Version 2.4.9 (Current)
October 2026
- Lookahead Balance Verification (Fixed Sr 82 & 83 Balance Distortion): Resolved single-digit OCR drops in closing balances causing cascading error amounts (e.g. ₹27,090 Dr and ₹37,000 Cr). Intelligent 1-3 step lookahead verification preserves genuine transaction amounts (Sr 82: ₹6,090 Dr; Sr 83: ₹16,000 Cr) and restores the unbroken running balance chain.
- Multi-Line Entry De-Clubbing (Recovered Sr 141/142 & 169/170): Fixed an issue where trailing reference numbers containing slashes or optical artifacts were mistaken for Value Dates with no trailing amounts. Smart backward date scanning cleanly separates Sr 141 (₹40,426 Dr) & Sr 142 (₹1,181 Dr), and Sr 169 (₹7,534 Dr) & Sr 170 (₹1,181 Dr), extracting all 190 out of 190 records.
- Anchored Decimal Balance Healer: Restricted trailing balance heuristics strictly to date-initiating rows, preventing reference strings (e.g.
PAYMENT FOR 104128) and pincodes from converting to spurious decimal amounts.
- Zero Discrepancy Parity: Total Debits (₹1,921,645.92 across 131 Dr) and Credits (₹1,956,889.40 across 59 Cr) match bank statement summary totals with ₹0.00 difference.
Version 2.4.6
October 2026
- Cheque Deposit (CHQ DEP) Classification: Fixed an issue where Cheque Deposits (e.g.,
08/04/26 Rs 52,817) were mistakenly classified as Debit withdrawals due to the word CHQ matching withdrawal logic. CHQ DEP, DEP, and DEPOSIT are now prioritized as Credit / Deposit across both HDFC and Universal parsers.
- CBDT Reference Number & Optical Decimal Healing: Fixed an optical typo where trailing punctuation on reference tokens (e.g.
0260430110042504.) was merged with the adjacent date (30/04/26), corrupting the transaction amount into a spurious 16-digit debit amount (026,04,30,11,00,42,504.3). The correct debit amount Rs 3,359.00 is now cleanly extracted.
- Auto-Detect Bank Robustness (190/190 Records Extracted): Enhanced HDFC signature detection in OCR mode so scanned statements consistently trigger the dedicated HDFC parser (extracting all 190 records with zero omissions) instead of falling back to the Universal parser.
Version 2.4.5
October 2026
- 100% OCR Record Extraction (All 190 Records Recovered): Resolved the remaining missing entries in OCR mode. Repaired date optical misread healers for single-digit months (e.g.,
2200526 → 22/05/26, 2000726 → 20/07/26, 3000726 → 30/07/26) and fixed optical balance double-dots (40.26.18 → 40,26.18), ensuring all 190 out of 190 statement entries are cleanly extracted.
- Date-Wise Sorting & MM/DD/YY Resolution: Fixed an issue where statement header date ranges with 4-digit years (such as
04/2026) were corrupted into fake 04/20/26 dates that sorted to the bottom of the statement after October. Added date parsing safeguards that automatically normalize and swap reversed month/day inputs (when month > 12), ensuring chronological sorting is always accurate without displaced bottom entries.
- Multi-Page Header Hygiene: Enforced complete removal of recurring statement header lines (
Statement From : ... To : ...) across all pages, preventing any metadata residue from leaking into transaction narrations or dates.
Version 2.4.4
October 2026
- OCR Mode Extraction Repair: Resolved the critical issue where OCR mode clubbed transactions into previous narrations. Individual transaction rows are recognized independently with resilient row-bucketing.
- Page Summary Exclusion: Statement summaries (Dr Count, Cr Count, Total Debits, Total Credits, Closing Balance, Generated By/On metadata) at page bottoms are strictly excluded from regular entries.
- OCR Healers for Optical Misreads: Automatically repairs OCR misread dates, reference number artifacts, and decimal numbers.
Version 2.4.3
October 2026
- Full Multi-Page Parsing Fix: Resolved an issue where only the 1st page of multi-page bank statements was converted. All pages are now thoroughly parsed across statements of any length (tested on 16 to 30+ pages).
- Non-Greedy Page Header Removal (HDFC): Corrected header stripping regex on HDFC multi-page statements that previously erased intermediate page transactions. Top headers on pages 2+ now cleanly strip only metadata without truncating transaction rows.
- Page Footer Bypass (BOI, PNB, Kotak): Replaced premature loop termination (
break) on repeated page footer notices (such as computer-generated disclaimers) with safe skips (continue), ensuring every subsequent page is parsed.
Version 2.4.2
September 2026
- Interactive Password Prompt Modal: Added a dedicated modal dialog that automatically prompts for the password whenever a password-protected / encrypted PDF bank statement is uploaded.
- Real-Time Validation & Retry Feedback: Seamlessly handles both initial password requests (
NEED_PASSWORD) and incorrect password attempts (INCORRECT_PASSWORD) with clear error messages without freezing or crashing.
- Show/Hide Password & Bank Hints: Includes an eye toggle to view entered passwords and hints for standard bank password formats (PAN + DOB, customer ID, account numbers).
- Dual Compatibility: Unlocked password-protected statements seamlessly integrate with both direct digital PDF text parsing and the Tesseract.js OCR rendering engine.
Version 2.4.1
September 2026
- Intelligent Page Header & Footer Sanitization: Automatically detects and removes bank address, IFSC/MICR codes, customer ID, account numbers, branch metadata, and legal disclaimers from transaction narrations across all bank statement formats (HDFC, Kotak, BOI, PNB, Indian Bank, ICICI, Universal).
- Cross-Page Continuation Stitching: Multi-line transaction narrations spanning across page boundaries are cleanly preserved and stitched together without injecting intervening page headers or footer notices.
- Pristine Accounting Export: Ensures that Tally XML vouchers (<NARRATION>), single/double-entry Excel, and DBF exports contain pure, unpolluted transaction narrations.
Version 2.4.0
September 2026
- Optical Character Recognition (OCR) Engine: Client-side Tesseract.js WebAssembly engine reads scanned PDF statements, image-only PDFs, and statement scans (PNG, JPG, JPEG) directly in the browser with zero server installation requirements.
- Automatic Scanned PDF Fallback: If an uploaded PDF has 0 selectable text characters, the app automatically switches to OCR mode, renders pages at high DPI, and extracts tabular statement data seamlessly.
- Dedicated OCR Mode Toggle: "OCR Mode (Scanned/Images)" toggle allows forcing OCR on poor quality or corrupted digital PDFs.
- OCR Text Healing & Normalization: Pre-parser cleaning pipeline fixes optical misrecognitions in amounts (spaces around commas/decimals, 0/O ambiguities) and broken date formats.
- Date Range Filtering & Export: Filter transactions by custom From Date and To Date with auto-detected date span boundaries and date-filtered Tally XML / Excel / DBF export.
Version 2.3.0
September 2026
- Date range filtering bar with auto-detected min/max boundaries and one-click clear.
- Date-filtered selective export to Tally XML, Single-Row Excel, Double-Entry Excel, and RECPAY1.DBF.
- Real-time financial summary recalibration for filtered date periods.
Version 2.2.0
September 2026
- HDFC statement parser with full negative amount reversals & negative overdraft balance support.
- Two-tier bank auto-detector suppressing false positives from narration bank mentions.
- Interactive instant re-parsing from in-memory text cache when changing bank format dropdown.
Version 2.1.0
September 2026
- Universal Fallback Parser for unlisted Indian banks (SBI, Axis, BOB, Canara, Union, Federal, Yes Bank, etc.).
- DRM-protected PDF auto-unlock pipeline with automatic RC4 permission stripping.
- Mathematical balance solver determining exact Debit/Credit directions across negative balances.
- Target bank ledger auto-fill and visual reading order coordinate text healing.
Version 2.0.0
September 2026
- Direct RECPAY1.DBF sync with auto-creation, date-range removal, and dBASE ZAP table support.
- Dual Excel Exports: Single-Row Standard Excel & Double-Entry Accounting Excel format.
- Tally Prime / ERP 9 XML generator for Receipts, Payments, and Contra vouchers.
- Dedicated parsers for ICICI, Indian Bank, HDFC, PNB, Kotak Mahindra, and Bank of India.
Designed & Developed by Alok Nagar (LogicTool.in)
Support: +91 9196144501
