Update swedish bank registry (mdomke#108)

* Update script to fetch data from PDF * Add Dockerfile to generate updated Swedish list * Update list of Swedish banks
withplum · Nov 22, 2022 · 1b58ad4 · 1b58ad4
1 parent fb6f063
commit 1b58ad4
Show file tree

Hide file tree

Showing 3 changed files with 73 additions and 25 deletions.
diff --git a/schwifty/bank_registry/generated_se.json b/schwifty/bank_registry/generated_se.json
@@ -10,7 +10,7 @@
   {
     "country_code": "SE",
     "primary": true,
-    "bic": "AABAFI22",
+    "bic": "AABASESS",
     "bank_code": "230",
     "name": "\u00c5landsbanken",
     "short_name": "\u00c5landsbanken"
@@ -63,6 +63,14 @@
     "name": "Citibank (filial)",
     "short_name": "Citibank (filial)"
   },
+  {
+    "country_code": "SE",
+    "primary": true,
+    "bic": "FEMAMTMT",
+    "bank_code": "907",
+    "name": "Ferratum Bank",
+    "short_name": "Ferratum Bank"
+  },
   {
     "country_code": "SE",
     "primary": true,
@@ -98,7 +106,7 @@
   {
     "country_code": "SE",
     "primary": true,
-    "bic": "DABASESX",
+    "bic": "MARGSESS",
     "bank_code": "923",
     "name": "Marginalen Bank",
     "short_name": "Marginalen Bank"
@@ -127,6 +135,14 @@
     "name": "Resurs Bank AB",
     "short_name": "Resurs Bank AB"
   },
+  {
+    "country_code": "SE",
+    "primary": true,
+    "bic": "SWEDSESS",
+    "bank_code": "930",
+    "name": "Swedbank",
+    "short_name": "Swedbank"
+  },
   {
     "country_code": "SE",
     "primary": true,
@@ -178,11 +194,19 @@
   {
     "country_code": "SE",
     "primary": true,
-    "bic": "SWEDSESS",
+    "bic": "SPSDSE23",
     "bank_code": "957",
     "name": "Sparbanken Syd",
     "short_name": "Sparbanken Syd"
   },
+  {
+    "country_code": "SE",
+    "primary": true,
+    "bic": "BMPBSESS",
+    "bank_code": "958",
+    "name": "AION Bank",
+    "short_name": "AION Bank"
+  },
   {
     "country_code": "SE",
     "primary": true,
@@ -202,10 +226,10 @@
   {
     "country_code": "SE",
     "primary": true,
-    "bic": "NOFBSES1",
+    "bic": "NOFBSESS",
     "bank_code": "964",
-    "name": "Nordax Finans AB (publ)",
-    "short_name": "Nordax Finans AB (publ)"
+    "name": "Nordax Bank",
+    "short_name": "Nordax Bank"
   },
   {
     "country_code": "SE",
@@ -247,6 +271,14 @@
     "name": "Ekobanken",
     "short_name": "Ekobanken"
   },
+  {
+    "country_code": "SE",
+    "primary": true,
+    "bic": "NOHLSESS",
+    "bank_code": "975",
+    "name": "Northmill Bank",
+    "short_name": "Northmill Bank"
+  },
   {
     "country_code": "SE",
     "primary": true,

diff --git a/scripts/Dockerfile_se b/scripts/Dockerfile_se
@@ -0,0 +1,14 @@
+FROM python:3.10
+
+# How use this Dockerfile to generate the Swedish bank registry:
+
+# docker build -t schwifty_se -f scripts/Dockerfile_se .
+# docker run --rm -v `pwd`:/schwifty --workdir /schwifty schwifty_se
+
+RUN apt update -y \
+    && apt install -y ghostscript python3-tk libgl1 \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN pip install "camelot-py[base]" opencv-python
+
+CMD [ "python", "scripts/get_bank_registry_se.py" ]
diff --git a/scripts/get_bank_registry_se.py b/scripts/get_bank_registry_se.py
@@ -1,37 +1,39 @@
 #!/usr/bin/env python
 import json
 
+import camelot
 import pandas
 
-
-# https://www.swedishbankers.se/fraagor-vi-arbetar-med/betalningar/ny-nordisk-betalningsinfrastruktur/iban-och-svenskt-nationellt-kontonummer/
+# https://www.bankinfrastruktur.se/framtidens-betalningsinfrastruktur/iban-och-svenskt-nationellt-kontonummer
 URL = (
-    "https://www.swedishbankers.se/media/4863/"
-    "kalkylblad-i-iban-och-svenskt-nationellt-kontonummer-2021-02-16.xlsx"
+    "https://www.bankinfrastruktur.se/media/d1tlidv0/"
+    "iban-id-och-bic-adress-for-banker-2022-06-23.pdf"
 )
 
 
 def process():
-    registry = []
+    registry = {}
 
-    datas = pandas.read_excel(URL, skiprows=2, sheet_name=0, dtype=str)
-    datas.fillna("", inplace=True)
+    tables = camelot.read_pdf(URL, pages="1,2")
+    datas = pandas.concat([tables[0].df, tables[1].df], ignore_index=True)
 
+    datas.drop(index=datas.index[0], inplace=True)
+    datas.drop(index=datas.index[0], inplace=True)
+    datas.fillna("", inplace=True)
+    datas.sort_values(1, inplace=True)
     for row in datas.itertuples(index=False):
-        bank_code, bic, name = row[:3]
-        registry.append(
-            {
-                "country_code": "SE",
-                "primary": True,
-                "bic": str(bic).upper(),
-                "bank_code": str(bank_code).split(".", 1)[0],
-                "name": str(name).strip(),
-                "short_name": str(name).strip(),
-            }
-        )
+        bank_code, bic, name = row[1:4]
+        registry[str(bank_code).strip()] = {
+            "country_code": "SE",
+            "primary": True,
+            "bic": str(bic).upper(),
+            "bank_code": str(bank_code).strip(),
+            "name": str(name).strip(),
+            "short_name": str(name).strip(),
+        }
 
     print(f"Fetched {len(registry)} bank records")
-    return registry
+    return list(registry.values())
 
 
 if __name__ == "__main__":