- 🐛 fix split feature adding additional pages to output files
- ✨ add support for crop extraction
- ✨ add support for split extraction
- ✨ add support for minimal V2 CLI
- 🐛 fix file saves from crop/split extractors breaking in some instances
- ♻️ fix cron
- ✨ add job information to the inference object
- ✨ add completed_at property to job
- 🐛 fix missing class property
- ✨ add support for V2 Split, Classification, Crop, and OCR utilities
- ✨ add object field convenience accessors
- ✨ allow use of str, dict, or object to create a data schema replace
- ✨ add support for text_context on v2
- ✨ add support for data schema replace on v2
- ⚰️ remove unused feedback calls from CLI
- 🏷️ better field typing for v2
- ✨ add RAG metadata to v2 inference response
- ✨ improved v2 error messages and exceptions
- ♻️ migrate to ruff for linting
- ✨ add support for python 3.14
- ♻️ harmonize getting page count from a local input source
- ✨ allow comparing v2 field confidence
- ⚰️ remove obsolete US Mail V2
- 🐛 fix for multiple webhooks
- 🐛 inference options should default to None
- ✨ add list field typed accessors
- ♻️ rework polygons, use them more like a real class
- ✨ add inference options
- ♻️ 💥 update raw text output from server
- ♻️ use direct import of input classes
- 🐛 fix broken location & confidence data for V2 fields
- ♻️ make importing v2 classes easier
- ✨ add support for page count, mimetype + fixes
- 🐛 fix jobs not deserializing webhook information into their proper class
- 🐛 fix typo in
InferenceFileclass:alais=>alias
- ✨ add support for URL inputs for V2
- ✨ add generic response loader
- ✨ Add support for mindee API V2 client & features
- 🔧 Tweak CI & testing
- ♻️ Uniformize variable naming across files
- 💥 remove support for Python 3.7
- ✨ merge enqueue parameters into a single object
- ✨ add dot access to object fields
- ♻️ remove deprecated files
- 🎨 touch up string representation for most objects.
- ✨ add support for mindee API V2 client & features
- ♻️ remove support for python 3.7
- 🔧 tweak CI & testing
- ♻️ uniformize variable naming across files
- ✨ add support for address fields
- ✨ add support for Financial Document V1.12
- ✨ add support for Invoices V4.10
- ✨ add support for US Healthcare Cards V1.2
- ✨ add support for workflow polling
- 🐛 fix for Python 3.10 and lower: server uses 'Z' in date string
- ✨ add support for RAG parameter in workflow executions
- ✨ add support for Financial Document V1.12
- ✨ add support for Invoices V4.10
- ✨ add support for US Healthcare Cards V1.2
- 🐛 fix for null objects in extras
- ✨ update structure for InvoiceSplitterV1
- ✨ update FR EnegryBillV1 to V1.2
- ✨ update US HealthcareCardV1 to V1.1
- ⚰️ remove support for EU Driver License
- ⚰️ remove support for PetrolReceiptV1
- ⚰️ remove support for ReceiptV4
- ⚰️ remove support for Proof of Address
- ⚰️ remove support for US Driver License
- ⚰️ remove support for US W9V1
- 🐛 fix polling waiting for a full timeout in case of parsing failure
- ♻️ add missing imports for some products
- ⬇️ downgrade pypdfium dependency due to very rare errors in local character extraction
- ✨ add support for image and pdf compression
- ♻️ increase async retry timers
- ✨ add support for us mail v3
- 🎨 linting fixes
- ✨ allow local downloading of remote sources
- ⚰️ remove support for (FR) Carte Vitale V1 in favor of French Health Card V1
- ⬆️ update pylint
- ✨ add support for workflows
- ✨ add support for French Health Card V1
- ✨ add support for Driver License V1
- ✨ add support for Payslip FR V3
- 🐛 fix circular imports error on install
- ⚰️ remove support for international ID V1
- ♻️ update import syntax
- ⬆️ loosen dependency pinning on requests
- 🐛 fix potential circular import issues
- ✨ add support for business cards V1
- ✨ add support for delivery note V1.1
- ✨ add support for indian passport V1
- ✨ add support for resume V1.1
- ♻️ adjust default values for async delays
- ⬆️ fully migrate project to pyproject.toml
- ✨ add support for Financial Document v1.10
- ✨ add support for Invoice v4.8
- ✨ add support for BillOfLadingV1
- ✨ add support for (US) UsMailV2
- ✨ add support for (FR) EnergyBillV1
- ✨ add support for (FR) PayslipV1
- ✨ add support for NutritionFactsLabelV1
- 🐛 fixed a bug that prevented longer decimals from appearing in the string representation of some objects
- 🐛 fixed a bug that caused non-table elements to unexpectedly appear truncated when printed to the console
- 🐛 fix full text ocr extra not properly parsing
- 📝 fix a few documentation errors & typos
- 🔧 updated CI dependencies
- ✨ add support for full text OCR extra
- ✨ add support for invoice splitter auto-extraction
- 📝 add sample code example for image splitter auto-extraction
- ♻️ refactor
image_extractionand change name toextraction - 📝 update documentation
- ♻️ fix many typos in product internals
- ♻️ add a few missing
__init__.pyfiles
- ✨ add support for Healthcare Card V1
- ✨ add support for Invoice V4.7
- ✨ add support for Financial Document V1.9
- ✨ add support for BooleanField
- ♻️ update company registration display format
- ♻️ switch default image save format to PNG for image extractor
- 🐛 fix invalid display when trying to cast null amount fields to string
- ✨ add support for complete multi-receipt extraction (#240)
- ✨ add support for OCR text print in CLI tool (#239)
- ♻️ fix miscellaneous typos (#241)
- ✨ add support for webhooks responses & HMAC validation
- ♻️ replaced PikePDF with PyPdfium2
- ✨ update receipt to 5.2 and financial document to 1.7
- ✨ add support for financial document v1.6 and invoice 4.6
- ✨ update Invoice to v4.5
- 🐛 fix invalid error code handling for some errors (#227)
- ♻️ update error handling to account for future evolutions
- 📝 update some documentation
- ✨ add support for Resume V1
- ✨ add support for EU Driver License V1
- ✨ add support for International ID V2
- 📝 fix miscellaneous documentation issues
- ⬆️ update invoices to v4.4
- ✨ add support for
raw_valuein string fields
- ✨ add support for International ID V1
- ✨ add support for Generated APIs
- ✨ add custom associated classes & namespace for Generated APIs
- 📝 update sample codes for default usage
- 📝 add documentation for Generated APIs
- 🐛 fix sphinx doc from mistakenly looking for static files
- 🐛 fix default async config to avoid timeouts on larger files
- 🐛 fix for build badges
- 🧪 optimize CI + add retry mechanism
- ♻️ homogenize typing & classes
- ⬆️ upgrade support for python 3.12
- ✨ add n_pages attribute to document
- 🧪 ✨ add experimental pdf-fixing utility
- 📝 fix invoice-splitter doc
- 🔧 fix wrongful instance variable assignments
- 🔧 rework custom internals & fix custom page_id
- 🐛 fix crashes when trying to parse from newer custom APIs
- 🐛 add missing internal imports for some products
- 🎨 💥 harmonize response types & syntax with other libraries
- 🎨 💥 change endpoint management & syntax
- 🎨 💥 move products to
productmodule
- ✨ add support for auto-poll asynchronous parsing
- ✨ add support for products with both sync & async modes
- ✨ add support for async custom products
- ✨ 📝 add auto-generated md doc
- ♻️ 📝 update rst doc
- ⬆️ update dependencies
- 🧪 add experimental support for feedback API
- ⬆️ improved support for custom Line Items reconstruction
- ⬆️ update error management system
- ♻️ update unit-tests
- ⬆️ implement regression testing
- ♻️ update CLI
- ♻️ re-organize geometry module
- ⚰️ remove support for ReceiptV3
- ⚰️ remove support for InvoiceV3
- 🐛 fix
raw_httpattribute displaying a python dict instead of raw JSON
- ⬆️ update
urllib&pillowdependencies
- ⬆️ update
urllib,charset-normalizer&packagedependencies
- ✨ add support for Multi Receipts Detector V1
- ✨ add support for Barcode Reader V1
- ✨ add support for US W9 V1
- ♻️ small internal tweaks to accomodate for new products
- ✨ add support for FR ID card v2
- ⬆️ GitHub CI and mypy updates
- ⬆️ update requests to 2.31
- 🐛 fix full_name for multiple given names
- 🐛 fix for parsing OCR response in CLI
- ✨ add basic support for line items
- ✨ add support for FR bank account details v2
- ⚰️ remove support for Shipping Container
- ✨ add OpenAPI GET request
- ✨ add OCR output in CLI
- ✨ add support for material certificate v1
- 🐛 fix potential problem where the words can be re-arranged by the user when calculating lines.
- 📝 add documentation for OCR extraction
- ✨ add support for financial documents v1.1
- ✨ add support for handling OCR return
- ♻️ update printing of receipt, invoice, financial doc
- ♻️ add a specific class for classifications, which are never None
- 🐛 fix for tax base amount
- 🐛 fix naming of job ID property
- ✨ add support for Invoice Splitter V1
- ✨ add support for asynchronous requests
- 🐛 fix EU imports, use short imports in tests
- ✨ add support for receipt v5
- ✅ use code samples for integration testing
- ✨ add an URL input source
- 📝 add proof of address documentation
- 🐛 fix file extension for sample code inclusion
- 🦺 make sure CI is run on various OSes
- ⬆️ general update to all dependencies
- ✨ add support for FR carte vitale v1
- ✨ add support for FR ID card v1
- ✨ add support for shipping container v1
- ✨ add EU license plate v1
- 📝 add sample code for all supported APIs
- 🐛 fix for null classification field on custom APIs
- ✨ add support for: financial document v1
- ✨ add support for: proof of address v1
The FinancialV1 class is now deprecated.
It's still usable with no code modifications but will be removed in a future release.
All users are encouraged to move to FinancialDocumentV1 which has much better performance.
Also field names in FinancialDocumentV1 now match those of Invoice and Receipt.
- sparkles: Add French carte grise v1 support
- ⬆️ Update pikepdf
- ✨ Add support for Python 3.11
- 🐛 Fix misleading docstrings
- ✨ add version option in CLI
- ♻️ harmonize printing of float values
- ✨ add support for Invoice 4.1
- ✨ add support for Receipt v4.1
- 📝 reorganize docs a bit
- ⬆️ update dependencies
- 🐛 fix for invoice v4 URL
- ✨ allow setting base URL from env
- ✨ add some helper functions for BBox
- ✨ Add orientation info on all pages
- ✨ add Cropper support
- ✨ allow setting timeout value from env
- ✨ Add Invoice V4 (clearer field names, line items) (#107)
- 🐛 page_n should always be set when available (#106)
- ✨ New PDF cut/merge system, allowing specifying exactly which pages to use.
- ♻️ PDF documents are no longer cut by default, use the
page_optionsparameter in theparsemethod. - ✨ Document (endpoints) are now versioned, providing better backward-compatible support.
- ✨ Pass the document class instead of a string to specify how to
parseinput sources. - ♻️ Some methods and parameters renamed for better clarity.
- ✨ Results from Custom documents are now deserialized into objects, rather than
dict.
- ✨ Add support for expense receipts V4.
- ♻️ minor improvements to geometry functions.
- 🐛 Make sure the user is specified when calling custom docs on CLI
- 🐛 Add default timeout of 120 seconds for endpoints.
- 🐛 don't print "None" when filename is empty
- 🗑️ deprecate setting singular and plural names for docs (#98)
- ✨ add x-axis geometry functions (#99)
- ✨ add getting bounding box for multiple polygons (#100)
- ✨ add support for classifications in custom docs
- ✨ allow setting only the
MINDEE_API_KEYenvironment variable - ⬆️ update pikepdf
- 🐛 never use mutable defaults in class definitions (#96)
- 💄 improve string output of documents
- ⬆️ general dependencies upgrade (#92)
- ✨ add functions for working with centroids (#93)
- ✨ Add bank checks documents (beta, US only) (#94)
- 🐛 📝 custom docs names refer to the API, not the type.
- 🐛 words should be separated when printing custom documents
- 🐛 empty date should return passport not in validity
- ✨ Add TIFF and HEIC support
- ✨ Add real bounding boxes
- 📝 publish documentation to pages
- 🐛 make sure the 'Token' keyword is sent in the auth headers
- ✨ now possible to read file contents at any time
- 🔨 run mypy in pre-commit
- ⬆️ upgrade pikepdf
- ♻️ minor cleanup / refactor
- 📝 Add basic automated class documentation
- 🐛 💥 fix for customer_company_registration being a list of values
- ✨ allow specifying file closing behavior
- ⬆️ loosen setup.py requirements; update pinned dependencies
- ✅ better testing of PDF pages
- 🐛 fix for locale constructor
- 🐛 fix custom document in CLI
- 🏷️ declare type info to mypy
- 🐛 fix for dumping JSON in CLI
- 🏷️ set stricter pylint and mypy settings
- 🧑💻 add pre-commit to ensure proper code formatting
- 🎨 fixes to import order (isort) and documentation (pydocstyle)
- ✨ update to Invoices API v3
- ♻️ refactor
Endpointclasses and document building - ⬆️ Update PikePDF to 5.0.1
- ✨ add a basic logger
- 🐛 fix sending financial document via the CLI
- ✨ allow getting OCR return in CLI
- ✨ Make sure all document information is printed
- 🐛 fix for invoice to string
- 🐛 fix for counting empty PDF pages
- ⬆️ Update PikePDF to 4.5.0
- ⬆️ Update Pillow to 9.0.1 (security fix)
- ✨ Allow using custom documents (API builder)
- ♻️
probabilityrenamed toconfidencein the return fields to match API return - ♻️
Clientinitialization reworked to be more extensible - ♻️ Document loading and parsing reworked to separate arguments
- ⬆️ Upgrade pikepdf to 4.4.1
- 📝 Documentation migrated mainly to https://developers.mindee.com/docs
- 🐛 fix probabilities not loaded from API #49
- ✨ add CLI tool for testing
- ✨ Add user-agent header with SDK and Python versions
- ✨ Use pikepdf to replace pyMuPDF
- 👷 Pass all code through Black
- 📌 use
pip-toolsandsetup.pyto pin all dependencies - 📄 change to MIT license
- ⬆️ upgrade pymupdf to 1.18.17
- 🐛 handle few PDF files considered blank
- ✅ re-initialize file cursor in test
- 🐛 Fixed #15
- ✨ Added pdf page number parameter for multi-pages pdfs
- ✨ Added a blank pages only PDF detection & error raising
- 🐛 Forward uploaded file name for file verification rules
- 🐛 Change token management for new header format
- ✨ Adapted SDK to the new Mindee API endpoint
- ⚡ Single page object reconstruction is now server-side
- ➖ Removed Numpy dependency
- ✅ Updated tests with new data
- ⚡ FinancialDoc attributes have same type from invoice or receipt
- 🐛 Fixed FinancialDoc invoice version and reconstruction
- Updated total tax reconstruction for invoice
- Updated invoice API endpoint to V2
- Fixed null tax rate issue
- 🎉 First release!