v0.11.0

Changelog

[0.11.0] - 2024-09-05

⛰️ Features

• (banks/hsbc) Add support for non-OCR credit statements

🛠️ Bug Fixes

• (write) Incorrect 'base' statement type in final result

🚜 Refactor

• (pdf) Make PdfDocument a child class of fitz.Document
• (pdf) Use file_path as first arg to PdfDocument
• (pipeline) Move parser & handler creation logic to extract
• Pass PdfPages instead of parser
• (pipeline) Move bank detection logic to CLI
• (detector) Move detector to banks namespace
• Remove unnecessary usage of pydantic dataclasses
• (pdf) Add metadata identifier attr to PdfDocument
• (banks/base) Fix type hint for identifiers
• (pdf) Lazily import ocrmypdf
• (pdf) Perform ocr based on metadata identifiers
• (pipeline) Move parser instantiation logic to CLI
• (pipeline) Allow custom document to be passed

📚 Documentation

• Remove false version from changelog
• (README) Add note about OCR feature

⚙️ Miscellaneous Tasks

• (generic) Add GenericParserError
• Remove unused import
• (pdf) Remove old get_byte_stream function
• Remove old mock_document fixture
• (constants) Remove case insensitive modifier from formats with no words
• (pdf) Improve ocrmypdf performance
• (pipeline) Shorten create_handler function signature
• (generic) Move GenericBank to generic init
• (pipeline) Import Transaction from statements namespace
• Rename generic/generic_handler to generic/handler
• Import from pymupdf instead of fitz
• Linting for ocr changes

Build

• (deps) Bump the deps group with 7 updates
• (deps) Add ocrmypdf as a system dependency
• (deps) Move ocrmypdf to extras