convert-stream 2.3.2__tar.gz → 2.3.5__tar.gz

{convert_stream-2.3.2 → convert_stream-2.3.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: convert_stream
-Version: 2.3.2
+Version: 2.3.5
 Summary: Manipulação de documentos PDF e Imagens. Converter, Ler, rotacionar exportar etc.
 Author: Bruno Chaves
 Requires-Python: >=3.11
@@ -21,8 +21,9 @@ pip3 install -r requirements.txt
 
 
 # Uso dividir páginas de um arquivo PDF:
-    >>> from file import File, Directory
+    >>> from soup_files import File, Directory
     >>> from pdflib import PdfStream
+
     # instâncie o caminho do arquivo PDF e o diretório para exportar os dados.
     file_pdf = File('path/to/file.pdf')
     output_dir = Directory('path/to/save')
@@ -32,8 +33,9 @@ pip3 install -r requirements.txt
     pdf_stream.to_files_pdf(output_dir) # isso irá converter cada página em um arquivo PDF
 
 # Uso para unir vários arquivos PDF.
-    >>> from file import File, Directory
+    >>> from soup_files import File, Directory
     >>> from pdflib import PdfStream
+
     # instâncie o caminho dos arquivos PDF que deseja unir.
     f = File('path/to/file.pdf')
     f2 = File('path/to/file2.pdf)

{convert_stream-2.3.2 → convert_stream-2.3.5}/README.md

@@ -3,8 +3,9 @@ pip3 install -r requirements.txt
 
 
 # Uso dividir páginas de um arquivo PDF:
-    >>> from file import File, Directory
+    >>> from soup_files import File, Directory
     >>> from pdflib import PdfStream
+
     # instâncie o caminho do arquivo PDF e o diretório para exportar os dados.
     file_pdf = File('path/to/file.pdf')
     output_dir = Directory('path/to/save')
@@ -14,8 +15,9 @@ pip3 install -r requirements.txt
     pdf_stream.to_files_pdf(output_dir) # isso irá converter cada página em um arquivo PDF
 
 # Uso para unir vários arquivos PDF.
-    >>> from file import File, Directory
+    >>> from soup_files import File, Directory
     >>> from pdflib import PdfStream
+
     # instâncie o caminho dos arquivos PDF que deseja unir.
     f = File('path/to/file.pdf')
     f2 = File('path/to/file2.pdf)

convert_stream-2.3.5/convert_stream/__init__.py

@@ -0,0 +1,84 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+from io import BytesIO
+from typing import Union
+from soup_files import File, Directory, LibraryDocs, InputFiles
+
+from ._version import (
+    __module_name__, __author__, __license__, __modify_date__, __version__
+)
+from .mod_types import (
+    MetaDataFile, MetaDataItem, create_metadata, ColumnsTable, ListString, ArrayString,
+    HeadCell, HeadValues, ColumnBody, TextTable, LibDate, LibPDF, LibImage, LibPdfToImage,
+    LibImageToPdf, create_dict_map_from_file, create_void_dict_map, RotationAngle
+)
+from .image import ImageObject, CollectionImages
+from .pdf import DocumentPdf, PageDocumentPdf, CollectionPagePdf
+from .pdf.pdf_document import SearchableTextPdf
+from .doc_stream import PdfStream, SplitPdf
+from .table_files import PdfFinder
+
+
+def __create_stream() -> PdfStream:
+    st = PdfStream()
+    st.clear()
+    return st
+
+
+def read_pdf(file: Union[str, File, bytes, BytesIO]) -> PdfStream:
+    _pdf_stream = __create_stream()
+
+    if isinstance(file, File):
+        _pdf_stream.add_file_pdf(file)
+    elif isinstance(file, str):
+        _pdf_stream.add_file_pdf(File(file))
+    elif isinstance(file, bytes):
+        _pdf_stream.add_document(DocumentPdf(BytesIO(file)))
+    elif isinstance(file, BytesIO):
+        _pdf_stream.add_document(DocumentPdf(file))
+    else:
+        raise ValueError()
+    return _pdf_stream
+
+
+def read_files_pdf(files: Union[list[str], list[File]]) -> PdfStream:
+    _pdf_stream = __create_stream()
+    if len(files) == 0:
+        return _pdf_stream
+
+    if isinstance(files[0], str):
+        for f in files:
+            _pdf_stream.add_file_pdf(File(f))
+    elif isinstance(files[0], File):
+        _pdf_stream.add_files_pdf(files)
+    else:
+        raise ValueError()
+    return _pdf_stream
+
+
+def read_image(img: Union[str, File, bytes, BytesIO]) -> PdfStream:
+    _pdf_stream = __create_stream()
+    if isinstance(img, str):
+        _pdf_stream.add_file_image(File(img))
+    elif isinstance(img, File):
+        _pdf_stream.add_image(ImageObject.create_from_file(img))
+    elif isinstance(img, bytes):
+        _pdf_stream.add_image(ImageObject.create_from_bytes(img))
+    elif isinstance(img, BytesIO):
+        img.seek(0)
+        _pdf_stream.add_image(ImageObject.create_from_bytes(img.getvalue()))
+    else:
+        raise ValueError()
+    return _pdf_stream
+
+
+__all__ = [
+    '__modify_date__', '__author__', '__license__', '__module_name__', '__version__',
+    'PdfStream', 'DocumentPdf', 'PageDocumentPdf', 'CollectionPagePdf', 'CollectionImages',
+    'ImageObject', 'File', 'Directory', 'LibraryDocs', 'InputFiles', 'SearchableTextPdf',
+    'PdfFinder', 'SplitPdf', 'read_pdf', 'read_files_pdf', 'read_image',
+    'MetaDataFile', 'MetaDataItem', 'create_metadata', 'ColumnsTable', 'ListString', 'ArrayString',
+    'HeadCell', 'HeadValues', 'ColumnBody', 'TextTable', 'LibDate', 'LibPDF', 'LibImage', 'LibPdfToImage',
+    'LibImageToPdf', 'create_dict_map_from_file', 'create_void_dict_map', 'RotationAngle'
+]
+

{convert_stream-2.3.2 → convert_stream-2.3.5}/convert_stream/doc_stream.py

@@ -183,7 +183,7 @@ class PdfStream(object):
 
 class SplitPdf(object):
 
-    def __init__(self, pages: list[PageDocumentPdf] = [], pbar: ProgressBarAdapter = ProgressBarAdapter()):
+    def __init__(self, pages: list[PageDocumentPdf], pbar: ProgressBarAdapter = ProgressBarAdapter()):
         self.pbar: ProgressBarAdapter = pbar
         self.__collection_pages: CollectionPagePdf = CollectionPagePdf(pages)
         self.__collection_pages.set_pbar(self.pbar)

convert_stream-2.3.5/convert_stream/image/__init__.py

@@ -0,0 +1,7 @@
+#!/usr/bin/env python3
+
+from .img_object import ImageObject, CollectionImages, ProcessImageScanner
+__all__ = ['ImageObject', 'CollectionImages', 'ProcessImageScanner']
+
+
+

{convert_stream-2.3.2 → convert_stream-2.3.5}/convert_stream/image/img_object.py

@@ -10,12 +10,12 @@ from soup_files import File, Directory, InputFiles, LibraryDocs, ProgressBarAdap
 from io import BytesIO
 import numpy as np
 from scipy.ndimage import uniform_filter
-from convert_stream.mod_types import get_void_metadata, get_hash_from_bytes
+from convert_stream.mod_types import get_hash_from_bytes, MetaDataFile, MetaDataItem, create_metadata
 from convert_stream.mod_types.enums import LibImage, RotationAngle
 from convert_stream.mod_types.table_types import ListItems
 from convert_stream.mod_types.modules import (
     MOD_IMG_PIL, ModuleImage, DEFAULT_LIB_IMAGE,
-    cv2, MatLike, Image, ImageFilter
+    cv2, MatLike, Image, ImageFilter, LibImage
 )
 
 
@@ -26,11 +26,11 @@ class ABCImageObject(ABC):
         self.max_size: tuple[int, int] = (1980, 720)
         self.module_image: ModuleImage = module_image
         self.lib_image: LibImage = LibImage.OPENCV
-        self.metadata: dict[str, str | None] = get_void_metadata()
+        self.metadata: MetaDataFile = MetaDataFile()
 
     @property
     def name(self) -> str | None:
-        return self.metadata['name']
+        return self.metadata.name if not self.metadata.name.is_empty else None
 
     @abstractmethod
     def get_dimensions(self) -> tuple[int, int]:
@@ -205,11 +205,7 @@ class ImplementOpenCv(ABCImageObject):
         if img is None:
             raise ValueError(f"Não foi possível abrir a imagem: {f.absolute()}")
         im_obj = cls(img)
-        im_obj.metadata['file_path'] = f.absolute()
-        im_obj.metadata['md5'] = f.md5()
-        im_obj.metadata['name'] = f.absolute()
-        im_obj.metadata['size'] = f.size()
-        im_obj.metadata['extension'] = f.extension()
+        im_obj.metadata = create_metadata(f)
         return im_obj
 
     @classmethod
@@ -221,10 +217,10 @@ class ImplementOpenCv(ABCImageObject):
         bytes_io = BytesIO(bt)
         name: str = get_hash_from_bytes(bytes_io)
         bytes_io.close()
-        _obj = cls(img)
-        _obj.metadata['md5'] = name
-        _obj.metadata['name'] = name
-        return _obj
+        im_obj = cls(img)
+        im_obj.metadata.name = MetaDataItem(name)
+        im_obj.metadata.md5 = MetaDataItem(name)
+        return im_obj
 
 
 class ImplementPIL(ABCImageObject):
@@ -352,13 +348,9 @@ class ImplementPIL(ABCImageObject):
 
     @classmethod
     def create_from_file(cls, f: File) -> ABCImageObject:
-        img = Image.open(f.path)
+        img = Image.open(f.absolute())
         im_obj = cls(img)
-        im_obj.metadata['file_path'] = f.absolute()
-        im_obj.metadata['md5'] = f.md5()
-        im_obj.metadata['name'] = f.absolute()
-        im_obj.metadata['size'] = f.size()
-        im_obj.metadata['extension'] = f.extension()
+        im_obj.metadata = create_metadata(f)
         return im_obj
 
     @classmethod
@@ -368,8 +360,8 @@ class ImplementPIL(ABCImageObject):
         name = get_hash_from_bytes(bytes_io)
         bytes_io.close()
         _obj = cls(img)
-        _obj.metadata['md5'] = name
-        _obj.metadata['name'] = name
+        _obj.metadata.name = MetaDataItem(name)
+        _obj.metadata.md5 = MetaDataItem(name)
         return _obj
 
 
@@ -419,9 +411,13 @@ class ImageObject(object):
         return self.img_adapter.name
 
     @property
-    def metadata(self) -> dict[str, str | None]:
+    def metadata(self) -> MetaDataFile:
         return self.img_adapter.metadata
 
+    @metadata.setter
+    def metadata(self, metadata: MetaDataFile) -> None:
+        self.img_adapter.metadata = metadata
+
     def to_pil(self) -> Image.Image:
         """
         Converte o ImageObject para um objeto PIL.Image.
@@ -600,3 +596,72 @@ class CollectionImages(ListItems):
     def set_pbar(self, p: ProgressBarAdapter) -> None:
         self.pbar = p
 
+
+class ProcessImageScanner(object):
+
+    def __init__(self, image: ImageObject):
+        self.image: ImageObject = image
+
+    def select_doc(self) -> ImageObject | None:
+        """
+        Detecta e recorta automaticamente a área do documento principal
+        (texto ou folha digitalizada) em uma imagem.
+        Retorna um novo ImageObject contendo apenas o recorte detectado.
+        """
+        # Converte para matriz OpenCV
+        img_cv: np.ndarray = cv2.imdecode(self.image.to_numpy(), cv2.IMREAD_COLOR)
+        if img_cv is None:
+            print(f'[NDARRAY] Falha na imagem: {self.image.metadata.name}')
+            return None
+
+        # Etapa 1: converter para tons de cinza
+        gray = cv2.cvtColor(img_cv, cv2.COLOR_BGR2GRAY)
+
+        # Etapa 2: aplicar blur para suavizar ruídos
+        gray_blur = cv2.GaussianBlur(gray, (5, 5), 0)
+
+        # Etapa 3: detecção de bordas
+        edges = cv2.Canny(gray_blur, 50, 150)
+
+        # Etapa 4: encontrar contornos
+        contours, _ = cv2.findContours(edges, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+        if not contours:
+            print(f'[CONTOURS] Falha na imagem: {self.image.metadata.name}')
+            return None
+
+        # Etapa 5: selecionar o maior contorno plausível (maior área retangular)
+        contours = sorted(contours, key=cv2.contourArea, reverse=True)
+        doc_contour = None
+
+        for c in contours:
+            perimeter = cv2.arcLength(c, True)
+            approx = cv2.approxPolyDP(c, 0.02 * perimeter, True)
+            if len(approx) == 4:
+                doc_contour = approx
+                break
+
+        if doc_contour is None:
+            # Caso não encontre um quadrilátero, usa o maior contorno
+            doc_contour = contours[0]
+
+        # Etapa 6: criar bounding box e recortar
+        x, y, w, h = cv2.boundingRect(doc_contour)
+        cropped = img_cv[y:y + h, x:x + w]
+
+        # Evitar recortes muito pequenos (ruído)
+        if w < 50 or h < 50:
+            return None
+
+        # Etapa 7: converter o recorte em ImageObject
+        success, buffer = cv2.imencode(".png", cropped)
+        if not success:
+            print(f'Falha na imagem: {self.image.metadata.name}')
+            return None
+
+        bytes_io = BytesIO(buffer.tobytes())
+        new_image = ImageObject(bytes_io, lib_image=LibImage.OPENCV)
+        # Herdar metadados originais
+        new_image.metadata = self.image.metadata
+        new_image.metadata.name = MetaDataItem(f"{self.image.metadata.name}_crop")
+        return new_image
+

convert_stream-2.3.5/convert_stream/mod_types/__init__.py

@@ -0,0 +1,100 @@
+#!/usr/bin/env python3
+#
+from __future__ import annotations
+from hashlib import md5 as md5_hash
+from io import BytesIO
+from soup_files import File
+from .enums import (
+    RotationAngle, ColumnsTable, LibImage, LibPdfToImage, LibPDF, LibImageToPdf, LibDate
+)
+
+from .table_types import (
+    TextTable, ListString, ArrayString, HeadCell, HeadValues,
+    ColumnBody, create_void_dict_map, create_dict_map_values, create_void_df_map,
+    create_dict_map_from_file, concat_maps, contains, get_text_from_file
+)
+
+
+def get_hash_from_bytes(bt: BytesIO | bytes) -> str:
+    if isinstance(bt, BytesIO):
+        return md5_hash(bt.getvalue()).hexdigest().upper()
+    elif isinstance(bt, bytes):
+        return md5_hash(bt).hexdigest().upper()
+    else:
+        raise TypeError(f"Unsupported type {type(bt)}")
+
+
+class MetaDataItem(str):
+
+    def __init__(self, text: str = 'nan'):
+        super().__init__()
+        self.text: str = text
+
+    @property
+    def is_empty(self) -> bool:
+        return self.text == 'nan'
+
+
+class MetaDataFile(object):
+
+    def __init__(
+                self, *,
+                file_path: MetaDataItem = MetaDataItem(),
+                dir_path: MetaDataItem = MetaDataItem(),
+                name: MetaDataItem = MetaDataItem(),
+                md5: MetaDataItem = MetaDataItem(),
+                size: MetaDataItem = MetaDataItem(),
+                extension: MetaDataItem = MetaDataItem(),
+                origin_src: MetaDataItem = MetaDataItem(),
+            ):
+        self.file_path: MetaDataItem = file_path
+        self.dir_path: MetaDataItem = dir_path
+        self.name: MetaDataItem = name
+        self.md5: MetaDataItem = md5
+        self.size: MetaDataItem = size
+        self.extension: MetaDataItem = extension
+        self.origin_src: MetaDataItem = origin_src
+
+    def to_dict(self) -> dict[str, str | None]:
+        return {
+            'file_path': self.file_path if not self.file_path.is_empty else None,
+            'dir_path': self.dir_path if not self.dir_path.is_empty else None,
+            'name': self.name if not self.name.is_empty else None,
+            'md5': self.md5 if not self.md5.is_empty else None,
+            'size': self.size if not self.size.is_empty else None,
+            'extension': self.extension if not self.extension.is_empty else None,
+            'origin_src': self.origin_src if not self.origin_src.is_empty else None,
+        }
+
+
+def get_void_metadata() -> dict[str, str | None]:
+    """
+        Salvar os metadados básico de documentos/imagens lidas em disco.
+    """
+    return MetaDataFile().to_dict()
+
+
+def create_metadata(file: str | File) -> MetaDataFile:
+    if isinstance(file, str):
+        file = File(file)
+    mt = MetaDataFile(
+        file_path=MetaDataItem(file.absolute()),
+        dir_path=MetaDataItem(file.dirname()),
+        extension=MetaDataItem(file.extension()),
+        origin_src=MetaDataItem('file'),
+        name=MetaDataItem(file.name()),
+    )
+    if file.exists():
+        mt.size = MetaDataItem(file.size())
+        mt.md5 = MetaDataItem(file.md5())
+    return mt
+
+
+__all__ = [
+    'get_hash_from_bytes', 'get_void_metadata', 'get_void_metadata', 'contains',
+    'concat_maps', 'create_dict_map_from_file', 'create_dict_map_values',
+    'create_void_df_map', 'create_void_dict_map', 'ColumnBody', 'ColumnsTable',
+    'ListString', 'HeadCell', 'HeadValues', 'TextTable', 'RotationAngle', 'LibPDF',
+    'LibImageToPdf', 'LibPdfToImage', 'LibImage', 'LibDate', 'ArrayString',
+    'MetaDataFile', 'MetaDataItem', 'create_metadata',
+]