pysru-mab

A small Python library and CLI for working with SRU files — the file format used by the Swedish Tax Agency (Skatteverket) for filing corporate income tax declarations (INK2 and its appendices), as defined in the SKV 269 technical specification.

pysru-mab cn:

• Parse BLANKETTER.SRU and INFO.SRU into typed Python objects.
• Write them back out, with byte-identical round-trip for unmodified data.
• Look up what a numeric field code (7011, 7104, ...) actually means, via a small built-in code table for INK2/INK2R/INK2S.
• A pysru-mab CLI for converting SRU files to/from JSON.

Install

pip install pysru-mab

Quick start

from pysru_mab import parse_file, describe

sru_file = parse_file("BLANKETTER.SRU")  # defaults to cp850 encoding

for blankett in sru_file.blanketter:
    print(blankett.form_id, blankett.namn)

    for uppgift in blankett.uppgifter:
        info = describe(blankett.form_name, uppgift.code)
        label = info["label"] if info else "?"
        print(f"  {uppgift.code} ({label}): {uppgift.value}")

Access individual fields:

ink2 = sru_file.find("INK2")[0]
ink2.get("7104")        # -> "-67222" (or None if absent)
ink2.get_all("7104")    # -> all values for that code, in order
ink2["7104"]            # -> "-67222" (raises KeyError if absent)

Writing / round-trip

from pysru_mab import parse_file, write_file

sru_file = parse_file("BLANKETTER.SRU")
# ... modify sru_file.blanketter[...] ...
write_file(sru_file, "BLANKETTER_OUT.SRU")

write_file emits CRLF line endings (as required by the SRU format). For unmodified parsed files, the writer also preserves whether the source had a trailing newline, so round-trips are byte-for-byte.

INFO.SRU

INFO.SRU uses a different, nested-section format and has its own parser/writer:

from pysru_mab import parse_info_file, write_info_file

info = parse_info_file("INFO.SRU")
print(info.orgnr, info.namn)

write_info_file(info, "INFO_OUT.SRU")

CLI usage

# Convert to JSON (optionally annotate fields with human-readable labels)
pysru-mab to-json BLANKETTER.SRU --describe -o blanketter.json

# Convert back to SRU
pysru-mab from-json blanketter.json -o BLANKETTER.SRU

# Look up what a field code means
pysru-mab describe INK2 7011
pysru-mab describe INK2          # list all known codes for a form

# INFO.SRU <-> JSON
pysru-mab info-to-json INFO.SRU -o info.json
pysru-mab info-from-json info.json -o INFO.SRU

If -o/--output is omitted, output goes to stdout.

Code tables

describe(form_name, code) looks up a field code against the bundled tables in pysru_mab/data/ (ink2.json, ink2r.json, ink2s.json). To add support for another form, drop in a new data/<form>.json file with the schema:

{
  "form": "INK2",
  "source": "SKV 269 (INK2, period 2025P4)",
  "codes": {
    "7011": {"label": "...", "description": "...", "field_type": "date"}
  }
}

field_type is one of "amount", "date", "checkbox", "text" (defaults to "amount"). You can also load a table from your own JSON file at runtime with load_custom_table(path).

Status: code tables are currently shipped for INK2, INK2R and INK2S covering the fields most commonly used. Labels are best-effort based on common SKV269/INK2 terminology — contributions and corrections against the official SKV269 appendix are welcome. Other forms parse and write correctly, but describe() returns None for them until a table is added.

Format notes

• Encoding: SKV269 specifies CP850/IBM850. cp850 is the default for all read/write functions, but it's configurable via the encoding keyword argument.
• Line endings: CRLF (\r\n) on output, while preserving whether the source had a trailing newline when round-tripping parsed files.
• Form IDs: #BLANKETT lines use IDs like INK2-2025P4, parsed into form_name="INK2", year=2025, period="P4".
• Values: #UPPGIFT values are kept as raw strings to preserve sign, leading zeros, and the "X" checkbox marker; use Uppgift.as_int() / Uppgift.is_checkbox for typed access.

License

MIT