||
# -------------------------------------------------------------------------
|
||||
@@ -100,18 +103,33 @@ class BaseStoreProfile(ABC):
|
||||
]
|
||||
|
||||
# Payment method patterns (pattern, method_type, confidence)
|
||||
# Handles ALL payment types: CARD, NUMERAR, and card brand names
|
||||
PAYMENT_PATTERNS = [
|
||||
# CARTE CREDIT variants (OMV/Petrom/Socar receipts)
|
||||
(r'CARTE\s+CREDIT\s*:?\s*([\d\s.,]+)', 'CARD', 0.98),
|
||||
(r'CARTE\s+CREDIT\s*:?\s*\n\s*([\d\s.,]+)', 'CARD', 0.97),
|
||||
(r'CARTE\s+DE\s+CREDIT\s*:?\s*([\d\s.,]+)', 'CARD', 0.98),
|
||||
(r'CARTE\s+DE\s+CREDIT\s*:?\s*\n\s*([\d\s.,]+)', 'CARD', 0.97),
|
||||
# CARD standard
|
||||
(r'(?:PLATA\s+)?CARD\s*[:\sA-Z]?\s*([\d\s.,]+)', 'CARD', 0.95),
|
||||
# Card brand names
|
||||
(r'VISA\s*:?\s*([\d\s.,]+)', 'CARD', 0.95),
|
||||
(r'MASTERCARD\s*:?\s*([\d\s.,]+)', 'CARD', 0.95),
|
||||
(r'MAESTR[O0]\s*:?\s*([\d\s.,]+)', 'CARD', 0.95),
|
||||
(r'CONTACTLESS\s*:?\s*([\d\s.,]+)', 'CARD', 0.90),
|
||||
(r'DEBIT\s*:?\s*([\d\s.,]+)', 'CARD', 0.90),
|
||||
(r'CREDIT\s*:?\s*([\d\s.,]+)', 'CARD', 0.88),
|
||||
# Cash variants
|
||||
(r'NUMERAR\s*:?\s*([\d\s.,]+)', 'NUMERAR', 0.95),
|
||||
(r'CASH\s*:?\s*([\d\s.,]+)', 'NUMERAR', 0.90),
|
||||
# Truncation recovery patterns (for OCR left-margin issues)
|
||||
(r'(?:^|\n|\s)RD\s*:?\s*(\d{1,6}[.,]\d{2})\b', 'CARD', 0.70),
|
||||
(r'(?:^|\n|\s)ARD\s*:?\s*(\d{1,6}[.,]\d{2})\b', 'CARD', 0.75),
|
||||
(r'(?:^|\n|\s)MERAR\s*:?\s*(\d{1,6}[.,]\d{2})\b', 'NUMERAR', 0.70),
|
||||
]
|
||||
|
||||
# Client section markers (for B2B receipts) - More flexible patterns
|
||||
# Includes OCR corruption variants (LIENT, C IENT, L IENT)
|
||||
CLIENT_MARKERS = [
|
||||
r'C\.?\s*[I1]\.?\s*F\.?\s+CLIENT', # "CIF CLIENT" (with or without colon)
|
||||
r'C\.?\s*U\.?\s*[I1]\.?\s+CLIENT', # "CUI CLIENT"
|
||||
@@ -121,24 +139,62 @@ class BaseStoreProfile(ABC):
|
||||
r'BENEFICIAR\s*:', # "BENEFICIAR:"
|
||||
r'CUMP[AĂ]R[AĂ]TOR', # "CUMPARATOR" without colon
|
||||
r'COD\s+FISCAL\s+CLIENT', # "COD FISCAL CLIENT"
|
||||
# OCR corruption patterns
|
||||
r'C[I1]F\s+[A-Z\s]{0,6}IENT\s*:', # "CIF a IENT:", "CIF CL IENT:", "CIF L IENT:"
|
||||
r'C[I1]F\s+LIENT\s*:', # "CIF LIENT:" (missing C)
|
||||
r'LIENT\s*:', # "LIENT:" (missing C and I/L)
|
||||
# Brick-specific (I→L OCR error)
|
||||
r'CLIENT\s+C\.?U\.?[LI1]\.?\s*/', # "CLIENT C.U.L./" (I read as L)
|
||||
]
|
||||
|
||||
# Client CUI patterns (pattern, confidence) - More flexible
|
||||
# Client CUI patterns (pattern, confidence) - Comprehensive
|
||||
# Handles: docTR reordering, doubled letters, corruption, CUMPARATOR, Brick L/I swap
|
||||
CLIENT_CUI_PATTERNS = [
|
||||
# "CIF CLIENT:XXXXXXX" or "CIF CLIENT: ROXXXXXXX" - most common format
|
||||
(r'C\.?[I1]\.?F\.?\s+CLIENT\s*:?\s*(R?[O0]?\s*\d{6,10})', 0.99),
|
||||
# "CLIENT CIF: XXXXXXX"
|
||||
(r'CLIENT\s+C\.?[I1]\.?F\.?\s*:?\s*(R?[O0]?\s*\d{6,10})', 0.98),
|
||||
# "CUI CLIENT: XXXXXXX"
|
||||
(r'C\.?U\.?[I1]\.?\s+CLIENT\s*:?\s*(R?[O0]?\s*\d{6,10})', 0.98),
|
||||
# "ROXXXXXXX" followed by CLIENT marker
|
||||
(r'(R[O0]\d{6,10})\s*\n\s*CLIENT', 0.97),
|
||||
# "C.I.F. CLIENT: XXXXXXX"
|
||||
(r'C\.?\s*I\.?\s*F\.?\s+CLIENT\s*:?\s*(R?[O0]?\s*\d{6,10})', 0.96),
|
||||
# "CLIENT: ROXXXXXXX" or "CLIENT: XXXXXXX"
|
||||
# === CUI on line BEFORE CLIENT marker (docTR/OCR reordering) ===
|
||||
(r'(R[O0]\d{6,10})\s*\n\s*CLIENT\s+C\.?\s*U\.?\s*[I1]\.?', 0.99),
|
||||
(r'(R[O0]\d{6,10})\s*\n\s*CLIENT\s+C\.?\s*[I1]\.?\s*F\.?', 0.99),
|
||||
(r'(R[O0]\d{6,10})\s*:?\s*\n\s*CLIENT', 0.98),
|
||||
# === "CIF I CLIENT:" format (OCR extra chars) ===
|
||||
(r'C[I1]F\s+[A-Z]*\s*CLIENT\s*:?\s*(R[O0]\d{6,10})', 0.98),
|
||||
(r'C[I1]F\s+[A-Z]*\s*CLIENT\s*:?\s*(R[O0]?\d{6,10})', 0.97),
|
||||
# === CIF CLIENT: (reversed - CIF before CLIENT) ===
|
||||
(r'C\.?\s*[I1]\.?\s*F\.?\s+CLIENT\s*:?\s*(R[O0]?\d{6,10})', 0.98),
|
||||
(r'C\.?\s*U\.?\s*[I1]\.?\s+CLIENT\s*:?\s*(R[O0]?\d{6,10})', 0.98),
|
||||
(r'C\.?\s*[I1]\.?\s*F\.?\s+CLIENT\s*:?\s*(?:R[O0])?(\d{6,10})', 0.98),
|
||||
(r'C\.?\s*U\.?\s*[I1]\.?\s+CLIENT\s*:?\s*(?:R[O0])?(\d{6,10})', 0.98),
|
||||
# === CLIENT C.U.I/C.I.F. (slash variants) ===
|
||||
(r'CLIENT\s+C\.?\s*U\.?\s*[I1]\.?\s*/\s*C\.?\s*[I1]\.?\s*F\.?\s*:?\s*(R[O0]?\d{6,10})', 0.97),
|
||||
(r'CLIENT\s+C\.?\s*U\.?\s*[I1]\.?\s*/\s*C\.?\s*[I1]\.?\s*F\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.97),
|
||||
# === Doubled letters (docTR artifact: "C.U U.I") ===
|
||||
(r'CLIENT\s+C\.?\s*U\.?\s*U?\.?\s*[I1]\.?\s*/\s*C\.?\s*[I1]\.?\s*F\.?\s*:?\s*(R[O0]?\d{6,10})', 0.96),
|
||||
(r'CLIENT\s+C\.?\s*U\.?\s*U?\.?\s*[I1]\.?\s*/\s*C\.?\s*[I1]\.?\s*F\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.96),
|
||||
# === CLIENT C.U.I. or CLIENT CUI (without slash) ===
|
||||
(r'CLIENT\s+C\.?\s*U\.?\s*[I1]\.?\s*:?\s*(R[O0]?\d{6,10})', 0.96),
|
||||
(r'CLIENT\s+C\.?\s*U\.?\s*[I1]\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.96),
|
||||
(r'CLIENT\s+C\.?\s*[I1]\.?\s*F\.?\s*:?\s*(R[O0]?\d{6,10})', 0.96),
|
||||
(r'CLIENT\s+C\.?\s*[I1]\.?\s*F\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.96),
|
||||
# === Corrupted CLIENT after CIF (OCR errors) ===
|
||||
(r'CIF\s+[a-zA-Z\s]{2,8}IENT\s*:?\s*(R[O0]?\d{6,10})', 0.93),
|
||||
(r'CIF\s+[a-zA-Z\s]{2,8}IENT\s*:?\s*(?:R[O0])?(\d{6,10})', 0.93),
|
||||
(r'CIF\s+LIENT\s*:?\s*(R[O0]?\d{6,10})', 0.92),
|
||||
(r'CIF\s+LIENT\s*:?\s*(?:R[O0])?(\d{6,10})', 0.92),
|
||||
# === CUMPARATOR variants ===
|
||||
(r'CUMPARATOR\s+C\.?\s*U\.?\s*[I1]\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.95),
|
||||
(r'CUMPARATOR\s+C\.?\s*[I1]\.?\s*F\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.95),
|
||||
# CUMPARATOR with CUI/CIF on next line
|
||||
(r'CUMPARATOR\s*:.*\n\s*C\.?\s*U\.?\s*[I1]\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.93),
|
||||
(r'CUMPARATOR\s*:.*\n\s*C\.?\s*[I1]\.?\s*[FT]\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.93),
|
||||
# CUMPARATOR with CUI/CIF two lines down
|
||||
(r'CUMPARATOR\s*:.*\n.*\n\s*C\.?\s*[I1]\.?\s*[FT]\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.90),
|
||||
# === CLIENT on next line ===
|
||||
(r'CLIENT\s*:\s*\n\s*C\.?\s*U\.?\s*[I1]\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.95),
|
||||
(r'CLIENT\s*:.*\n\s*C\.?\s*U\.?\s*[I1]\.?\s*:?\s*(?:R[O0])?(\d{6,10})', 0.90),
|
||||
# === Standard fallback patterns ===
|
||||
(r'CLIENT\s*:?\s*(R?[O0]?\s*\d{6,10})', 0.90),
|
||||
# "COD FISCAL CLIENT: XXXXXXX"
|
||||
(r'COD\s+FISCAL\s+CLIENT\s*:?\s*(R?[O0]?\s*\d{6,10})', 0.95),
|
||||
# === Brick-specific (I→L OCR error) ===
|
||||
# Matches: "CLIENT C.U.L./C.IF. :R01879855"
|
||||
(r'CLIENT\s+C\.?U\.?[LI1]\.?\s*/\s*C\.?[LI1]\.?F\.?\s*:?\s*(R?O?\d{6,10})', 0.99),
|
||||
]
|
||||
|
||||
# Company type indicators (for identifying company names)
|
||||
@@ -158,15 +214,133 @@ class BaseStoreProfile(ABC):
|
||||
# Maximum reasonable payment amount (to filter OCR errors)
|
||||
MAX_PAYMENT = Decimal('100000')
|
||||
|
||||
# -------------------------------------------------------------------------
|
||||
# TVA (VAT) patterns - ALL FORMATS unified
|
||||
# OCR tolerant: T[VU][AR] matches TVA, TUA, TVR
|
||||
# -------------------------------------------------------------------------
|
||||
TVA_PATTERNS = [
|
||||
# === FORMAT 1: INLINE cu cod și procent (Lidl-style) ===
|
||||
# "TVA A 21,00% 7.71" sau "TVA B 11,00% 2.13"
|
||||
(r'T[VU][AR]\s+([A-D])\s+(\d{1,2})[.,]?\d{0,2}\s*%\s+([\d.,]+)', 0.98, 'inline'),
|
||||
(r'T[VU][AR]\s+([A-D])\s+\\?(\d{1,2})[.,]?\d{0,2}\s*%\s+([\d.,]+)', 0.97, 'inline'),
|
||||
(r'IVA\s+([A-D])\s+(\d{1,2})[.,]?\d{0,2}\s*%\s+([\d.,]+)', 0.95, 'inline'),
|
||||
|
||||
# === FORMAT 2: REVERSED (Stepout-style) ===
|
||||
# "5.00% TUA*B" - procent ÎNAINTE de TVA
|
||||
(r'(\d{1,2})[.,]\d{0,2}\s*%\s*T[UV]A\*?([A-D])', 0.97, 'reversed'),
|
||||
|
||||
# === FORMAT 3: TABLE (OMV-style) ===
|
||||
# "A-21,00% 285,66 49,58" (cod-procent bază tva)
|
||||
(r'([A-D])\s*[-:]\s*(\d{1,2})[.,\s]*\d{0,2}\s*%\s+([\d.,\s]+)\s+([\d.,\s]+)', 0.96, 'table'),
|
||||
(r'TOTAL\s+TAXE\s*:?\s*([\d.,\s]+)', 0.95, 'taxe'),
|
||||
|
||||
# === FORMAT 4: MULTILINE (Brick/Electrobering) ===
|
||||
# "TOTAL TVA A - 19%" pe o linie, amount pe următoarea
|
||||
(r'TOTAL\s+TVA\s*([A-D])\s*[-\s]+(\d{1,2})\s*%', 0.96, 'multiline'),
|
||||
(r'TOTAL\s+TVA\s+([A-D])\s+(\d{1,2})\s*%', 0.95, 'multiline'),
|
||||
(r'TOTAL\s+T[VU][AR]\s*([A-D])?\s*[-:]?\s*(\d{1,2})\s*%', 0.94, 'multiline'),
|
||||
|
||||
# === FORMAT 5: STANDARD (din extractor) ===
|
||||
(r'TOTAL\s+(?:T[VU][AR]|IVA)\s+BON\s*:?\s*([\d\s.,]+)', 0.98, 'bon'),
|
||||
(r'T[O0]TAL\s+(?:T[VU][AR]|IVA)\s*:?\s*([\d\s.,]+)', 0.95, 'standard'),
|
||||
(r'TOTAL\s+IVA\s*:?\s*([\d\s.,]+)', 0.95, 'standard'),
|
||||
(r'T[VU][AR]\s+(?:A\s*[-:]?\s*)?(\d{1,2})\s*%\s*:?\s*([\d\s.,]+)', 0.95, 'percent'),
|
||||
(r'T[VU][AR]\s+[A-Z]\s*[-:]\s*(\d{1,2})\s*%\s*([\d\s.,]+)', 0.93, 'percent'),
|
||||
(r'T[VU][AR]\s*[C5]\s*[-:]\s*5\s*%\s*:?\s*([\d\s.,]+)', 0.93, 'books'),
|
||||
(r'(?:T[VU][AR]|IVA)\s+5\s*%\s*:?\s*([\d\s.,]+)', 0.92, 'books'),
|
||||
|
||||
# === FORMAT 6: CODED inline (cu code A-D) ===
|
||||
(r'T[UV]A\s*[-:]?\s*([A-D])\s*:?\s*(\d{1,2})\s*%\s*[=:]?\s*([\d.,\s]+)', 0.95, 'coded'),
|
||||
(r'([A-D])\s*[-:]\s*(\d{1,2})[.,]?\d{0,2}\s*%\s*[=:]?\s*([\d.,\s]+)', 0.93, 'coded'),
|
||||
|
||||
# === FALLBACK patterns ===
|
||||
(r'T[O0]T[AE]L\s+(?:T[VUAI]+[AR]?|IVA)\s*:?\s*([\d\s.,]+)', 0.88, 'fallback'),
|
||||
(r'TOTAL\s+TA\s*[F\s]?\s*\d*\s*:?\s*([\d\s.,]+)', 0.85, 'fallback'),
|
||||
(r'(?:T[VU][AR]|IVA)\s*:?\s*([\d\s.,]+)', 0.85, 'fallback'),
|
||||
(r'TOTAL\s+T[UV]A\s*:?\s*([\d.,\s]+)', 0.90, 'standard'),
|
||||
]
|
||||
|
||||
# Non-VAT payer patterns - NEPLATITOR DE TVA
|
||||
# OCR variants: NEPLATTOR, NEPLATITOR, NEPLATOR, ANEPLATHTOR, MEPLATITOR
|
||||
NON_VAT_PATTERNS = [
|
||||
r'NEPLAT\w*OR', # NEPLATITOR, NEPLATTOR, NEPLATOR
|
||||
r'[ANM]EPLAT\w*O?R', # OCR errors: ANEPLATHTOR, MEPLATITOR
|
||||
r'TOTAL\s+NEPLAT', # TOTAL NEPLATITOR...
|
||||
r'TOTAL\s+[ANM]EPLAT', # TOTAL ANEPLAT... (OCR error)
|
||||
r'SCUTIT\s*(?:DE\s+)?T[VU]A', # SCUTIT DE TVA
|
||||
r'NEPLAT\w*\s+T[VU]A', # NEPLATITOR TVA
|
||||
r'NEPLAT\w*\s+DE\s+T', # NEPLATITOR DE T... (truncated)
|
||||
]
|
||||
|
||||
# CUI (fiscal code) patterns - VENDOR CUI (exclude CLIENT)
|
||||
# OCR errors: R0 instead of RO, C1F instead of CIF
|
||||
CUI_PATTERNS = [
|
||||
# CIF at start of line (definitely vendor)
|
||||
(r'^CIF\s*:?\s*(R[O0]?\d{6,10})', 0.98),
|
||||
(r'^CIF\s*:?\s*(\d{6,10})', 0.97),
|
||||
(r'^C[I1]F\s*:?\s*(R[O0]?\d{6,10})', 0.95),
|
||||
(r'^C[I1]F\s*:?\s*(\d{6,10})', 0.94),
|
||||
# CIF not preceded by CLIENT (negative lookbehind)
|
||||
(r'(?<!CLIENT\s)(?<!LIENT\s)CIF\s*:?\s*(R[O0]?\d{6,10})', 0.95),
|
||||
(r'(?<!CLIENT\s)(?<!LIENT\s)CIF\s*:?\s*(\d{6,10})', 0.94),
|
||||
# Standalone CIF with word boundary
|
||||
(r'\bC[I1]F\s*:?\s*(R[O0]?\d{6,10})\b', 0.90),
|
||||
(r'\bC[I1]F\s*:?\s*(\d{6,10})\b', 0.89),
|
||||
# COD FISCAL (vendor)
|
||||
(r'COD\s+FISCAL\s*:?\s*(R[O0]?\d{6,10})', 0.90),
|
||||
(r'COD\s+FISCAL\s*:?\s*(\d{6,10})', 0.89),
|
||||
# C. I. F. format with SPACES (OCR artifact)
|
||||
(r'C\.\s*I\.\s*F\.?\s*[:\s]+(R[O0]?\d{6,10})', 0.92),
|
||||
(r'C\.\s*I\.\s*F\.?\s*[:\s]+(\d{6,10})', 0.91),
|
||||
# C.I.F. format (with dots, no spaces)
|
||||
(r'(?<!CLIENT\s)C\.[I1]\.F\.?\s*:?\s*(R[O0]?\d{6,10})', 0.88),
|
||||
(r'(?<!CLIENT\s)C\.[I1]\.F\.?\s*:?\s*(\d{6,10})', 0.87),
|
||||
# CUI format
|
||||
(r'(?<!CLIENT\s)C\.?U\.?[I1]\.?\s*:?\s*(R[O0]?\d{6,10})', 0.85),
|
||||
(r'(?<!CLIENT\s)C\.?U\.?[I1]\.?\s*:?\s*(\d{6,10})', 0.84),
|
||||
# Lidl format: "Cod Identificare fiscala" (OCR corrupted)
|
||||
(r'[IC](?:od|ed)\s*Identific[a-z/]*\s*(R[O0]\d{6,10})', 0.90),
|
||||
# Generic: anything with "fiscal" followed by RO + digits
|
||||
(r'fiscal[a-z]*\s*:?\s*(R[O0]\d{6,10})', 0.85),
|
||||
]
|
||||
|
||||
# CUI REVERSED format (number BEFORE label)
|
||||
CUI_REVERSED_PATTERNS = [
|
||||
(r'(R[O0]\d{6,10})\s*\n\s*C\.?\s*I\.?\s*F\.?', 0.98),
|
||||
(r'(\d{6,10})\s*\n\s*C\.?\s*I\.?\s*F\.?', 0.95),
|
||||
]
|
||||
|
||||
# Items count patterns - NR POZ ART IN BON
|
||||
ITEMS_COUNT_PATTERNS = [
|
||||
(r'NR\.?\s*P?[O0]Z\.?\s*ART\.?\s*(?:IN\s+BON)?\s*:?\s*(\d+)', 0.98),
|
||||
(r'(\d{1,2})\s*\n\s*[O0]Z\.?\s*ART', 0.95),
|
||||
(r'[O0]Z\.?\s*ART\.?\s*(?:IN\s+BON)?\s*:?\s*[\n\s]*(\d+)', 0.93),
|
||||
(r'NR\.?\s*(?:P?[O0]Z\.?)?\s*ART(?:ICOLE)?\.?\s*(?:IN\s+BON)?\s*:?\s*[\n\s]*(\d+)', 0.90),
|
||||
(r'ARTIC[O0]LE\s*:?\s*(\d+)', 0.88),
|
||||
(r'(?:^|\s|:)P?[O0]Z\.?\s*(?:ART)?\.?\s*(?:IN\s+BON)?\s*:?\s*(\d{1,3})(?:\s|$)', 0.85),
|
||||
]
|
||||
|
||||
# Series patterns - Romanian fiscal receipt series
|
||||
SERIES_PATTERNS = [
|
||||
(r'SERIE\s*:?\s*([A-Z]{1,4})', 0.90),
|
||||
(r'(?:^|\s)Z\s*:\s*(\d{4})', 0.85),
|
||||
(r'(?:^|\s)BF\s*:\s*(\d{4})', 0.85),
|
||||
]
|
||||
|
||||
# -------------------------------------------------------------------------
|
||||
# Extraction methods - override in subclasses as needed
|
||||
# -------------------------------------------------------------------------
|
||||
|
||||
def extract_tva_entries(self, text: str) -> List[dict]:
|
||||
"""
|
||||
Extract TVA entries from receipt text.
|
||||
Extract TVA entries from receipt text - GENERIC implementation.
|
||||
|
||||
Override this method in subclasses to handle store-specific TVA formats.
|
||||
Handles ALL formats:
|
||||
- Multi-rate inline (Lidl): "TVA A 21% 7.71"
|
||||
- Reversed (Stepout): "5.00% TUA*B"
|
||||
- Table (OMV): "A-21,00% 285,66 49,58"
|
||||
- Multiline: "TOTAL TVA A - 19%" + amount on next line
|
||||
- Non-VAT payers: Returns empty list
|
||||
|
||||
Args:
|
||||
text: Raw OCR text from receipt
|
||||
@@ -174,12 +348,252 @@ class BaseStoreProfile(ABC):
|
||||
Returns:
|
||||
List of dicts with keys: code, percent, amount
|
||||
"""
|
||||
return []
|
||||
entries = []
|
||||
text_upper = text.upper()
|
||||
|
||||
# Step 1: Check for known non-VAT payer (by class flag or text detection)
|
||||
if self.IS_NON_VAT_PAYER or self._is_non_vat_payer(text_upper):
|
||||
return [] # No TVA entries for non-VAT payers
|
||||
|
||||
# Step 2: Normalize OCR spaces in numbers
|
||||
normalized = re.sub(r'(\d+)[.,]\s+(\d{2})', r'\1.\2', text_upper)
|
||||
lines = normalized.split('\n')
|
||||
|
||||
# Step 3: Try all formats, collect candidates
|
||||
candidates = []
|
||||
|
||||
# Try inline multi-rate (Lidl-style)
|
||||
candidates.extend(self._try_tva_inline(normalized))
|
||||
|
||||
# Try reversed format (Stepout-style)
|
||||
candidates.extend(self._try_tva_reversed(normalized, lines))
|
||||
|
||||
# Try multiline format (Brick/Electrobering)
|
||||
candidates.extend(self._try_tva_multiline(normalized, lines))
|
||||
|
||||
# Try table format (OMV-style)
|
||||
candidates.extend(self._try_tva_table(normalized))
|
||||
|
||||
# Try standard/fallback patterns
|
||||
if not candidates:
|
||||
candidates.extend(self._try_tva_standard(normalized))
|
||||
|
||||
# Step 4: Deduplicate and return
|
||||
seen = set()
|
||||
for entry in candidates:
|
||||
key = (entry.get('code', 'A'), entry.get('percent', 19))
|
||||
if key not in seen and entry.get('amount') and entry['amount'] > 0:
|
||||
entries.append(entry)
|
||||
seen.add(key)
|
||||
|
||||
return entries
|
||||
|
||||
def _is_non_vat_payer(self, text: str) -> bool:
|
||||
"""Check if receipt is from non-VAT payer."""
|
||||
for pattern in self.NON_VAT_PATTERNS:
|
||||
if re.search(pattern, text, re.IGNORECASE):
|
||||
return True
|
||||
return False
|
||||
|
||||
def _try_tva_inline(self, text: str) -> List[dict]:
|
||||
"""Try Lidl-style inline format: 'TVA A 21,00% 7.71'"""
|
||||
entries = []
|
||||
# Pattern: "TVA A 21,00% 7.71" or "TVA B 11,00% 2.13"
|
||||
for pattern, confidence, fmt in self.TVA_PATTERNS:
|
||||
if fmt != 'inline':
|
||||
continue
|
||||
for match in re.finditer(pattern, text, re.IGNORECASE):
|
||||
try:
|
||||
groups = match.groups()
|
||||
if len(groups) >= 3:
|
||||
code = groups[0].upper() if groups[0] else 'A'
|
||||
percent = int(groups[1])
|
||||
amount = self._parse_decimal(self._clean_ocr_number(groups[2]))
|
||||
if amount and amount > 0:
|
||||
entries.append({
|
||||
'code': code,
|
||||
'percent': percent,
|
||||
'amount': amount
|
||||
})
|
||||
except (ValueError, InvalidOperation, IndexError):
|
||||
continue
|
||||
return entries
|
||||
|
||||
def _try_tva_reversed(self, text: str, lines: List[str]) -> List[dict]:
|
||||
"""Try Stepout-style reversed format: '5.00% TUA*B 2.00' (rate BEFORE TVA marker)"""
|
||||
entries = []
|
||||
# Pattern: "5.00% TUA*B 2.00" - procent BEFORE TVA, amount same line or next
|
||||
for i, line in enumerate(lines):
|
||||
# Try pattern with amount on SAME line: "5.00% TUA*B 2.00"
|
||||
match = re.search(
|
||||
r'(\d{1,2})[.,]?\d{0,2}\s*%\s*T[UV][AR]\s*\*?\s*([A-D])?\s+([\d\s.,]+)',
|
||||
line, re.IGNORECASE
|
||||
)
|
||||
if match:
|
||||
try:
|
||||
percent = int(match.group(1))
|
||||
code = match.group(2).upper() if match.group(2) else 'A'
|
||||
amount_str = match.group(3).strip()
|
||||
amount = self._parse_decimal(amount_str)
|
||||
if amount and amount > 0:
|
||||
entries.append({
|
||||
'code': code,
|
||||
'percent': percent,
|
||||
'amount': amount
|
||||
})
|
||||
continue # Check for more entries
|
||||
except (ValueError, InvalidOperation, IndexError):
|
||||
pass
|
||||
|
||||
# Fallback: amount on NEXT line
|
||||
match = re.search(r'(\d{1,2})[.,]?\d{0,2}\s*%\s*T[UV][AR]\s*\*?\s*([A-D])?$', line, re.IGNORECASE)
|
||||
if match:
|
||||
try:
|
||||
percent = int(match.group(1))
|
||||
code = match.group(2).upper() if match.group(2) else 'A'
|
||||
if i + 1 < len(lines):
|
||||
amount_str = lines[i + 1].strip()
|
||||
amount = self._parse_decimal(amount_str)
|
||||
if amount and amount > 0:
|
||||
entries.append({
|
||||
'code': code,
|
||||
'percent': percent,
|
||||
'amount': amount
|
||||
})
|
||||
except (ValueError, InvalidOperation, IndexError):
|
||||
continue
|
||||
return entries
|
||||
|
||||
def _try_tva_multiline(self, text: str, lines: List[str]) -> List[dict]:
|
||||
"""Try multiline format: 'TOTAL TVA A - 19%' + amount on next line"""
|
||||
entries = []
|
||||
# Pattern: "TOTAL TVA A - 19%" or "TOTAL TVA A 19%" on one line, amount on next
|
||||
multiline_patterns = [
|
||||
r'TOTAL\s+TVA\s*([A-D])\s*[-\s]+(\d{1,2})\s*%',
|
||||
r'TOTAL\s+TVA\s+([A-D])\s+(\d{1,2})\s*%',
|
||||
r'TOTAL\s+T[VU][AR]\s*([A-D])?\s*[-:]?\s*(\d{1,2})\s*%',
|
||||
r'O?TAL\s+[IT]VA\s*([A-D])\s*[-:]?\s*(\d{1,2})\s*%',
|
||||
]
|
||||
for i, line in enumerate(lines):
|
||||
for pattern in multiline_patterns:
|
||||
match = re.search(pattern, line, re.IGNORECASE)
|
||||
if match:
|
||||
try:
|
||||
code = match.group(1).upper() if match.group(1) else 'A'
|
||||
percent = int(match.group(2))
|
||||
# Amount is on next line
|
||||
if i + 1 < len(lines):
|
||||
amount_str = lines[i + 1].strip()
|
||||
amount = self._parse_decimal(amount_str)
|
||||
if amount and amount > 0:
|
||||
entries.append({
|
||||
'code': code,
|
||||
'percent': percent,
|
||||
'amount': amount
|
||||
})
|
||||
return entries
|
||||
except (ValueError, InvalidOperation, IndexError):
|
||||
continue
|
||||
return entries
|
||||
|
||||
def _try_tva_table(self, text: str) -> List[dict]:
|
||||
"""Try OMV-style table format: 'A-21,00% 285,66 49,58'"""
|
||||
entries = []
|
||||
# Pattern: "A-21,00% 285,66 49,58" (code-percent base_amount tva_amount)
|
||||
table_pattern = r'([A-D])\s*[-:]\s*(\d{1,2})[.,\s]*\d{0,2}\s*%\s+([\d.,\s]+)\s+([\d.,\s]+)'
|
||||
for match in re.finditer(table_pattern, text, re.IGNORECASE):
|
||||
try:
|
||||
code = match.group(1).upper()
|
||||
percent = int(match.group(2))
|
||||
# Group 4 is the TVA amount (last column in table)
|
||||
tva_amount_str = self._clean_ocr_number(match.group(4))
|
||||
tva_amount = self._parse_decimal(tva_amount_str)
|
||||
if tva_amount and tva_amount > 0:
|
||||
entries.append({
|
||||
'code': code,
|
||||
'percent': percent,
|
||||
'amount': tva_amount
|
||||
})
|
||||
except (ValueError, InvalidOperation, IndexError):
|
||||
continue
|
||||
|
||||
# Fallback: "TOTAL TAXE: 55,22"
|
||||
if not entries:
|
||||
taxe_match = re.search(r'TOTAL\s+TAXE\s*:?\s*([\d.,\s]+)', text, re.IGNORECASE)
|
||||
if taxe_match:
|
||||
try:
|
||||
amount_str = self._clean_ocr_number(taxe_match.group(1))
|
||||
amount = self._parse_decimal(amount_str)
|
||||
if amount and amount > 0:
|
||||
entries.append({
|
||||
'code': 'A',
|
||||
'percent': 19, # Default rate
|
||||
'amount': amount
|
||||
})
|
||||
except (ValueError, InvalidOperation):
|
||||
pass
|
||||
return entries
|
||||
|
||||
def _try_tva_standard(self, text: str) -> List[dict]:
|
||||
"""Try standard TVA patterns as fallback"""
|
||||
entries = []
|
||||
standard_fmts = ['standard', 'bon', 'percent', 'coded', 'fallback', 'books']
|
||||
for pattern, confidence, fmt in self.TVA_PATTERNS:
|
||||
if fmt not in standard_fmts:
|
||||
continue
|
||||
match = re.search(pattern, text, re.IGNORECASE)
|
||||
if match:
|
||||
try:
|
||||
groups = match.groups()
|
||||
if len(groups) >= 2:
|
||||
# Could be (percent, amount) or (code, percent, amount)
|
||||
if groups[0] and groups[0].isalpha():
|
||||
code = groups[0].upper()
|
||||
percent = int(groups[1]) if len(groups) > 1 else 19
|
||||
amount_str = groups[2] if len(groups) > 2 else None
|
||||
else:
|
||||
code = 'A'
|
||||
percent = int(groups[0]) if groups[0] and groups[0].isdigit() else 19
|
||||
amount_str = groups[1] if len(groups) > 1 else groups[0]
|
||||
if amount_str:
|
||||
amount = self._parse_decimal(self._clean_ocr_number(amount_str))
|
||||
if amount and amount > 0:
|
||||
entries.append({
|
||||
'code': code,
|
||||
'percent': percent,
|
||||
'amount': amount
|
||||
})
|
||||
return entries
|
||||
elif len(groups) == 1:
|
||||
# Just amount
|
||||
amount = self._parse_decimal(self._clean_ocr_number(groups[0]))
|
||||
if amount and amount > 0:
|
||||
entries.append({
|
||||
'code': 'A',
|
||||
'percent': 19,
|
||||
'amount': amount
|
||||
})
|
||||
return entries
|
||||
except (ValueError, InvalidOperation, IndexError):
|
||||
continue
|
||||
return entries
|
||||
|
||||
def _clean_ocr_number(self, value: str) -> str:
|
||||
"""Remove OCR spaces from numbers (e.g., '55, 22' -> '55,22')."""
|
||||
if not value:
|
||||
return ""
|
||||
value = re.sub(r'\s*([.,])\s*', r'\1', value)
|
||||
value = value.replace(' ', '')
|
||||
return value
|
||||
|
||||
def extract_total(self, text: str) -> Tuple[Optional[Decimal], float]:
|
||||
"""
|
||||
Extract total amount from receipt text.
|
||||
|
||||
Supports both single-line and multiline formats:
|
||||
- Single line: "TOTAL: 78.00", "SUMA TOTALA: 78.00"
|
||||
- Multiline: "SUMA\nTOTALA:\n78.00" (common in thermal receipts)
|
||||
|
||||
Args:
|
||||
text: Raw OCR text from receipt
|
||||
|
||||
@@ -187,7 +601,54 @@ class BaseStoreProfile(ABC):
|
||||
Tuple of (amount, confidence) or (None, 0.0)
|
||||
"""
|
||||
text_upper = text.upper()
|
||||
lines = text_upper.split('\n')
|
||||
|
||||
# =====================================================================
|
||||
# STRATEGY 1: Multiline "SUMA TOTALA" pattern (thermal receipts)
|
||||
# Format: SUMA on one line, TOTALA: on next, amount on third
|
||||
# =====================================================================
|
||||
for i, line in enumerate(lines):
|
||||
line_clean = line.strip()
|
||||
|
||||
# Check for "SUMA" keyword (with OCR variants: SUNA, SUHA, SUM A)
|
||||
if re.search(r'S[UU]M[AĂ\s]', line_clean):
|
||||
# Look at next 3 lines for "TOTALA" and amount
|
||||
for j in range(i, min(i + 4, len(lines))):
|
||||
check_line = lines[j].strip()
|
||||
|
||||
# Check for "TOTALA:" or "TOTALA -" followed by amount
|
||||
match = re.search(r'T[O0]TALA\s*[:\-]\s*([\d\s.,]+)', check_line)
|
||||
if match:
|
||||
amount = self._parse_decimal(self._clean_ocr_number(match.group(1)))
|
||||
if amount and amount > 0 and amount < self.MAX_PAYMENT:
|
||||
return (amount, 0.98)
|
||||
|
||||
# Check for "TOTALA" without amount, amount on next line
|
||||
if re.search(r'T[O0]TALA\s*[:\-]?\s*$', check_line):
|
||||
if j + 1 < len(lines):
|
||||
amount_line = lines[j + 1].strip()
|
||||
amount = self._parse_decimal(amount_line)
|
||||
if amount and amount > 0 and amount < self.MAX_PAYMENT:
|
||||
return (amount, 0.97)
|
||||
|
||||
# Check for "SUMA TOTALA" on single line with amount
|
||||
match = re.search(r'S[UU]M[AĂ]\s+T[O0]TALA\s*[:\-]\s*([\d\s.,]+)', line_clean)
|
||||
if match:
|
||||
amount = self._parse_decimal(self._clean_ocr_number(match.group(1)))
|
||||
if amount and amount > 0 and amount < self.MAX_PAYMENT:
|
||||
return (amount, 0.98)
|
||||
|
||||
# Check for "SUMA TOTALA" without amount, amount on next line
|
||||
if re.search(r'S[UU]M[AĂ]\s+T[O0]TALA\s*[:\-]?\s*$', line_clean):
|
||||
if i + 1 < len(lines):
|
||||
next_line = lines[i + 1].strip()
|
||||
amount = self._parse_decimal(next_line)
|
||||
if amount and amount > 0 and amount < self.MAX_PAYMENT:
|
||||
return (amount, 0.96)
|
||||
|
||||
# =====================================================================
|
||||
# STRATEGY 2: Standard single-line patterns
|
||||
# =====================================================================
|
||||
for pattern, confidence in self.TOTAL_PATTERNS:
|
||||
match = re.search(pattern, text_upper)
|
||||
if match:
|
||||
@@ -259,28 +720,76 @@ class BaseStoreProfile(ABC):
|
||||
"""
|
||||
Extract payment methods (CARD/NUMERAR) from receipt.
|
||||
|
||||
Supports multiple payments of the same type (e.g., 2x CARD for split payments).
|
||||
Each payment is returned as a separate entry with its amount.
|
||||
Supports:
|
||||
- Multiline patterns: "CARD\n78.00" (common in thermal receipts)
|
||||
- Multiple payments (split CARD + NUMERAR)
|
||||
- REST (change) detection to calculate actual CARD amount
|
||||
- Keyword-only CARD/NUMERAR that infers from total
|
||||
- Fallback for fiscal receipts without explicit payment
|
||||
|
||||
Args:
|
||||
text: Raw OCR text from receipt
|
||||
|
||||
Returns:
|
||||
List of dicts: [{'method': 'CARD'/'NUMERAR', 'amount': Decimal, 'confidence': float}]
|
||||
Multiple entries of same method type are allowed for split payments.
|
||||
"""
|
||||
text_upper = text.upper()
|
||||
lines = text_upper.split('\n')
|
||||
methods = []
|
||||
# Track (method, amount) pairs to avoid exact duplicates from overlapping patterns
|
||||
seen_entries = set()
|
||||
|
||||
# =====================================================================
|
||||
# STEP 0: Try MULTILINE patterns first (thermal receipts)
|
||||
# Format: "CARD" on one line, amount on next line
|
||||
# =====================================================================
|
||||
for i, line in enumerate(lines):
|
||||
line_clean = line.strip()
|
||||
|
||||
# Standalone CARD keyword (not part of MASTERCARD, etc.)
|
||||
if re.match(r'^CARD\s*$', line_clean):
|
||||
if i + 1 < len(lines):
|
||||
next_line = lines[i + 1].strip()
|
||||
# Must be a valid amount (not another keyword)
|
||||
if re.match(r'^[\d\s.,]+$', next_line):
|
||||
amount = self._parse_decimal(next_line)
|
||||
if amount and amount > 0 and amount < self.MAX_PAYMENT:
|
||||
entry_key = ('CARD', amount)
|
||||
if entry_key not in seen_entries:
|
||||
methods.append({
|
||||
'method': 'CARD',
|
||||
'amount': amount,
|
||||
'confidence': 0.95
|
||||
})
|
||||
seen_entries.add(entry_key)
|
||||
|
||||
# Standalone NUMERAR keyword
|
||||
if re.match(r'^NUMERAR\s*$', line_clean):
|
||||
if i + 1 < len(lines):
|
||||
next_line = lines[i + 1].strip()
|
||||
if re.match(r'^[\d\s.,]+$', next_line):
|
||||
amount = self._parse_decimal(next_line)
|
||||
if amount and amount > 0 and amount < self.MAX_PAYMENT:
|
||||
entry_key = ('NUMERAR', amount)
|
||||
if entry_key not in seen_entries:
|
||||
methods.append({
|
||||
'method': 'NUMERAR',
|
||||
'amount': amount,
|
||||
'confidence': 0.95
|
||||
})
|
||||
seen_entries.add(entry_key)
|
||||
|
||||
# If multiline extraction found methods, return them
|
||||
if methods:
|
||||
return methods
|
||||
|
||||
# =====================================================================
|
||||
# STEP 1: Try pattern-based extraction with explicit amounts
|
||||
# =====================================================================
|
||||
for pattern, method, confidence in self.PAYMENT_PATTERNS:
|
||||
for match in re.finditer(pattern, text_upper):
|
||||
try:
|
||||
amount = self._parse_decimal(match.group(1))
|
||||
if amount and amount > 0 and amount < self.MAX_PAYMENT:
|
||||
# Deduplicate by (method, amount) to avoid same entry from multiple patterns
|
||||
# But allow different amounts for same method (split payments)
|
||||
entry_key = (method, amount)
|
||||
if entry_key not in seen_entries:
|
||||
methods.append({
|
||||
@@ -292,6 +801,70 @@ class BaseStoreProfile(ABC):
|
||||
except (ValueError, InvalidOperation):
|
||||
continue
|
||||
|
||||
# If we found explicit amounts, we're done
|
||||
if methods:
|
||||
return methods
|
||||
|
||||
# Step 2: Try keyword-only detection with REST logic
|
||||
# Get total amount for inference
|
||||
total_amount, _ = self.extract_total(text)
|
||||
if not total_amount:
|
||||
return []
|
||||
|
||||
# Check for payment keywords
|
||||
has_card = any(kw in text_upper for kw in ['CARD', 'CARTE CREDIT', 'VISA', 'MASTERCARD', 'DEBIT', 'CREDIT', 'CONTACTLESS'])
|
||||
has_numerar = any(kw in text_upper for kw in ['NUMERAR', 'CASH'])
|
||||
|
||||
# Find REST (change) amount
|
||||
rest_amount = Decimal('0')
|
||||
for i, line in enumerate(lines):
|
||||
if 'REST' in line:
|
||||
# REST on same line: "REST 0.00" or "REST: 0.00"
|
||||
match = re.search(r'REST\s*:?\s*([\d.,]+)', line)
|
||||
if match:
|
||||
rest_amount = self._parse_decimal(match.group(1)) or Decimal('0')
|
||||
elif i + 1 < len(lines):
|
||||
# REST on separate line
|
||||
rest_amount = self._parse_decimal(lines[i + 1].strip()) or Decimal('0')
|
||||
break
|
||||
|
||||
# Calculate payment amounts
|
||||
if has_card:
|
||||
card_amount = total_amount - rest_amount
|
||||
if card_amount > 0:
|
||||
methods.append({
|
||||
'method': 'CARD',
|
||||
'amount': card_amount,
|
||||
'confidence': 0.90
|
||||
})
|
||||
|
||||
if has_numerar:
|
||||
if has_card and rest_amount > 0:
|
||||
# Mixed payment: NUMERAR is the change given back
|
||||
methods.append({
|
||||
'method': 'NUMERAR',
|
||||
'amount': rest_amount,
|
||||
'confidence': 0.85
|
||||
})
|
||||
elif not has_card:
|
||||
# Cash only
|
||||
methods.append({
|
||||
'method': 'NUMERAR',
|
||||
'amount': total_amount,
|
||||
'confidence': 0.90
|
||||
})
|
||||
|
||||
# Step 3: Fallback for fiscal receipts without explicit payment
|
||||
if not methods and total_amount and total_amount > 0:
|
||||
is_fiscal = 'BON FISCAL' in text_upper or 'FISCAL' in text_upper
|
||||
if is_fiscal:
|
||||
# Default to CARD for business purchases (most common)
|
||||
methods.append({
|
||||
'method': 'CARD',
|
||||
'amount': total_amount,
|
||||
'confidence': 0.70
|
||||
})
|
||||
|
||||
return methods
|
||||
|
||||
def extract_client_cui(self, text: str) -> Tuple[Optional[str], float]:
|
||||
|
||||
Reference in New Issue
Block a user