ocr-stream 1.4__tar.gz → 2.2__tar.gz

{ocr_stream-1.4 → ocr_stream-2.2}/PKG-INFO

@@ -1,32 +1,18 @@
 Metadata-Version: 2.4
 Name: ocr_stream
-Version: 1.4
+Version: 2.2
 Summary: Extrair textos de documentos digitalizados e imagens.
 Author: Bruno Chaves
-Requires: PyPDF2
-Requires: PyMuPDF
-Requires: reportlab
-Requires: openpyxl
-Requires: openpyxl
-Requires: pandas
-Requires: Pillow
-Requires: opencv_python
-Requires: requests
-Requires: pytesseract
-Requires: soup_files
-Requires: convert_stream
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown
 Requires-Dist: PyMuPDF
 Requires-Dist: reportlab
 Requires-Dist: openpyxl
 Requires-Dist: pandas
-Requires-Dist: Pillow>=11.3.0
 Requires-Dist: requests>=2.32.4
 Requires-Dist: opencv-python
-Requires-Dist: convert_stream>=1.4
+Requires-Dist: convert_stream>=2.2.4
 Requires-Dist: soup_files
-Dynamic: requires
 
 # ocrlib
 

ocr_stream-2.2/ocr_stream/__init__.py

@@ -0,0 +1,5 @@
+#!/usr/bin/env python3
+from ocr_stream.modules import *
+from ocr_stream.bin_tess import BinTesseract, get_path_tesseract_sys
+from ocr_stream.ocr import TesseractOcr, TextRecognized, DEFAULT_LIB_OCR
+from ocr_stream.recognize import RecognizeImage, RecognizePdf

ocr_stream-2.2/ocr_stream/bin_tess/__init__.py

@@ -0,0 +1,95 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+import os
+from shutil import which
+from soup_files import File, Directory, KERNEL_TYPE
+
+
+def __get_path_tesseract_unix() -> File | None:
+    out = which('tesseract')
+    if out is None:
+        return None
+    return File(out)
+
+
+def __get_path_tesseract_windows() -> File | None:
+    out = which('tesseract.exe')
+    if out is None:
+        return None
+    return File(out)
+
+
+def get_path_tesseract_sys() -> File | None:
+    if KERNEL_TYPE == 'Windows':
+        return __get_path_tesseract_windows()
+    return __get_path_tesseract_unix()
+
+
+class BinTesseract(object):
+    """
+        Fornece o caminho absoluto do tesseract instalado no sistema, se
+    disponível. Você pode usar um binário alternativo, basta informar
+    o caminho do binário desejado no construtor.
+    """
+    _instance = None  # Atributo de classe para armazenar a instância singleton
+
+    def __new__(cls, *args, **kwargs):
+        if cls._instance is None:
+            cls._instance = super(BinTesseract, cls).__new__(cls)
+        return cls._instance
+
+    def __init__(self, path: File = None):
+        # Garante que __init__ não será executado mais de uma vez
+        if hasattr(self, '_initialized') and self._initialized:
+            return
+
+        self._initialized = True
+        self.__tess_data_dir = None
+        self.__lang = None
+        if path is not None:
+            self.__path_tesseract: File = path
+        else:
+            self.__path_tesseract = get_path_tesseract_sys()
+
+    def set_tesseract(self, path: File):
+        if not isinstance(path, File):
+            return
+        self.__path_tesseract = path
+
+    def get_tesseract(self) -> File | None:
+        return self.__path_tesseract
+
+    def set_lang(self, lang: str):
+        if not isinstance(lang, str):
+            return
+        self.__lang = lang
+
+    def get_lang(self):
+        return self.__lang
+
+    def set_tessdata_dir(self, tessdata_dir: str):
+        if not isinstance(tessdata_dir, str):
+            return
+        self.__tess_data_dir = tessdata_dir
+
+    def get_tessdata_dir(self) -> str | None:
+        if self.__tess_data_dir is not None:
+            return self.__tess_data_dir
+        try:
+            if KERNEL_TYPE == 'Windows':
+                if os.path.exists("C:\\Program Files (x86)\\Tesseract-OCR\\tessdata"):
+                    self.__tess_data_dir = "C:\\Program Files (x86)\\Tesseract-OCR\\tessdata"
+            elif KERNEL_TYPE == 'Linux':
+                if os.path.exists('/usr/share/tesseract-ocr/5/tessdata'):
+                    self.__tess_data_dir = '/usr/share/tesseract-ocr/5/tessdata'
+        except Exception as err:
+            print(err)
+        return self.__tess_data_dir
+
+    def exists(self) -> bool:
+        """Verifica se o binário tesseract existe"""
+        if self.__path_tesseract is None:
+            return False
+        return self.__path_tesseract.exists()
+
+

ocr_stream-2.2/ocr_stream/modules/__init__.py

@@ -0,0 +1,43 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+from enum import Enum
+from typing import Union, TypeAlias
+
+
+class LibOcr(Enum):
+
+    PYTESSERACT = "pytesseract"
+    PYOCR = "pyocr"
+    NOT_IMPLEMENTED = "not_implemented"
+
+
+MOD_PYTESSERACT: bool = False
+MOD_PYOCR: bool = False
+
+try:
+    from pytesseract import pytesseract, Output
+    MOD_PYTESSERACT = True
+except:
+    pytesseract = None
+    Output = None
+
+try:
+    import pyocr
+    import pyocr.tesseract
+    MOD_PYOCR = True
+except:
+    pyocr = None
+    pyocr.tesseract = None
+
+if MOD_PYTESSERACT and MOD_PYOCR:
+    ModuleOcr: TypeAlias = Union[pyocr, pytesseract]
+    DEFAULT_LIB_OCR = LibOcr.PYTESSERACT
+elif MOD_PYTESSERACT:
+    ModuleOcr: TypeAlias = Union[pytesseract]
+    DEFAULT_LIB_OCR = LibOcr.PYTESSERACT
+elif MOD_PYOCR:
+    ModuleOcr: TypeAlias = Union[pyocr]
+    DEFAULT_LIB_OCR = LibOcr.PYOCR
+else:
+    ModuleOcr: TypeAlias = None
+    DEFAULT_LIB_OCR = LibOcr.NOT_IMPLEMENTED

ocr_stream-2.2/ocr_stream/ocr/__init__.py

@@ -0,0 +1,218 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+
+import os.path
+from abc import abstractmethod, ABC
+from typing import Union
+from io import BytesIO
+from pandas import DataFrame
+from soup_files import File, Directory
+from convert_stream import (
+    ImageObject, DocumentPdf, PageDocumentPdf, MapText, ColumnsTable,
+    get_hash_from_bytes, DEFAULT_LIB_IMAGE, LibImage
+)
+from ocr_stream.modules import (
+    LibOcr, ModuleOcr, DEFAULT_LIB_OCR, MOD_PYOCR, MOD_PYTESSERACT,
+    pytesseract, Output, pyocr,
+)
+from ocr_stream.bin_tess import BinTesseract
+
+try:
+    import pyocr.tesseract
+except:
+    pyocr.tesseract = None
+
+
+class TextRecognized(object):
+    """
+        Recebe os bytes de uma página PDF reconhecida de imagem
+    e exporta para vários tipos de dados.
+    """
+
+    def __init__(self, bytes_recognized: bytes):
+        self.bytes_recognized: bytes = bytes_recognized
+        self.list_bad_char: list[str] = [
+            ':', ',', ';', '$', '=',
+            '!', '}', '{', '(', ')',
+            '|', '\\', '‘', '*'
+            '¢', '“', '\'', '¢', '"',
+            '#', '.', '<', '?', '>',
+            '»', '@', '+', '[', ']',
+            '%', '~', '¥', '♀',
+        ]
+
+    def to_string(self) -> str | None:
+        return self.to_page_pdf().to_string()
+
+    def to_page_pdf(self) -> PageDocumentPdf:
+        doc = self.to_document()
+        return doc.to_pages()[0]
+
+    def to_document(self) -> DocumentPdf:
+        return DocumentPdf.create_from_bytes(BytesIO(self.bytes_recognized))
+
+    def to_map(self, separator: str = '\n') -> MapText:
+        text = self.to_string()
+        if (text is None) or (text == '') or (text == 'nas'):
+            return MapText([])
+        try:
+            return MapText(text.split(separator))
+        except Exception as err:
+            print(err)
+            return MapText([])
+
+    def to_dict(self, separator: str = '\n') -> dict[str, list[str]]:
+        return self.to_map(separator).map_elements
+
+    def to_dataframe(self, separator='\n') -> DataFrame:
+        return self.to_map(separator).to_data()
+
+
+class ABCOcrTesseract(ABC):
+
+    def __init__(self, bin_tess: BinTesseract, *, lang: str = None, tess_data_dir: str = None):
+        if not bin_tess.exists():
+            raise FileNotFoundError('tesseract binary not found')
+        self.bin_tess: BinTesseract = bin_tess
+        self.lang: str = lang
+        self.tess_data_dir: str = tess_data_dir
+        self.current_library: LibOcr = DEFAULT_LIB_OCR
+
+    @abstractmethod
+    def to_string(self, img: Union[File, ImageObject]) -> str:
+        pass
+
+    @abstractmethod
+    def to_reconize(self, img: Union[File, ImageObject]) -> TextRecognized:
+        pass
+
+
+class IPytesseract(ABCOcrTesseract):
+
+    def __init__(self, bin_tess: BinTesseract, *, lang: str = None, tess_data_dir: str = None):
+        super().__init__(bin_tess, lang=lang, tess_data_dir=tess_data_dir)
+        self.current_library: LibOcr = LibOcr.PYTESSERACT
+        self._MOD_TESS: pytesseract = pytesseract
+        self._MOD_TESS.tesseract_cmd = self.bin_tess.get_tesseract().absolute()
+
+    def __get_tess_dir_config(self) -> str | None:
+        """
+        https://github.com/h/pytesseract
+
+        Example config: r'--tessdata-dir "C:\Program Files (x86)\Tesseract-OCR\tessdata"'
+        tessdata_dir_config = r'--tessdata-dir <replace_with_your_tessdata_dir_path>'
+        It's important to add double quotes around the dir path.
+        """
+        # Caminho para os dados de idioma, por, eng etc...
+        # os.environ["TESSDATA_PREFIX"] = self.tess_data_dir.absolute()
+        if self.tess_data_dir is None:
+            return ''
+        if not os.path.exists(self.tess_data_dir):
+            return ''
+        return r'--tessdata-dir "{}"'.format(self.tess_data_dir)
+
+    def to_string(self, img: Union[File, ImageObject]) -> str:
+        if isinstance(img, File):
+            img = ImageObject.create_from_file(img)
+
+        if DEFAULT_LIB_IMAGE == LibImage.OPENCV:
+            _im = img.to_opencv()
+        elif DEFAULT_LIB_IMAGE == LibImage.PIL:
+            _im = img.to_pil()
+        else:
+            raise NotImplementedError()
+
+        if self.lang is None:
+            return self._MOD_TESS.image_to_string(_im, config=self.__get_tess_dir_config())
+        else:
+            return self._MOD_TESS.image_to_string(_im, lang=self.lang, config=self.__get_tess_dir_config())
+
+    def to_reconize(self, img: Union[File, ImageObject]) -> TextRecognized:
+
+        if isinstance(img, File):
+            img = ImageObject.create_from_file(img)
+
+        if DEFAULT_LIB_IMAGE == LibImage.OPENCV:
+            _im = img.to_opencv()
+        elif DEFAULT_LIB_IMAGE == LibImage.PIL:
+            _im = img.to_pil()
+        else:
+            raise NotImplementedError()
+        bt = self._MOD_TESS.image_to_pdf_or_hocr(
+            _im,
+            lang=self.lang,
+            config=self.__get_tess_dir_config()
+        )
+        return TextRecognized(bt)
+
+
+# ======================================================================#
+# Modulo OCR pyocr
+# ======================================================================#
+
+
+class IPyOcr(ABCOcrTesseract):
+
+    def __init__(self, bin_tess: BinTesseract, *, lang: str = None, tess_data_dir: str = None):
+        super().__init__(bin_tess, lang=lang, tess_data_dir=tess_data_dir)
+
+        self.current_library: LibOcr = LibOcr.PYOCR
+        pyocr_modules: list = pyocr.get_available_tools()
+        if len(pyocr_modules) == 0:
+            raise ValueError(f"{__class__.__name__} No OCR tool found")
+        # The tools are returned in the recommended order of usage
+        self._pyOcr = pyocr_modules[0]
+        langs: list[str] = self._pyOcr.get_available_languages()
+        if lang in lang:
+            self.lang = lang
+        else:
+            self.lang = langs[0]
+
+        # Ex: Will use tool 'libtesseract'
+        print(f"Will use tool {self._pyOcr.get_name()}")
+
+    def to_string(self, img: Union[File, ImageObject]) -> str:
+        if isinstance(img, File):
+            img = ImageObject.create_from_file(img)
+
+        if DEFAULT_LIB_IMAGE == LibImage.OPENCV:
+            _im = img.to_opencv()
+        elif DEFAULT_LIB_IMAGE == LibImage.PIL:
+            _im = img.to_pil()
+
+        return self._pyOcr.to_string(
+            _im,
+            lang=self.lang,
+            builder=pyocr.builders.TextBuilder()
+        )
+
+    def to_reconize(self, img: Union[File, ImageObject]) -> TextRecognized:
+        raise NotImplementedError(f'{__class__.__name__} método não implementado')
+
+
+class TesseractOcr(object):
+
+    def __init__(self, mod_ocr: ABCOcrTesseract):
+        self.mod_ocr: ABCOcrTesseract = mod_ocr
+
+    def to_string(self, img: Union[File, ImageObject]) -> str:
+        return self.mod_ocr.to_string(img)
+
+    def to_reconize(self, img: Union[File, ImageObject]) -> TextRecognized:
+        return self.mod_ocr.to_reconize(img)
+
+    @classmethod
+    def create(
+                cls,
+                bin_tess: BinTesseract, *,
+                lang: str = None,
+                tess_data_dir: str = None,
+                lib_ocr: LibOcr = DEFAULT_LIB_OCR,
+            ) -> TesseractOcr:
+        if lib_ocr == LibOcr.PYTESSERACT:
+            md = IPytesseract(bin_tess, lang=lang, tess_data_dir=tess_data_dir)
+        elif lib_ocr == LibOcr.PYOCR:
+            md = IPyOcr(bin_tess, lang=lang, tess_data_dir=tess_data_dir)
+        else:
+            raise ValueError(f'{lib_ocr} is not supported')
+        return cls(md)

ocr_stream-2.2/ocr_stream/recognize.py

@@ -0,0 +1,105 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+from convert_stream import ImageObject, LibImage, DocumentPdf, PageDocumentPdf, ConvertPdfToImages, PdfStream
+from ocr_stream.bin_tess import BinTesseract
+from ocr_stream.modules import LibOcr, DEFAULT_LIB_OCR, ModuleOcr
+from ocr_stream.ocr import TesseractOcr, TextRecognized
+from pandas import DataFrame
+from soup_files import File, ProgressBarAdapter, CreatePbar
+
+
+class RecognizeImage(object):
+    """
+        Reconhecer textos em imagens com pytesseract|pyocr usando o tesseract
+    """
+
+    def __init__(self, bin_tess: BinTesseract, *, lib_ocr: LibOcr = DEFAULT_LIB_OCR):
+        super().__init__()
+        self.module_ocr: TesseractOcr = TesseractOcr.create(
+            bin_tess,
+            lib_ocr=lib_ocr,
+            lang=bin_tess.get_lang(),
+            tess_data_dir=bin_tess.get_tessdata_dir(),
+        )
+
+    @property
+    def bin_tesseract(self) -> BinTesseract:
+        return self.module_ocr.mod_ocr.bin_tess
+
+    def image_content_data(self, img: ImageObject | File) -> DataFrame:
+        """
+        Retorna uma tabela estruturada com os dados reconhecidos da imagem.
+
+        Este método extrai as informações contidas em uma imagem e as organiza em
+        um DataFrame com colunas como o texto detectado, coordenadas da caixa delimitadora,
+        nível do OCR e confiabilidade, entre outras.
+
+        Parâmetros:
+        img (ImageObject | File): A imagem a ser processada, podendo ser um objeto
+        `ImageObject` ou um arquivo `File`.
+
+        Retorna:
+        DataFrame: Uma tabela contendo os dados reconhecidos na imagem.
+
+        Lança:
+        ValueError: Se o tipo do parâmetro `img` não for `ImageObject` ou `File`.
+
+        @param img: Imagem a ser processada, podendo ser uma instância de `ImageObject`
+            ou um arquivo `File`.
+        @type img: ImageObject | File
+
+        @rtype: pandas.DataFrame
+        @return: Um DataFrame com os dados reconhecidos na imagem.
+
+        @raise ValueError: Caso o tipo de `img` não seja `ImageObject` nem `File`.
+        """
+        return self.image_recognize(img).to_dataframe()
+
+    def image_recognize(self, img: ImageObject | File) -> TextRecognized:
+        rec: TextRecognized = self.module_ocr.to_reconize(img)
+        return rec
+
+    def image_to_string(self, img: ImageObject | File) -> str:
+        return self.module_ocr.to_string(img)
+
+
+class RecognizePdf(object):
+
+    def __init__(self, bin_tess: BinTesseract, *, lib_ocr: LibOcr = DEFAULT_LIB_OCR):
+        self.recognize_image: RecognizeImage = RecognizeImage(bin_tess, lib_ocr=lib_ocr)
+        self.pbar: ProgressBarAdapter = CreatePbar().get()
+
+    def set_pbar(self, p: ProgressBarAdapter):
+        self.pbar = p
+
+    def recognize_page_pdf(self, page: PageDocumentPdf) -> DocumentPdf:
+        """
+            Converte a página em Imagem, reconhece o texto com OCR e
+        retorna uma nova página com o texto reconhecido.
+        """
+        _doc = DocumentPdf.create_from_pages([page])
+        _conv = ConvertPdfToImages.create(_doc)
+        img: ImageObject = _conv.to_images()[0]
+        text_recognized: TextRecognized = self.recognize_image.image_recognize(img)
+        return text_recognized.to_document()
+
+    def recognize_document(self, doc: DocumentPdf) -> DocumentPdf:
+        _stream = PdfStream()
+        _conv = ConvertPdfToImages.create(doc)
+        images: list[ImageObject] = _conv.to_images()
+        max_num: int = len(images)
+        for num, img in enumerate(images):
+            self.pbar.update(
+                ((num + 1) / max_num) * 100,
+                f'Aplicando OCR [{num+1} de {max_num}]'
+            )
+            text_recognized = self.recognize_image.image_recognize(img)
+            _stream.add_page(text_recognized.to_page_pdf())
+        print()
+        self.pbar.stop()
+        return _stream.to_document()
+
+    def recognize_file_pdf(self, file: File) -> DocumentPdf:
+        _doc = DocumentPdf.create_from_file(file)
+        return self.recognize_document(_doc)
+

{ocr_stream-1.4 → ocr_stream-2.2}/ocr_stream.egg-info/PKG-INFO

@@ -1,32 +1,18 @@
 Metadata-Version: 2.4
 Name: ocr_stream
-Version: 1.4
+Version: 2.2
 Summary: Extrair textos de documentos digitalizados e imagens.
 Author: Bruno Chaves
-Requires: PyPDF2
-Requires: PyMuPDF
-Requires: reportlab
-Requires: openpyxl
-Requires: openpyxl
-Requires: pandas
-Requires: Pillow
-Requires: opencv_python
-Requires: requests
-Requires: pytesseract
-Requires: soup_files
-Requires: convert_stream
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown
 Requires-Dist: PyMuPDF
 Requires-Dist: reportlab
 Requires-Dist: openpyxl
 Requires-Dist: pandas
-Requires-Dist: Pillow>=11.3.0
 Requires-Dist: requests>=2.32.4
 Requires-Dist: opencv-python
-Requires-Dist: convert_stream>=1.4
+Requires-Dist: convert_stream>=2.2.4
 Requires-Dist: soup_files
-Dynamic: requires
 
 # ocrlib
 

ocr_stream-2.2/ocr_stream.egg-info/SOURCES.txt

@@ -0,0 +1,12 @@
+README.md
+pyproject.toml
+ocr_stream/__init__.py
+ocr_stream/recognize.py
+ocr_stream.egg-info/PKG-INFO
+ocr_stream.egg-info/SOURCES.txt
+ocr_stream.egg-info/dependency_links.txt
+ocr_stream.egg-info/requires.txt
+ocr_stream.egg-info/top_level.txt
+ocr_stream/bin_tess/__init__.py
+ocr_stream/modules/__init__.py
+ocr_stream/ocr/__init__.py

{ocr_stream-1.4 → ocr_stream-2.2}/ocr_stream.egg-info/requires.txt

@@ -2,8 +2,7 @@ PyMuPDF
 reportlab
 openpyxl
 pandas
-Pillow>=11.3.0
 requests>=2.32.4
 opencv-python
-convert_stream>=1.4
+convert_stream>=2.2.4
 soup_files

{ocr_stream-1.4 → ocr_stream-2.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "ocr_stream"
-version = "1.4"
+version = "2.2"
 authors = [
   { name="Bruno Chaves"},
 ]
@@ -18,12 +18,16 @@ dependencies = [
     "reportlab",
     "openpyxl",
     "pandas",
-    "Pillow>=11.3.0",
     "requests>=2.32.4",
     "opencv-python",
-    "convert_stream>=1.4",
+    "convert_stream>=2.2.4",
     "soup_files",
 ]
 
 [tool.setuptools]
-packages = ["ocr_stream", "ocr_stream.models", "ocr_stream.utils"]
+packages = [
+    "ocr_stream",
+    "ocr_stream.bin_tess",
+    "ocr_stream.modules",
+    "ocr_stream.ocr",
+]

ocr_stream-1.4/ocr_stream/__init__.py

@@ -1,8 +0,0 @@
-#!/usr/bin/env python3
-
-from .models.m_modules_ocr import LibraryOCR
-from .modules_ocr import TextRecognized, BinaryTesseract
-from .utils import File, Directory, InputFiles, UserFileSystem
-from .extractors import RecognizeImage, RecognizePdf
-from ._version import __version__, __modify_data__
-

ocr_stream-1.4/ocr_stream/__main__.py

@@ -1,46 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-"""
-------------------------------------------------------------
-    Configuração instalação:
-------------------------------------------------------------
- $ pip3 install -r requirements.txt
- 
-------------------------------------------------------------
-    Referências:
-------------------------------------------------------------
-
-https://gist.github.com/endolith/334196bac1cac45a4893
-
-Tesseract OCR
-https://github.com/tesseract-ocr/tesseract
-
-pytesseract
-https://github.com/h/pytesseract
-
-pyocr
-https://gitlab.gnome.org/World/OpenPaperwork/pyocr
-"""
-
-import sys, os
-this_file = os.path.abspath(os.path.relpath(__file__))
-dir_of_project = os.path.dirname(this_file)
-sys.path.insert(0, dir_of_project)
-
-from convert_stream import print_title
-from ocr_stream import __version__, __modify_data__
-
-
-def test():
-    # /usr/bin/tesseract
-    # /usr/share/tesseract-ocr/5/tessdata
-    pass
-    
-    
-def main():
-    print_title(f' ocrlib Versão: {__version__} | Atualização: {__modify_data__} ')
-    test()
-
-
-if __name__ == '__main__':
-    main()

ocr_stream-1.4/ocr_stream/_version.py

@@ -1,5 +0,0 @@
-#!/usr/bin/env python3
-#
-
-__version__ = '1.4'
-__modify_data__ = '2025-07-11'