refactor: decompose Statement constructor into pure factory + persistence

[boscorat]

Parent issue

#194 — function purification

Problem

Statement.__init__ (statements.py:347) is a monolithic constructor that:

• Opens the PDF file (pdf_open)
• Validates/initialises the project directory (validate_or_initialise_project)
• Reads config from disk (ImportConfigManager)
• Runs the full extraction pipeline (get_results)
• Writes debug files (_write_debug_json, _write_debug_excel)
• Prints errors to stderr
• Never raises, masking errors via self.success/self.error_message

Testing any single step requires real PDFs and a real project directory.

Proposal

Split into a pure factory/classmethod parse(file, config, ...) -> StatementResult that returns a dataclass with all extracted fields, and a separate side-effecting persist() method for file I/O and debug writes. The validation logic (is_successfull()) and debug writing should be decoupled from construction.

Code relationships

Target:

• Statement.__init__() — statements.py:347

Callers (will need updating):

• process_pdf_statement() — statements.py:828 (creates Statement)
• StatementBatch.debug() — statements.py:1682 (re-creates Statement objects)

Related (called from init):

• validate_or_initialise_project() — paths.py
• pdf_open() — pdf_functions.py:42
• ImportConfigManager.__init__() — import_config.py:147
• Statement.get_results() — statements.py:631
• _write_debug_json() — statements.py:198

Existing tests:

• tests/test_statements.py::TestGoodStatements — integration tests create Statement via StatementBatch (indirect coverage only)
• tests/conftest.py — session fixtures use StatementBatch which creates Statements

No direct unit tests exist for Statement.__init__.

Siblings

• refactor: BatchMetadata dataclass to replace 13 duplicated parameters #195 — BatchMetadata dataclass
• refactor: ExtractionContext dataclass for extraction pipeline #196 — ExtractionContext dataclass
• refactor: decompose process_pdf_statement into pure extraction + persistence #198 — process_pdf_statement decomposition
• refactor: stop mutating inputs in extraction pipeline #199 — Stop mutating inputs
• refactor: add optional conn parameter to build_datamart and related functions #200 — DB connection threading
• refactor: move processing out of StatementBatch constructor #201 — StatementBatch constructor
• refactor: remove dead logs and file_path parameters from pipeline functions #202 — Dead parameter cleanup
• refactor: Parquet class accepts pre-loaded data instead of reading from disk #203 — Parquet class purification
• refactor: _handle_parquet_write_error returns error dict instead of mutating #204 — _handle_parquet_write_error purification