Upload 5 files

samples/generate_samples.py

@@ -0,0 +1,240 @@
+"""Generate synthetic bank statements + ground-truth JSON.
+
+Produces:
+  samples/sample_digital.pdf        -- real reportlab table (exercises pdfplumber)
+  samples/sample_scan.png           -- rendered + rotated + noised (exercises MiniCPM-V)
+  samples/sample_digital_truth.json -- ground-truth transactions (test fixture)
+  samples/sample_scan_truth.json    -- same data, used for vision-accuracy check
+
+No real bank data is used (privacy + licensing). Narrations are deliberately
+messy to mimic real Indian bank statements.
+
+Run:  python samples/generate_samples.py
+"""
+
+import json
+import os
+import random
+
+HERE = os.path.dirname(os.path.abspath(__file__))
+
+OPENING_BALANCE = 152340.50
+
+# (narration, debit, credit) -- one tuple per transaction row.
+# Narrations are intentionally messy and cover most ledger heads so the
+# rules layer + reconciliation both get a real workout.
+RAW = [
+    ("UPI/CR/512345678901/RAMESH TRADERS/PAYTM/Payment for goods", None, 45000.00),
+    ("NEFT-HDFC0001234-ACME PVT LTD-INV 4521", None, 128500.00),
+    ("UPI/DR/423451234567/SHARMA RENT/GPAY/April office rent", 35000.00, None),
+    ("ACH/NACH/SBILIFE INSURANCE/PREMIUM/POL 9921", 12500.00, None),
+    ("SALARY APR 2026 - STAFF PAYROLL BATCH 04", 86000.00, None),
+    ("ATM CSH WDL/AXIS/KANPUR MALL ROAD/04APR", 10000.00, None),
+    ("UPI/CR/612345098765/SUNIL ENTERPRISES/PHONEPE/sale", None, 22750.00),
+    ("GST PMT-CBIC-26AAAPL1234C1Z5-GSTR3B APR", 18900.00, None),
+    ("NEFT-ICIC0004567-MEGHA SUPPLIERS-PO 7781", 64200.00, None),
+    ("INT.PD-SAVINGS A/C-QTR ENDING", None, 1842.00),
+    ("BANK CHRG-SMS CHGS APR 2026", 23.60, None),
+    ("UPI/DR/734512345098/BSNL BROADBAND/INTERNET BILL", 1499.00, None),
+    ("RTGS-UTIB0000123-DELTA EXPORTS-SETTLEMENT", None, 215000.00),
+    ("ACH/NACH/HDFC HOME LOAN/EMI/LN44219", 28750.00, None),
+    ("UPI/DR/845612340987/KESHAV ELECTRICALS/utility KESCO bill", 6540.00, None),
+    ("CSH DEP/BRANCH/KANPUR/CASH DEPOSIT BY CLIENT", None, 50000.00),
+    ("TDS PMT-CPC-TDS Q4 26AAAPL1234C", 9400.00, None),
+    ("UPI/CR/923451234560/VIKAS AGENCY/sale receipt", None, 17800.00),
+    ("NEFT-PUNB0123456-RAVI CONSULTING-PROF FEES", 25000.00, None),
+    ("IRCTC TRAVEL/RAIL TICKET/PNR 8841200012", 3450.00, None),
+    ("UPI/DR/103451234561/SWIGGY/staff lunch conveyance", 1280.00, None),
+    ("ELECTRICITY KESCO BILL PMT/CONS 4412", 8920.00, None),
+    ("UPI/CR/213451234562/ANJALI STORES/PHONEPE/sale", None, 9650.00),
+    ("AMC CHGS-DEBIT CARD ANNUAL FEE", 590.00, None),
+    ("NEFT-SBIN0009988-OWN ACCOUNT TRANSFER-SELF", 40000.00, None),
+    ("UPI/DR/313451234563/GARG REPAIRS/AC servicing maintenance", 4200.00, None),
+    ("INCOME TAX PMT-CBDT-ADVANCE TAX 26AAAPL", 35000.00, None),
+    ("UPI/CR/413451234564/MOHIT TRADING/sale", None, 31200.00),
+    ("LIC OF INDIA/ACH/PREMIUM/POLICY 5512", 9800.00, None),
+    ("UPI/DR/513451234565/JIO RECHARGE/mobile postpaid", 799.00, None),
+    ("NEFT-YESB0001122-GLOBE IMPORTS-PURCHASE", 52400.00, None),
+    ("DRAWINGS-PROPRIETOR-PERSONAL WITHDRAWAL", 30000.00, None),
+    ("UPI/CR/613451234566/CAPITAL INTRODUCED BY PARTNER", None, 200000.00),
+    ("INT CREDIT-FIXED DEPOSIT INTEREST", None, 4521.00),
+    ("UPI/DR/713451234567/PETROL HPCL/fuel conveyance", 2500.00, None),
+    ("NEFT-KKBK0007788-ZENITH PVT LTD-INV 9921", None, 76300.00),
+    ("ATM CSH WDL/HDFC/CIVIL LINES/22APR", 15000.00, None),
+    ("UPI/DR/813451234568/PRINTWELL/office stationery purchase", 3120.00, None),
+    ("INTEREST PAID ON OVERDRAFT-OD A/C", 5640.00, None),
+    ("RTGS-HDFC0009090-NEXUS SOLUTIONS-FINAL PMT", None, 98000.00),
+]
+
+DATES = None
+
+
+def build_transactions():
+    """Return the ground-truth transaction list with running balances."""
+    txns = []
+    balance = OPENING_BALANCE
+    # Dates: spread across April 2026, monotonically non-decreasing.
+    day = 1
+    for i, (narration, debit, credit) in enumerate(RAW):
+        # advance the day every ~1-2 rows, stay within April
+        if i > 0 and i % 2 == 0:
+            day = min(day + 1, 30)
+        date_iso = f"2026-04-{day:02d}"
+        balance = round(balance - (debit or 0) + (credit or 0), 2)
+        ref = "".join([c for c in narration if c.isdigit()][:6]) or None
+        txns.append({
+            "date": date_iso,
+            "narration": narration,
+            "ref_no": ref,
+            "debit": debit,
+            "credit": credit,
+            "balance": balance,
+        })
+    return txns
+
+
+def _printed_date(iso):
+    """ISO -> DD/MM/YYYY as printed on the statement."""
+    y, m, d = iso.split("-")
+    return f"{d}/{m}/{y}"
+
+
+def _fmt_amt(v):
+    if v is None:
+        return ""
+    # Indian-style grouping is hard to parse; keep it plain with thousands commas
+    return f"{v:,.2f}"
+
+
+def render_pdf(txns, path):
+    from reportlab.lib import colors
+    from reportlab.lib.pagesizes import A4
+    from reportlab.lib.units import mm
+    from reportlab.platypus import (SimpleDocTemplate, Table, TableStyle,
+                                    Paragraph, Spacer)
+    from reportlab.lib.styles import getSampleStyleSheet
+
+    styles = getSampleStyleSheet()
+    doc = SimpleDocTemplate(path, pagesize=A4,
+                            leftMargin=10 * mm, rightMargin=10 * mm,
+                            topMargin=12 * mm, bottomMargin=12 * mm)
+    story = []
+    story.append(Paragraph("STATE BANK STYLE DEMO — Account Statement", styles["Title"]))
+    story.append(Paragraph("Account: XXXXXX4412    Period: 01/04/2026 - 30/04/2026",
+                           styles["Normal"]))
+    story.append(Paragraph(f"Opening Balance: {_fmt_amt(OPENING_BALANCE)}", styles["Normal"]))
+    story.append(Spacer(1, 6 * mm))
+
+    header = ["Date", "Narration", "Ref No", "Debit", "Credit", "Balance"]
+    data = [header]
+    narr_style = styles["BodyText"]
+    narr_style.fontSize = 7
+    narr_style.leading = 8
+    for t in txns:
+        data.append([
+            _printed_date(t["date"]),
+            Paragraph(t["narration"], narr_style),
+            t["ref_no"] or "",
+            _fmt_amt(t["debit"]),
+            _fmt_amt(t["credit"]),
+            _fmt_amt(t["balance"]),
+        ])
+
+    col_widths = [20 * mm, 78 * mm, 16 * mm, 22 * mm, 22 * mm, 24 * mm]
+    table = Table(data, colWidths=col_widths, repeatRows=1)
+    table.setStyle(TableStyle([
+        ("BACKGROUND", (0, 0), (-1, 0), colors.HexColor("#1a3c6e")),
+        ("TEXTCOLOR", (0, 0), (-1, 0), colors.white),
+        ("FONTSIZE", (0, 0), (-1, -1), 7),
+        ("FONTNAME", (0, 0), (-1, 0), "Helvetica-Bold"),
+        ("ALIGN", (3, 0), (5, -1), "RIGHT"),
+        ("GRID", (0, 0), (-1, -1), 0.4, colors.grey),
+        ("ROWBACKGROUNDS", (0, 1), (-1, -1), [colors.white, colors.HexColor("#eef3fa")]),
+        ("VALIGN", (0, 0), (-1, -1), "MIDDLE"),
+    ]))
+    story.append(table)
+    doc.build(story)
+
+
+def render_scan(txns, path):
+    """Render a statement page to an image, then rotate + add noise to mimic a scan."""
+    from PIL import Image, ImageDraw, ImageFont
+
+    W, H = 1240, 1754  # ~A4 at 150dpi
+    img = Image.new("RGB", (W, H), "white")
+    draw = ImageDraw.Draw(img)
+    try:
+        font = ImageFont.truetype("arial.ttf", 16)
+        font_b = ImageFont.truetype("arialbd.ttf", 20)
+        font_s = ImageFont.truetype("arial.ttf", 13)
+    except Exception:
+        font = ImageFont.load_default()
+        font_b = font
+        font_s = font
+
+    draw.text((40, 30), "DEMO BANK — Account Statement (scanned)", fill="black", font=font_b)
+    draw.text((40, 64), "Account: XXXXXX4412   Period: 01/04/2026 - 30/04/2026",
+              fill="black", font=font_s)
+    draw.text((40, 84), f"Opening Balance: {_fmt_amt(OPENING_BALANCE)}", fill="black", font=font_s)
+
+    # column x positions
+    cols = [40, 140, 660, 760, 920, 1080]
+    y = 120
+    headers = ["Date", "Narration", "Ref", "Debit", "Credit", "Balance"]
+    for x, h in zip(cols, headers):
+        draw.text((x, y), h, fill="black", font=font)
+    y += 26
+    draw.line((40, y, W - 40, y), fill="black", width=1)
+    y += 6
+
+    for t in txns:
+        draw.text((cols[0], y), _printed_date(t["date"]), fill="black", font=font_s)
+        narr = t["narration"]
+        if len(narr) > 70:
+            narr = narr[:70]
+        draw.text((cols[1], y), narr, fill="black", font=font_s)
+        draw.text((cols[2], y), (t["ref_no"] or ""), fill="black", font=font_s)
+        draw.text((cols[3], y), _fmt_amt(t["debit"]), fill="black", font=font_s)
+        draw.text((cols[4], y), _fmt_amt(t["credit"]), fill="black", font=font_s)
+        draw.text((cols[5], y), _fmt_amt(t["balance"]), fill="black", font=font_s)
+        y += 30
+
+    # Rotate 1.5 deg and add mild gaussian-ish noise to simulate a scan.
+    img = img.rotate(1.5, expand=False, fillcolor="white")
+    px = img.load()
+    rnd = random.Random(42)
+    for _ in range(45000):
+        x = rnd.randint(0, W - 1)
+        yy = rnd.randint(0, H - 1)
+        v = rnd.randint(-25, 25)
+        r, g, b = px[x, yy]
+        px[x, yy] = (max(0, min(255, r + v)),
+                     max(0, min(255, g + v)),
+                     max(0, min(255, b + v)))
+    img.save(path, "PNG")
+
+
+def main():
+    txns = build_transactions()
+
+    pdf_path = os.path.join(HERE, "sample_digital.pdf")
+    scan_path = os.path.join(HERE, "sample_scan.png")
+    truth_digital = os.path.join(HERE, "sample_digital_truth.json")
+    truth_scan = os.path.join(HERE, "sample_scan_truth.json")
+
+    render_pdf(txns, pdf_path)
+    render_scan(txns, scan_path)
+    with open(truth_digital, "w", encoding="utf-8") as f:
+        json.dump(txns, f, indent=2)
+    with open(truth_scan, "w", encoding="utf-8") as f:
+        json.dump(txns, f, indent=2)
+
+    print(f"Wrote {len(txns)} transactions")
+    print(f"  {pdf_path}")
+    print(f"  {scan_path}")
+    print(f"  {truth_digital}")
+    print(f"  {truth_scan}")
+
+
+if __name__ == "__main__":
+    main()

samples/sample_digital.pdf

@@ -0,0 +1,93 @@
+%PDF-1.4
+%���� ReportLab Generated PDF document (opensource)
+1 0 obj
+<<
+/F1 2 0 R /F2 3 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BaseFont /Helvetica-Bold /Encoding /WinAnsiEncoding /Name /F2 /Subtype /Type1 /Type /Font
+>>
+endobj
+4 0 obj
+<<
+/Contents 9 0 R /MediaBox [ 0 0 595.2756 841.8898 ] /Parent 8 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+5 0 obj
+<<
+/Contents 10 0 R /MediaBox [ 0 0 595.2756 841.8898 ] /Parent 8 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+6 0 obj
+<<
+/PageMode /UseNone /Pages 8 0 R /Type /Catalog
+>>
+endobj
+7 0 obj
+<<
+/Author (\(anonymous\)) /CreationDate (D:20260613051145+05'00') /Creator (\(unspecified\)) /Keywords () /ModDate (D:20260613051145+05'00') /Producer (ReportLab PDF Library - \(opensource\)) 
+  /Subject (\(unspecified\)) /Title (\(anonymous\)) /Trapped /False
+>>
+endobj
+8 0 obj
+<<
+/Count 2 /Kids [ 4 0 R 5 0 R ] /Type /Pages
+>>
+endobj
+9 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 4087
+>>
+stream
+GatU7>Bci%&q9:VTuZL>%(ZQ4L,#s$?q^YtGNS6b-LLGO:HI`7Ve=?Rqj"D(-"DVt`i%:JX!ij$V7NT.!`48@oD@Jc!6Z:*f*EK-l:6B'90o,"V#KkU%nL[Ql4V5g#JM:kJPX6rJdNI^ldlY6!,?)&X1nsd(na.:_cTO+JW"qMW1_aL^f=B)FTRk+IKQVH/k`u0g$ZX?1%mp*nF6#?q$g;ApMYh*LZc-rJGS/rrWOj-15()u`TNbO*-F:P<5VTke0rg/0U)q0VZ9f`f"Zg_bn$Y&'f<D7!h_Nc*acYZ$6\m^oe=<(/-(6ITX?VT744eH-g7pL-V^dfh[e&RkHA9=s)e;=E;1.jL)iUSbFlg[$$HErOhC(3+`r"F+9b)lUas,`"ApSW>tc(F.<$_%`$\I3B5nq&0jogJ+_2!4`!!gr^uI(7W;L<MnVeKLR/Rb[)\VrKge\=nG6e:AKK+ApZ]IlE3M%\ALegWT@;pd.(;OY+r9SK!8Z$-mF=8$!IBG$@nEc>K,h)8LF%o3&rq$UsG0(.j1-L7(>r%$H:<t$h/t##BVe8i-cq>YEkntc1Or;1RWuLr&bTLNH1Pj.!1PdbJkA:Edd7()T=fcL_dHX1:WuG?O\rjts^1_LjXR1q3[`-BQ^T^q.-@*GCmfK?lL:sP.!=tX/HF4@l75&'BH*AeC>DQJ6r;QB^rpBLOI%DDb`:ZJe>5n*^#kF>RbX>!W81<?[EVSShG+P%0jDri9&\fC$N:I<%+G[%0<r#%B]:WZU[))qY57-T&S"?-&D<R&oZXLpc6Cl4T[g\]fAm_??>^?g8*0,#'#fXL"BU_s?:m]QZ.4_TXmcpoUh#AZ65OeI=Wg1*jqXX4=a1HHVm#7n+38J.V7RRm4RMe+?T>;:/f-`/f'kIl.d"QFL.XW^D4.iqY..mP2TagOh1[ufc$#animl\.`(Rrrt/6%sL/>@BYJE"!A=hS"7q5jqC(ti;sF`$25@7nQoL"m,NpZ)Sb%^Q4:>sL)b!]-dBSeT/0TOK_o<,':Od\G7%PTbD(-7ZJ:.$M=;f#!*Rkkj]9$+rAa*Zkm/?Tl0WJ?3iW8KIn8PX*0"6-DqKm#mqi+iS?9=:)nhSAH-6Go/n((ZQq#Ge*`P/^l8a:-b7Lq_H=S`%/u@S*kugfOZ06mAsA<r/U_,XGRa$ZXmlR:[BE1PsF[@'E3Lc_$O.O#:g1@rq@D6<Kor6*JfRcW:ErQiRc[H=DYOk1/X(`nHM6nJZ74$ar+H'%TUanE%>EulipFs5STfhK86E%E#u%7T5C\&(;qDMI9L2goJs%dRj]r`-.kZgdW-/+`17(m2P=er-NR5+#hf<*@Q+=f"U:0@1)1.p-(o@'fMZDbDnIS/".qBQg?5HV=n4U&M__CIOXY*iGu>pEO8Gi-%CSlIDNbhak4,fD9>fEPR7.6oq4[<0Tb1WukM]'G8$I,9K48%<iaj#]0TVs,a<W5iqOK2@Yp;tMBs:34gn6]>[#,_):2".EA1D,4!kEs-THY,VQmJ\EG0"E=XY`daXsJ.@`AUk;,#IZb,"k@:]-qj<pB-n0eS<i0RR1"/F6[-P<7+[`dK1"M:2!%\L;>E?BfJ1`V0VGDX$mcZVYRe)OV]Z7etqj<.l"J_;Q4>3lD#=5;4"a)Y)p_^nFT%cfeIIJJkbG4b1Ui/o_dC)WS(U=2s#?_D(G.36r:jO)sif@n(7l!/TSbU%m!bNnd,CtX<n4<YQdE2Lq8'M6=$nN6HCd:73^L\Fg<[;OghE4OXnPp/c]4'W=2CE;?N&^Ai@.KX\U]BOCd(`>U$678V`&ZjQ`V;j_P=-Q7!jmJY<BTWWJ<$#qhbhLrt$L/bX=uKVu\-YtD;_!m0V7=4Yor9'SH1[NaqiTW6OZ6qtUD`B7J$C2"Mff)>FH6Mr9;0qX4O^:\iV[6N=17+%XoU>-EIPF(Bq@MIN=WYBm-IkD(B?;])ho-%tKGUa&7JHQ/e@n&SSMF)k?SoAie;GJ>Gd9&F1R?DkTM(&4=6'<O6?)DV+<Sl'2j4!Pc0-EgU%fC<oHEoh>WIGOEjUDn[G8R_Q),A^Z=rr7IJ]NL"$/mG"8ME50aVnoZ'])_'qAWF'H#[s&H<6r[_qorHA5cM=6"t0)!i$_rbL[9L-H[kKVrefe2$Kn],$/Fa-_DEil;_uLnr%`9s8GPT5M*A1:anu*oR=Y6207tXM_nCQS$5l_#fR94A'*ps/D3I=I\J"q]0/CBh53,naJ[T\*oAhD.!Rm;*0/on[edARTkUR22,cCKXL+BFBuh+dP)YkZI_qdGc-R)'$_0<?E"6&l=l!'Do-HL"]-p+N&)pl$mB]tAq2;!&?%_GgWGHNRqEe,CY@El1`cbRbMVWrd&kE:t_A*7GE(0@KV_YDbfiGXX19ZYb%+K!=PLNp$QQQP^_XA/7;,EO`Vi.G-J79i8[E&Fa[g>tbO,.P8)_m&HCp<9(i,Ejm;L&7oXW'N3odO1MK`cS3Is_Ap5G'%s"Wh!As#g5PXqNaOY#ZJ[>X=:NAp/-$JM*HB[Br%_TYgjF\lc?WXjQfS#:uE<V=gBOTk$G8A!eY"JUB!fF4H_n@m:3rAdt7A:W'0A%<-'-"=(rG)CssmAR"k)0Y]GV@FQ2Gp7hJo-qcj+n&kGfW*8AJ`/1.`Qq+Vc['7p92Z'D@?+L'CCW2Oa6^H"NYI8GWR3Kf(V+=5=W*Fm%chC9`j-1"9F<m(^(:XX-fVG-kKbehl-ETXrnogE+fSCSLJ]Y3'K(83Ce-,\aRH>^mKOGECG+Wdd`URi<f.p[oeTFcIg-<3X5uQroOu3j@2)DZ+ROE=mY"E*N1H"2"BAs8Xl[A,=Ka:GL>OTL78*(YR8J.e1j[Cp!EHqHRpi(W[g7G>>G(4mg4$6^O/,^='j8rNRWgN03VX\3%Wo`2i45>WMdd9?D*X\@n8knnZDQD:7B@TUWeVe*`.pLNJQ`AQ@^f2gG]#gts=#4$gr6]$k[Cn@`;h40,TIli.];R2adE[-=\(toL`[Aj`IbQQQFeq!)$B93u<GeS/M7@'&iYGJ302d?-Q[i%.&6X0Pd/6@JQGu0(:MIO8,>[_4p'slonJu#].\kF\ZK^akbo3$$;/N&&D:TIth)G"`"3kfMAtU`GoD>E-_m7QURJ0ps'+QqW#Y1A/hd6:A4rqH5QQL/'muLglTN,/OUI/$QJS+U-3WWp"H`EoN+$5iT'=>=qB\?u`(NrUCP"Km!&Ai[sc?@A6Ut\fqT\,S4ilWUW6='0sNLFYK;AD"*<kDCGZ?`GR@O,!)&QIR6[)@'8lZ?,rT4HDMDP.\TRY[J`;']_\b=/$!.BPOl+gQr$UHMQcA#OU6XTHsb_c;>fTa@SMe;Zf<:<NJ=C3pSi?1H^.UEZfDKP=bZ]@B)?<cVEULrqbSCsG^&V7B<F)XWfi2HAM*[G7EUr+$FCh'KErG(*duTEjI3=f(fN3rLT`e6DiF2`7RNG50+HS_Zled8N/%%B>K_3Xd6k6`@#n&#ho51,_mmq-*Ke/r`4^lCC0l&n-b;fG&>f4oRQ+c.dnL.f2,<^$^h_*KV9.ZZ^aqKdq*E,H@acBT)0RVJ-]qDJkb!@O:iaW(VV9EoA;La/(?uSDlF0#sjao-0>>0amuu0Cf\(6g%)pE7Qg>E9$7D;ru?o\?"bldDpo,*BqA9F`cfDG<-^56`RUa9PsnWf8(F]"r#J"YbAZCt^H]5%]d3FAs)S(SV;T,US!)[?(K]\e#uh??$3KRBdkX5<'g42J[=$H/It\`J,W<:ul\*Bh./P:<]1V-;I?GF^P'fX&8!:cU'FaVt_j84GC<0gVU'!:6YU/7Z)KOJni(sm8W?L/&m#gA5-2@X#jKKr8JJrR]P/olt3G_9#TS=Gp,WJa*G&#)W$%!pXCe9#IL)dWI3`&0R>7r!U+bqWU%FgJ,9u99H`'$K*6]8fEMit:5ogu5`6b^+_-]q.5q0HA>KKD'<d%X@_U<q8]:sYl_d%X@_U<q8]:sYl_d%X@_U<q8]:sYkj#1^&XTfMpsUE8-2DgmJjgO%H/PA6`*Q0W9dd"BUk@,F1/8P*h**M$Tt+r^$,h@=l(J=Qa<6e?sN:9!.G91d]aC[/t&"$ZVnk21;4~>endstream
+endobj
+10 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 847
+>>
+stream
+Gat%!?$"^R&:EYBggccRq=JCeOT\5>:d^2jp6AP)-52QnU!q,umod)PP@0QH_9C`<<d-&;F$LZ!4kEngZQ,\G#;..V!*;i[5(dk3qpXR7;d*cGN='-'=V<6R7O[<<Xl7'KeA<\C8obeB."CsYBNcrPF3Q7N%0N1e=hs0a,cf<hZoUT`<`U;M[ZVqbS"R.*:Qjj??-44`\:t1*Y$]Ru91e$D!CPNlS;^L\iWJ7#OIn_7#\L0Bgb[DZ=fur"pNMDW,*=Fb>'bKqC`pDUba"u`S>u9O']6*<LT*mbAooH\]I\MR;Q0^SW'0lkFh>4h\%>#Jo_.cap*f#Z\;a<R!ek@1lq^@FM"]19R.^kL*(!tGKa!ZT]p#M91`Bq&_PP[cd).-V,+GJ*fC<Ymh/>4'lN/,NcLIWVK\q:&orln5]:DS5LN]>i]pA*=!A0ZmET4O7cNQ)B%VHN#s,!1l!07lRZ1_p)@im"pe(J`J5T!!*pbmJQI':4oj>\)3F'&jK,FGA_s"_khfl3)bjs(^4kATjp0(OgD9c(BZeEc7>W<Ml325O8Zg107kn+'5gf=ENJ=:RNT-Kg1uBaiG[\#Uf+kHh4:3_u)paYY$`@.?7,K84TO[8P%BN%NN/AWV$/_i_pDJPU+1-a@Qq]pU`^/+M'2b+hc)^$.9->jp!fZ\om<,6%C"I_3@0.Ct50^Y-(RmPF&>Pcka/P%24j#T7Hr@6-Tsj3r]ho!S!,mVj8%j[:*?>N5,`"^DDV\?`!O6aXlI?11^Oq^CAp`U96XfEX!PkbZi[Rj&B)8.2*S=[1PV-aXg2^ihX;%eX>DR6Bo[a*'*:p>oebPm4mC"+'tJO<;F^~>endstream
+endobj
+xref
+0 11
+0000000000 65535 f 
+0000000061 00000 n 
+0000000102 00000 n 
+0000000209 00000 n 
+0000000321 00000 n 
+0000000524 00000 n 
+0000000728 00000 n 
+0000000796 00000 n 
+0000001076 00000 n 
+0000001141 00000 n 
+0000005319 00000 n 
+trailer
+<<
+/ID 
+[<c29a6a2e4bbaf2b532c0617c2451c2f8><c29a6a2e4bbaf2b532c0617c2451c2f8>]
+% ReportLab generated PDF document -- digest (opensource)
+
+/Info 7 0 R
+/Root 6 0 R
+/Size 11
+>>
+startxref
+6257
+%%EOF

samples/sample_digital_truth.json

@@ -0,0 +1,322 @@
+[
+  {
+    "date": "2026-04-01",
+    "narration": "UPI/CR/512345678901/RAMESH TRADERS/PAYTM/Payment for goods",
+    "ref_no": "512345",
+    "debit": null,
+    "credit": 45000.0,
+    "balance": 197340.5
+  },
+  {
+    "date": "2026-04-01",
+    "narration": "NEFT-HDFC0001234-ACME PVT LTD-INV 4521",
+    "ref_no": "000123",
+    "debit": null,
+    "credit": 128500.0,
+    "balance": 325840.5
+  },
+  {
+    "date": "2026-04-02",
+    "narration": "UPI/DR/423451234567/SHARMA RENT/GPAY/April office rent",
+    "ref_no": "423451",
+    "debit": 35000.0,
+    "credit": null,
+    "balance": 290840.5
+  },
+  {
+    "date": "2026-04-02",
+    "narration": "ACH/NACH/SBILIFE INSURANCE/PREMIUM/POL 9921",
+    "ref_no": "9921",
+    "debit": 12500.0,
+    "credit": null,
+    "balance": 278340.5
+  },
+  {
+    "date": "2026-04-03",
+    "narration": "SALARY APR 2026 - STAFF PAYROLL BATCH 04",
+    "ref_no": "202604",
+    "debit": 86000.0,
+    "credit": null,
+    "balance": 192340.5
+  },
+  {
+    "date": "2026-04-03",
+    "narration": "ATM CSH WDL/AXIS/KANPUR MALL ROAD/04APR",
+    "ref_no": "04",
+    "debit": 10000.0,
+    "credit": null,
+    "balance": 182340.5
+  },
+  {
+    "date": "2026-04-04",
+    "narration": "UPI/CR/612345098765/SUNIL ENTERPRISES/PHONEPE/sale",
+    "ref_no": "612345",
+    "debit": null,
+    "credit": 22750.0,
+    "balance": 205090.5
+  },
+  {
+    "date": "2026-04-04",
+    "narration": "GST PMT-CBIC-26AAAPL1234C1Z5-GSTR3B APR",
+    "ref_no": "261234",
+    "debit": 18900.0,
+    "credit": null,
+    "balance": 186190.5
+  },
+  {
+    "date": "2026-04-05",
+    "narration": "NEFT-ICIC0004567-MEGHA SUPPLIERS-PO 7781",
+    "ref_no": "000456",
+    "debit": 64200.0,
+    "credit": null,
+    "balance": 121990.5
+  },
+  {
+    "date": "2026-04-05",
+    "narration": "INT.PD-SAVINGS A/C-QTR ENDING",
+    "ref_no": null,
+    "debit": null,
+    "credit": 1842.0,
+    "balance": 123832.5
+  },
+  {
+    "date": "2026-04-06",
+    "narration": "BANK CHRG-SMS CHGS APR 2026",
+    "ref_no": "2026",
+    "debit": 23.6,
+    "credit": null,
+    "balance": 123808.9
+  },
+  {
+    "date": "2026-04-06",
+    "narration": "UPI/DR/734512345098/BSNL BROADBAND/INTERNET BILL",
+    "ref_no": "734512",
+    "debit": 1499.0,
+    "credit": null,
+    "balance": 122309.9
+  },
+  {
+    "date": "2026-04-07",
+    "narration": "RTGS-UTIB0000123-DELTA EXPORTS-SETTLEMENT",
+    "ref_no": "000012",
+    "debit": null,
+    "credit": 215000.0,
+    "balance": 337309.9
+  },
+  {
+    "date": "2026-04-07",
+    "narration": "ACH/NACH/HDFC HOME LOAN/EMI/LN44219",
+    "ref_no": "44219",
+    "debit": 28750.0,
+    "credit": null,
+    "balance": 308559.9
+  },
+  {
+    "date": "2026-04-08",
+    "narration": "UPI/DR/845612340987/KESHAV ELECTRICALS/utility KESCO bill",
+    "ref_no": "845612",
+    "debit": 6540.0,
+    "credit": null,
+    "balance": 302019.9
+  },
+  {
+    "date": "2026-04-08",
+    "narration": "CSH DEP/BRANCH/KANPUR/CASH DEPOSIT BY CLIENT",
+    "ref_no": null,
+    "debit": null,
+    "credit": 50000.0,
+    "balance": 352019.9
+  },
+  {
+    "date": "2026-04-09",
+    "narration": "TDS PMT-CPC-TDS Q4 26AAAPL1234C",
+    "ref_no": "426123",
+    "debit": 9400.0,
+    "credit": null,
+    "balance": 342619.9
+  },
+  {
+    "date": "2026-04-09",
+    "narration": "UPI/CR/923451234560/VIKAS AGENCY/sale receipt",
+    "ref_no": "923451",
+    "debit": null,
+    "credit": 17800.0,
+    "balance": 360419.9
+  },
+  {
+    "date": "2026-04-10",
+    "narration": "NEFT-PUNB0123456-RAVI CONSULTING-PROF FEES",
+    "ref_no": "012345",
+    "debit": 25000.0,
+    "credit": null,
+    "balance": 335419.9
+  },
+  {
+    "date": "2026-04-10",
+    "narration": "IRCTC TRAVEL/RAIL TICKET/PNR 8841200012",
+    "ref_no": "884120",
+    "debit": 3450.0,
+    "credit": null,
+    "balance": 331969.9
+  },
+  {
+    "date": "2026-04-11",
+    "narration": "UPI/DR/103451234561/SWIGGY/staff lunch conveyance",
+    "ref_no": "103451",
+    "debit": 1280.0,
+    "credit": null,
+    "balance": 330689.9
+  },
+  {
+    "date": "2026-04-11",
+    "narration": "ELECTRICITY KESCO BILL PMT/CONS 4412",
+    "ref_no": "4412",
+    "debit": 8920.0,
+    "credit": null,
+    "balance": 321769.9
+  },
+  {
+    "date": "2026-04-12",
+    "narration": "UPI/CR/213451234562/ANJALI STORES/PHONEPE/sale",
+    "ref_no": "213451",
+    "debit": null,
+    "credit": 9650.0,
+    "balance": 331419.9
+  },
+  {
+    "date": "2026-04-12",
+    "narration": "AMC CHGS-DEBIT CARD ANNUAL FEE",
+    "ref_no": null,
+    "debit": 590.0,
+    "credit": null,
+    "balance": 330829.9
+  },
+  {
+    "date": "2026-04-13",
+    "narration": "NEFT-SBIN0009988-OWN ACCOUNT TRANSFER-SELF",
+    "ref_no": "000998",
+    "debit": 40000.0,
+    "credit": null,
+    "balance": 290829.9
+  },
+  {
+    "date": "2026-04-13",
+    "narration": "UPI/DR/313451234563/GARG REPAIRS/AC servicing maintenance",
+    "ref_no": "313451",
+    "debit": 4200.0,
+    "credit": null,
+    "balance": 286629.9
+  },
+  {
+    "date": "2026-04-14",
+    "narration": "INCOME TAX PMT-CBDT-ADVANCE TAX 26AAAPL",
+    "ref_no": "26",
+    "debit": 35000.0,
+    "credit": null,
+    "balance": 251629.9
+  },
+  {
+    "date": "2026-04-14",
+    "narration": "UPI/CR/413451234564/MOHIT TRADING/sale",
+    "ref_no": "413451",
+    "debit": null,
+    "credit": 31200.0,
+    "balance": 282829.9
+  },
+  {
+    "date": "2026-04-15",
+    "narration": "LIC OF INDIA/ACH/PREMIUM/POLICY 5512",
+    "ref_no": "5512",
+    "debit": 9800.0,
+    "credit": null,
+    "balance": 273029.9
+  },
+  {
+    "date": "2026-04-15",
+    "narration": "UPI/DR/513451234565/JIO RECHARGE/mobile postpaid",
+    "ref_no": "513451",
+    "debit": 799.0,
+    "credit": null,
+    "balance": 272230.9
+  },
+  {
+    "date": "2026-04-16",
+    "narration": "NEFT-YESB0001122-GLOBE IMPORTS-PURCHASE",
+    "ref_no": "000112",
+    "debit": 52400.0,
+    "credit": null,
+    "balance": 219830.9
+  },
+  {
+    "date": "2026-04-16",
+    "narration": "DRAWINGS-PROPRIETOR-PERSONAL WITHDRAWAL",
+    "ref_no": null,
+    "debit": 30000.0,
+    "credit": null,
+    "balance": 189830.9
+  },
+  {
+    "date": "2026-04-17",
+    "narration": "UPI/CR/613451234566/CAPITAL INTRODUCED BY PARTNER",
+    "ref_no": "613451",
+    "debit": null,
+    "credit": 200000.0,
+    "balance": 389830.9
+  },
+  {
+    "date": "2026-04-17",
+    "narration": "INT CREDIT-FIXED DEPOSIT INTEREST",
+    "ref_no": null,
+    "debit": null,
+    "credit": 4521.0,
+    "balance": 394351.9
+  },
+  {
+    "date": "2026-04-18",
+    "narration": "UPI/DR/713451234567/PETROL HPCL/fuel conveyance",
+    "ref_no": "713451",
+    "debit": 2500.0,
+    "credit": null,
+    "balance": 391851.9
+  },
+  {
+    "date": "2026-04-18",
+    "narration": "NEFT-KKBK0007788-ZENITH PVT LTD-INV 9921",
+    "ref_no": "000778",
+    "debit": null,
+    "credit": 76300.0,
+    "balance": 468151.9
+  },
+  {
+    "date": "2026-04-19",
+    "narration": "ATM CSH WDL/HDFC/CIVIL LINES/22APR",
+    "ref_no": "22",
+    "debit": 15000.0,
+    "credit": null,
+    "balance": 453151.9
+  },
+  {
+    "date": "2026-04-19",
+    "narration": "UPI/DR/813451234568/PRINTWELL/office stationery purchase",
+    "ref_no": "813451",
+    "debit": 3120.0,
+    "credit": null,
+    "balance": 450031.9
+  },
+  {
+    "date": "2026-04-20",
+    "narration": "INTEREST PAID ON OVERDRAFT-OD A/C",
+    "ref_no": null,
+    "debit": 5640.0,
+    "credit": null,
+    "balance": 444391.9
+  },
+  {
+    "date": "2026-04-20",
+    "narration": "RTGS-HDFC0009090-NEXUS SOLUTIONS-FINAL PMT",
+    "ref_no": "000909",
+    "debit": null,
+    "credit": 98000.0,
+    "balance": 542391.9
+  }
+]

samples/sample_scan_truth.json

@@ -0,0 +1,322 @@
+[
+  {
+    "date": "2026-04-01",
+    "narration": "UPI/CR/512345678901/RAMESH TRADERS/PAYTM/Payment for goods",
+    "ref_no": "512345",
+    "debit": null,
+    "credit": 45000.0,
+    "balance": 197340.5
+  },
+  {
+    "date": "2026-04-01",
+    "narration": "NEFT-HDFC0001234-ACME PVT LTD-INV 4521",
+    "ref_no": "000123",
+    "debit": null,
+    "credit": 128500.0,
+    "balance": 325840.5
+  },
+  {
+    "date": "2026-04-02",
+    "narration": "UPI/DR/423451234567/SHARMA RENT/GPAY/April office rent",
+    "ref_no": "423451",
+    "debit": 35000.0,
+    "credit": null,
+    "balance": 290840.5
+  },
+  {
+    "date": "2026-04-02",
+    "narration": "ACH/NACH/SBILIFE INSURANCE/PREMIUM/POL 9921",
+    "ref_no": "9921",
+    "debit": 12500.0,
+    "credit": null,
+    "balance": 278340.5
+  },
+  {
+    "date": "2026-04-03",
+    "narration": "SALARY APR 2026 - STAFF PAYROLL BATCH 04",
+    "ref_no": "202604",
+    "debit": 86000.0,
+    "credit": null,
+    "balance": 192340.5
+  },
+  {
+    "date": "2026-04-03",
+    "narration": "ATM CSH WDL/AXIS/KANPUR MALL ROAD/04APR",
+    "ref_no": "04",
+    "debit": 10000.0,
+    "credit": null,
+    "balance": 182340.5
+  },
+  {
+    "date": "2026-04-04",
+    "narration": "UPI/CR/612345098765/SUNIL ENTERPRISES/PHONEPE/sale",
+    "ref_no": "612345",
+    "debit": null,
+    "credit": 22750.0,
+    "balance": 205090.5
+  },
+  {
+    "date": "2026-04-04",
+    "narration": "GST PMT-CBIC-26AAAPL1234C1Z5-GSTR3B APR",
+    "ref_no": "261234",
+    "debit": 18900.0,
+    "credit": null,
+    "balance": 186190.5
+  },
+  {
+    "date": "2026-04-05",
+    "narration": "NEFT-ICIC0004567-MEGHA SUPPLIERS-PO 7781",
+    "ref_no": "000456",
+    "debit": 64200.0,
+    "credit": null,
+    "balance": 121990.5
+  },
+  {
+    "date": "2026-04-05",
+    "narration": "INT.PD-SAVINGS A/C-QTR ENDING",
+    "ref_no": null,
+    "debit": null,
+    "credit": 1842.0,
+    "balance": 123832.5
+  },
+  {
+    "date": "2026-04-06",
+    "narration": "BANK CHRG-SMS CHGS APR 2026",
+    "ref_no": "2026",
+    "debit": 23.6,
+    "credit": null,
+    "balance": 123808.9
+  },
+  {
+    "date": "2026-04-06",
+    "narration": "UPI/DR/734512345098/BSNL BROADBAND/INTERNET BILL",
+    "ref_no": "734512",
+    "debit": 1499.0,
+    "credit": null,
+    "balance": 122309.9
+  },
+  {
+    "date": "2026-04-07",
+    "narration": "RTGS-UTIB0000123-DELTA EXPORTS-SETTLEMENT",
+    "ref_no": "000012",
+    "debit": null,
+    "credit": 215000.0,
+    "balance": 337309.9
+  },
+  {
+    "date": "2026-04-07",
+    "narration": "ACH/NACH/HDFC HOME LOAN/EMI/LN44219",
+    "ref_no": "44219",
+    "debit": 28750.0,
+    "credit": null,
+    "balance": 308559.9
+  },
+  {
+    "date": "2026-04-08",
+    "narration": "UPI/DR/845612340987/KESHAV ELECTRICALS/utility KESCO bill",
+    "ref_no": "845612",
+    "debit": 6540.0,
+    "credit": null,
+    "balance": 302019.9
+  },
+  {
+    "date": "2026-04-08",
+    "narration": "CSH DEP/BRANCH/KANPUR/CASH DEPOSIT BY CLIENT",
+    "ref_no": null,
+    "debit": null,
+    "credit": 50000.0,
+    "balance": 352019.9
+  },
+  {
+    "date": "2026-04-09",
+    "narration": "TDS PMT-CPC-TDS Q4 26AAAPL1234C",
+    "ref_no": "426123",
+    "debit": 9400.0,
+    "credit": null,
+    "balance": 342619.9
+  },
+  {
+    "date": "2026-04-09",
+    "narration": "UPI/CR/923451234560/VIKAS AGENCY/sale receipt",
+    "ref_no": "923451",
+    "debit": null,
+    "credit": 17800.0,
+    "balance": 360419.9
+  },
+  {
+    "date": "2026-04-10",
+    "narration": "NEFT-PUNB0123456-RAVI CONSULTING-PROF FEES",
+    "ref_no": "012345",
+    "debit": 25000.0,
+    "credit": null,
+    "balance": 335419.9
+  },
+  {
+    "date": "2026-04-10",
+    "narration": "IRCTC TRAVEL/RAIL TICKET/PNR 8841200012",
+    "ref_no": "884120",
+    "debit": 3450.0,
+    "credit": null,
+    "balance": 331969.9
+  },
+  {
+    "date": "2026-04-11",
+    "narration": "UPI/DR/103451234561/SWIGGY/staff lunch conveyance",
+    "ref_no": "103451",
+    "debit": 1280.0,
+    "credit": null,
+    "balance": 330689.9
+  },
+  {
+    "date": "2026-04-11",
+    "narration": "ELECTRICITY KESCO BILL PMT/CONS 4412",
+    "ref_no": "4412",
+    "debit": 8920.0,
+    "credit": null,
+    "balance": 321769.9
+  },
+  {
+    "date": "2026-04-12",
+    "narration": "UPI/CR/213451234562/ANJALI STORES/PHONEPE/sale",
+    "ref_no": "213451",
+    "debit": null,
+    "credit": 9650.0,
+    "balance": 331419.9
+  },
+  {
+    "date": "2026-04-12",
+    "narration": "AMC CHGS-DEBIT CARD ANNUAL FEE",
+    "ref_no": null,
+    "debit": 590.0,
+    "credit": null,
+    "balance": 330829.9
+  },
+  {
+    "date": "2026-04-13",
+    "narration": "NEFT-SBIN0009988-OWN ACCOUNT TRANSFER-SELF",
+    "ref_no": "000998",
+    "debit": 40000.0,
+    "credit": null,
+    "balance": 290829.9
+  },
+  {
+    "date": "2026-04-13",
+    "narration": "UPI/DR/313451234563/GARG REPAIRS/AC servicing maintenance",
+    "ref_no": "313451",
+    "debit": 4200.0,
+    "credit": null,
+    "balance": 286629.9
+  },
+  {
+    "date": "2026-04-14",
+    "narration": "INCOME TAX PMT-CBDT-ADVANCE TAX 26AAAPL",
+    "ref_no": "26",
+    "debit": 35000.0,
+    "credit": null,
+    "balance": 251629.9
+  },
+  {
+    "date": "2026-04-14",
+    "narration": "UPI/CR/413451234564/MOHIT TRADING/sale",
+    "ref_no": "413451",
+    "debit": null,
+    "credit": 31200.0,
+    "balance": 282829.9
+  },
+  {
+    "date": "2026-04-15",
+    "narration": "LIC OF INDIA/ACH/PREMIUM/POLICY 5512",
+    "ref_no": "5512",
+    "debit": 9800.0,
+    "credit": null,
+    "balance": 273029.9
+  },
+  {
+    "date": "2026-04-15",
+    "narration": "UPI/DR/513451234565/JIO RECHARGE/mobile postpaid",
+    "ref_no": "513451",
+    "debit": 799.0,
+    "credit": null,
+    "balance": 272230.9
+  },
+  {
+    "date": "2026-04-16",
+    "narration": "NEFT-YESB0001122-GLOBE IMPORTS-PURCHASE",
+    "ref_no": "000112",
+    "debit": 52400.0,
+    "credit": null,
+    "balance": 219830.9
+  },
+  {
+    "date": "2026-04-16",
+    "narration": "DRAWINGS-PROPRIETOR-PERSONAL WITHDRAWAL",
+    "ref_no": null,
+    "debit": 30000.0,
+    "credit": null,
+    "balance": 189830.9
+  },
+  {
+    "date": "2026-04-17",
+    "narration": "UPI/CR/613451234566/CAPITAL INTRODUCED BY PARTNER",
+    "ref_no": "613451",
+    "debit": null,
+    "credit": 200000.0,
+    "balance": 389830.9
+  },
+  {
+    "date": "2026-04-17",
+    "narration": "INT CREDIT-FIXED DEPOSIT INTEREST",
+    "ref_no": null,
+    "debit": null,
+    "credit": 4521.0,
+    "balance": 394351.9
+  },
+  {
+    "date": "2026-04-18",
+    "narration": "UPI/DR/713451234567/PETROL HPCL/fuel conveyance",
+    "ref_no": "713451",
+    "debit": 2500.0,
+    "credit": null,
+    "balance": 391851.9
+  },
+  {
+    "date": "2026-04-18",
+    "narration": "NEFT-KKBK0007788-ZENITH PVT LTD-INV 9921",
+    "ref_no": "000778",
+    "debit": null,
+    "credit": 76300.0,
+    "balance": 468151.9
+  },
+  {
+    "date": "2026-04-19",
+    "narration": "ATM CSH WDL/HDFC/CIVIL LINES/22APR",
+    "ref_no": "22",
+    "debit": 15000.0,
+    "credit": null,
+    "balance": 453151.9
+  },
+  {
+    "date": "2026-04-19",
+    "narration": "UPI/DR/813451234568/PRINTWELL/office stationery purchase",
+    "ref_no": "813451",
+    "debit": 3120.0,
+    "credit": null,
+    "balance": 450031.9
+  },
+  {
+    "date": "2026-04-20",
+    "narration": "INTEREST PAID ON OVERDRAFT-OD A/C",
+    "ref_no": null,
+    "debit": 5640.0,
+    "credit": null,
+    "balance": 444391.9
+  },
+  {
+    "date": "2026-04-20",
+    "narration": "RTGS-HDFC0009090-NEXUS SOLUTIONS-FINAL PMT",
+    "ref_no": "000909",
+    "debit": null,
+    "credit": 98000.0,
+    "balance": 542391.9
+  }
+]