Bank Statement Parser jẹ ile-ikawe Python orisun-ṣiṣi ti o ṣe itupalẹ awọn alaye banki lati awọn ọna kika meje (CAMT.053, PAIN.001, CSV, OFX, QFX, MT940, ati PDF) sinu pandas DataFrames ti iṣeto. Gbogbo sisẹ n ṣiṣẹ ni agbegbe — iṣelọpọ ipinnu, isọdọtun PII laifọwọyi, ati opo gigun PDF aladapọ ti o lo awọn LLM agbegbe nigbati o nilo.
Bẹrẹ ni Iṣẹju-aaya
pip install bankstatementparser
from bankstatementparser import create_parser, detect_statement_format
fmt = detect_statement_format("statement.xml")
parser = create_parser("statement.xml", fmt)
df = parser.parse() # pandas DataFrame, ready to use
# Parse PDFs with the hybrid pipeline (v0.0.5+)
from bankstatementparser.hybrid import smart_ingest
result = smart_ingest("statement.pdf")
print(result.source_method) # "deterministic" | "llm" | "vision"
print(result.verification.status) # VERIFIED | DISCREPANCY | FAILED
Ile-ikawe Kan, Awọn Ọna Kika Meje
Ṣe itupalẹ CAMT.053, PAIN.001, CSV, OFX, QFX, MT940, ati PDF sinu pandas DataFrames ti iṣeto pẹlu API iṣọkan kan. Ko si ye lati fi sori ẹrọ awọn idii lọtọ fun ọna kika kọọkan.
| Ẹya ara ẹrọ | Bank Statement Parser | OSS ọna kika ẹyọkan (mt940, ofxparse) | SaaS (Ocrolus, Parseur) |
|---|---|---|---|
| Awọn ọna kika ti a ṣe atilẹyin | 7, API iṣọkan | 1 kọọkan | Ọpọlọpọ (nipasẹ OCR) |
| Atilẹyin PDF | Opo gigun aladapọ (ipinnu + LLM + iran) | Rara | Bẹẹni (cloud OCR) |
| Aṣiri data | 100% agbegbe (awọn LLM n ṣiṣẹ ni agbegbe nipasẹ Ollama) | 100% agbegbe | Data rán si ita |
| Iye owo | Ọfẹ, Apache 2.0 | Ọfẹ | $49-$1,000+/oṣu |
| Iṣayẹwo iwọntunwọnsi | Ofin Goolu (iwọntunwọnsi ṣiṣi + awọn kirẹditi − awọn debiti = iwọntunwọnsi ipari) | Rara | O yatọ |
| Isọdọtun PII | Ti a ṣe sinu, titan nipasẹ aiyipada | Rara | O yatọ |
| Sisanwọle | Iranti ti o ni opin | Rara | N/A |
| REST API | Iṣẹ-kekere FastAPI ti a ṣe sinu | Rara | Bẹẹni |
| Yiyọ ẹda | Awọn hash idanimọ iṣowo aláìyípadà | Rara | Diẹ ninu |
| Okeere iwe iroyin | hledger + beancount | Rara | Rara |
Opo Gigun PDF Aladapọ
Bank Statement Parser v0.0.5+ pẹlu opo gigun aladapọ ọna-mẹta fun awọn alaye banki PDF:
- Ọna A (Ipinnu): Awọn tabili PDF ti iṣeto ti a ṣe itupalẹ taara — ọfẹ, yara julọ, ko nilo LLM.
- Ọna B (Ọrọ-LLM): Awọn PDF oni-nọmba pẹlu awọn eto eka ti a yọ jade nipasẹ LLM agbegbe (LiteLLM/Ollama).
- Ọna C (Iran-LLM): Awọn alaye ti a ṣe scan tabi fọtokopi ti a ṣe ilana pẹlu awọn awoṣe iran olopọ-iru.
A ṣe idanwo gbogbo isediwon pẹlu Ofin Goolu: opening balance + credits − debits == closing balance.
Ti a Ṣe fun Iṣilọ ISO 20022
SWIFT ti ṣeto awọn ọjọ ipari to duro ṣinṣin: gbogbo awọn ile-iṣẹ inawo gbọdọ gba CAMT.053 ṣaaju Oṣu kọkanla 2027, ati MT940/MT942/MT950 yoo fẹhinti ni kikun ṣaaju Oṣu kọkanla 2028. Bank Statement Parser ṣe itọju mejeeji MT940 atijọ ati awọn ọna kika ISO 20022 ode oni (CAMT.053, PAIN.001) ninu API kan, nitorinaa opo gigun itupalẹ rẹ yoo ṣiṣẹ lakoko iyipada ati lẹhin rẹ.
Iṣẹ
- 27,000+ iṣowo/iṣẹju-aaya fun itupalẹ CAMT.053
- 52,000+ iṣowo/iṣẹju-aaya fun itupalẹ PAIN.001
- < 2 ms akoko si abajade akọkọ
- Iranti igbagbogbo lati 1K si awọn iṣowo 50K+ nipasẹ sisanwọle
- Awọn idanwo 718 pẹlu agbegbe 100% kọja Python 3.10 si 3.14
Kilode ti Bank Statement Parser?
- Isediwon PDF Aladapọ:
smart_ingest()ṣe itọju awọn PDF oni-nọmba ati ti a ṣe scan pẹlu itọsọna aifọwọyi ati iṣayẹwo iwọntunwọnsi. - Iwari Ọna Kika Aifọwọyi:
detect_statement_format()ṣe idanimọ awọn faili laifọwọyi aticreate_parser()da parser to tọ padà. - Ikọkọ Ni Akọkọ: Isọdọtun PII wa ni titan nipasẹ aiyipada. Awọn LLM n ṣiṣẹ ni agbegbe nipasẹ Ollama — ko si data ti o kuro ninu ẹrọ rẹ.
- REST API: Ṣe imuṣiṣẹ gẹgẹbi iṣẹ-kekere FastAPI pẹlu awọn endpoint
/ingestati/health. - Imudara: Isọri iṣowo ti LLM ṣe pẹlu awọn ero ti a le ṣe atunṣe (aiyipada ẹka 13 Plaid).
- Okeere Iwe Iroyin: Okeere si awọn ọna kika iwe iroyin hledger ati beancount fun ṣiṣan iṣiro ọrọ-lasan.
- Ayẹwo Pupọ:
scan_and_ingest()ṣe ilana awọn igi folda pẹlu yiyọ ẹda laifọwọyi kọja awọn faili. - Owo Pupọ:
verify_balance_multi_currency()ṣe iṣayẹwo Ofin Goolu fun ẹgbẹ owo kọọkan. - Ṣetan fun Iṣelọpọ: Gbigbe ZIP to ni aabo, afọwọsi titẹ sii, idena ipa ọna, ati ipo atunyẹwo ifọrọwanilẹnuwo.
- Ijade Rọ: Okeere si CSV, JSON, Excel, Polars, hledger, tabi beancount.
- Sisẹ ni Afiwe: Ṣe itupalẹ awọn faili lọpọlọpọ ni akoko kanna pẹlu
parse_files_parallel().
Ti a Ṣe fun Iṣelọpọ
Bank Statement Parser jẹ apẹrẹ fun awọn ẹgbẹ ile-iṣura, awọn olupilẹṣẹ fintech, ati awọn oṣiṣẹ ibamu ti n ṣakoso data owo ifura. A lo ile-ikawe naa ninu awọn opo gigun iṣilọ MT940-si-CAMT, awọn eto ilaja adaṣe, gbigbe alaye PDF, ati ṣiṣan ṣiṣayẹwo ilana ofin kọja awọn ile-iṣẹ inawo.
- Awọn idanwo 718 pẹlu agbegbe 100% kọja Python 3.10 si 3.14
- Awọn igbẹkẹle ti SHA-256 hash tiipa pẹlu CycloneDX SBOM fun gbogbo idasilẹ
- Iṣelọpọ ipinnu — titẹ sii kanna ṣe awọn abajade baiti-kan-naa, gbogbo igba
- Apache 2.0 ni iwe-aṣẹ — lo larọwọto ninu awọn eto iṣowo ati inu
Ṣe o n ṣe ayẹwo awọn omiiran? Wo bi Bank Statement Parser ṣe afiwe ❯ | Ṣawari awọn ọran lilo gidi ❯