ocr-stream 2.3__tar.gz → 2.3.2__tar.gz

{ocr_stream-2.3 → ocr_stream-2.3.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ocr_stream
-Version: 2.3
+Version: 2.3.2
 Summary: Extrair textos de documentos digitalizados e imagens.
 Author: Bruno Chaves
 Requires-Python: >=3.11
@@ -12,7 +12,7 @@ Requires-Dist: pandas
 Requires-Dist: requests>=2.32.4
 Requires-Dist: opencv-python
 Requires-Dist: pytesseract
-Requires-Dist: convert_stream>=2.3
+Requires-Dist: convert_stream>=2.3.2
 Requires-Dist: soup_files>=1.4
 
 # ocrlib

ocr_stream-2.3.2/ocr_stream/__init__.py

@@ -0,0 +1,269 @@
+#!/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
+from soup_files import File
+from convert_stream import (
+    ImageObject, DocumentPdf, ConvertPdfToImages, ColumnsTable, CollectionImages, FindText
+)
+from pandas import DataFrame
+from soup_files import JsonConvert, ProgressBarAdapter, CreatePbar
+
+
+def read_images(
+            images: list[ImageObject], *,
+            pbar: ProgressBarAdapter = CreatePbar().get()
+        ) -> DocumentPdf:
+    pbar.start()
+    print()
+    max_num = len(images)
+    pages = []
+    rec = RecognizeImage()
+    for idx, image in enumerate(images):
+        pbar.update(
+            ((idx + 1) / max_num) * 100,
+            f'[OCR] {idx + 1} / {max_num}',
+        )
+        out = rec.image_recognize(image).to_page_pdf()
+        pages.append(out)
+    tmp_doc = DocumentPdf.create_from_pages(pages)
+    print()
+    pbar.stop()
+    return tmp_doc
+
+
+def read_document(
+            document: DocumentPdf, *,
+            dpi: int = 300,
+            pbar: ProgressBarAdapter = CreatePbar().get()
+        ) -> DocumentPdf:
+    pbar.start()
+    print()
+    convert = ConvertPdfToImages.create(document)
+    convert.set_pbar(pbar)
+    images = convert.to_images(dpi=dpi)
+    print()
+    return read_images(images, pbar=pbar)
+
+
+class SearchableTextImage(object):
+    """
+        Esta classe serve para armazenar textos filtrados em documentos PDF
+    guardando o texto buscado e outros elementos como a página em que o texto foi
+    encontrado, linha e o nome do respetivo arquivo.
+    """
+    def __init__(self, silent: bool = False):
+        self.silent = silent
+        self.elements: dict[str, list[str]] = {
+            ColumnsTable.NUM_LINE.value: [],
+            ColumnsTable.TEXT.value: [],
+            ColumnsTable.FILE_PATH.value: [],
+        }
+
+    def __repr__(self):
+        return f'{__class__.__name__}: {self.elements}'
+
+    def is_empty(self) -> bool:
+        return len(self.elements[ColumnsTable.TEXT.value]) == 0
+
+    @property
+    def first(self) -> dict[str, str]:
+        if self.is_empty():
+            return {}
+        return {
+            ColumnsTable.NUM_LINE.value: self.elements[ColumnsTable.NUM_LINE.value][0],
+            ColumnsTable.TEXT.value: self.elements[ColumnsTable.TEXT.value][0],
+            ColumnsTable.FILE_PATH.value: self.elements[ColumnsTable.FILE_PATH.value][0],
+        }
+
+    @property
+    def last(self) -> dict[str, str]:
+        if self.is_empty():
+            return {}
+        return {
+            ColumnsTable.NUM_LINE.value: self.elements[ColumnsTable.NUM_LINE.value][-1],
+            ColumnsTable.TEXT.value: self.elements[ColumnsTable.TEXT.value][-1],
+            ColumnsTable.FILE_PATH.value: self.elements[ColumnsTable.FILE_PATH.value][-1],
+        }
+
+    @property
+    def length(self) -> int:
+        return len(self.elements[ColumnsTable.TEXT.value])
+
+    @property
+    def files(self) -> list[str]:
+        if self.is_empty():
+            return []
+        return self.elements[ColumnsTable.FILE_PATH.value]
+
+    def get_item(self, idx: int) -> dict[str, str]:
+        try:
+            return {
+                ColumnsTable.NUM_LINE.value: self.elements[ColumnsTable.NUM_LINE.value][idx],
+                ColumnsTable.TEXT.value: self.elements[ColumnsTable.TEXT.value][idx],
+                ColumnsTable.FILE_PATH.value: self.elements[ColumnsTable.FILE_PATH.value][idx],
+            }
+        except Exception as err:
+            if not self.silent:
+                print(err)
+            return {}
+
+    def add_line(self, text, *, num_line: str, file: str) -> None:
+        if not self.silent:
+            print('----------------------')
+            print(f'{file} => {text}')
+        self.elements[ColumnsTable.NUM_LINE.value].append(num_line)
+        self.elements[ColumnsTable.TEXT.value].append(text)
+        self.elements[ColumnsTable.FILE_PATH.value].append(file)
+
+    def clear(self) -> None:
+        for _k in self.elements.keys():
+            self.elements[_k].clear()
+
+    def to_string(self) -> str:
+        """
+            Retorna o texto da coluna TEXT em formato de string
+        ou 'nas' em caso de erro nas = Not a String
+        """
+        try:
+            return ' '.join(self.elements[ColumnsTable.TEXT.value])
+        except Exception as e:
+            print(e)
+            return 'nas'
+
+    def to_data_frame(self) -> DataFrame:
+        return DataFrame.from_dict(self.elements)
+
+    def to_file_json(self, file: File):
+        """Exporta os dados da busca para arquivo .JSON"""
+        dt = JsonConvert.from_dict(self.elements).to_json_data()
+        dt.to_file(file)
+
+    def to_file_excel(self, file: File):
+        """Exporta os dados da busca para um arquivo .XLSX"""
+        try:
+            self.to_data_frame().to_excel(file.absolute())
+        except Exception as e:
+            print(e)
+
+
+class FindTextImages(object):
+
+    def __init__(
+                self, collection_images: CollectionImages,
+                *,
+                rec_img: RecognizeImage = RecognizeImage(),
+                pbar: ProgressBarAdapter = CreatePbar().get(),
+            ):
+        self.collection_images: CollectionImages = collection_images
+        self._rec_img: RecognizeImage = rec_img
+        self.pbar: ProgressBarAdapter = pbar
+
+    def find(
+            self, text: str,
+            separator: str = '\n',
+            iqual: bool = False,
+            case: bool = False,
+            silent: bool = False,
+    ) -> SearchableTextImage:
+        """
+            Filtrar texto retornando a primeira ocorrência do Documento PDF.
+        """
+        print()
+        self.pbar.start()
+        _searchable = SearchableTextImage(silent)
+
+        max_num: int = self.collection_images.length
+        for idx, image in enumerate(self.collection_images):
+            self.pbar.update(
+                ((idx + 1) / max_num) * 100,
+                f'[Pesquisando]: {idx + 1} / {max_num} Aguarde...',
+            )
+            text_str_in_page: str = self._rec_img.image_to_string(image)
+
+            if (text_str_in_page == 'nas') or (text_str_in_page is None):
+                continue
+            try:
+                fd = FindText(text_str_in_page, separator=separator)
+                idx = fd.find_index(text, iqual=iqual, case=case)
+                if idx is None:
+                    continue
+                math_text = fd.get_index(idx)
+            except Exception as err:
+                print(f'{__class__.__name__} {err}')
+            else:
+                _searchable.add_line(
+                    math_text,
+                    num_line=str(idx + 1),
+                    file=image.name,
+                )
+                return _searchable
+        self.pbar.stop()
+        print()
+        return _searchable
+
+    def find_all(
+            self, text: str,
+            separator: str = '\n',
+            iqual: bool = False,
+            case: bool = False,
+            silent: bool = False,
+    ) -> SearchableTextImage:
+        """
+            Filtrar texto em documento PDF e retorna todas as ocorrências do texto
+        encontradas no documento, incluindo o número da linha, página e nome do arquivo
+        em cada ocorrência.
+        """
+        print()
+        self.pbar.start()
+        _searchable = SearchableTextImage(silent)
+        max_num: int = self.collection_images.length
+        for idx, image in enumerate(self.collection_images):
+            self.pbar.update(
+                ((idx + 1) / max_num) * 100,
+                f'[Pesquisando]: {idx + 1} / {max_num} Aguarde...',
+            )
+
+            text_str_in_page = self._rec_img.image_to_string(image)
+            if (text_str_in_page == 'nas') or (text_str_in_page is None):
+                continue
+            try:
+                _values = text_str_in_page.split(separator)
+                for num, item in enumerate(_values):
+                    if case:
+                        if iqual:
+                            if text == item:
+                                _searchable.add_line(
+                                    item,
+                                    num_line=str(num + 1),
+                                    file=image.name,
+                                )
+                        else:
+                            if text in item:
+                                _searchable.add_line(
+                                    item,
+                                    num_line=str(num + 1),
+                                    file=image.name,
+                                )
+                    else:
+                        if iqual:
+                            if text.lower() == item.lower():
+                                _searchable.add_line(
+                                    item,
+                                    num_line=str(num + 1),
+                                    file=image.name,
+                                )
+                        else:
+                            if text.lower() in item.lower():
+                                _searchable.add_line(
+                                    item,
+                                    num_line=str(num + 1),
+                                    file=image.name,
+                                )
+            except Exception as e:
+                print(f'{__class__.__name__} {e}')
+        self.pbar.stop()
+        print()
+        return _searchable

{ocr_stream-2.3 → ocr_stream-2.3.2}/ocr_stream/bin_tess/__init__.py

@@ -2,7 +2,7 @@
 from __future__ import annotations
 import os
 from shutil import which
-from soup_files import File, Directory, KERNEL_TYPE
+from soup_files import File, KERNEL_TYPE
 
 
 def __get_path_tesseract_unix() -> File | None:
@@ -93,5 +93,3 @@ class BinTesseract(object):
         if self.__path_tesseract is None:
             return False
         return self.__path_tesseract.exists()
-
-

{ocr_stream-2.3 → ocr_stream-2.3.2}/ocr_stream/modules/__init__.py

@@ -17,7 +17,8 @@ MOD_PYOCR: bool = False
 try:
     from pytesseract import pytesseract, Output
     MOD_PYTESSERACT = True
-except:
+except Exception as err:
+    print(err)
     pytesseract = None
     Output = None
 
@@ -25,7 +26,8 @@ try:
     import pyocr
     import pyocr.tesseract
     MOD_PYOCR = True
-except:
+except Exception as err:
+    print(err)
     pyocr = object
     #pyocr.tesseract = object
 

{ocr_stream-2.3 → ocr_stream-2.3.2}/ocr_stream/ocr/__init__.py

@@ -6,20 +6,21 @@ from abc import abstractmethod, ABC
 from typing import Union
 from io import BytesIO
 from pandas import DataFrame
-from soup_files import File, Directory
+from soup_files import File
 from convert_stream import (
-    ImageObject, DocumentPdf, PageDocumentPdf, MapText, ColumnsTable,
-    get_hash_from_bytes, DEFAULT_LIB_IMAGE, LibImage
+    ImageObject, DocumentPdf, PageDocumentPdf, DEFAULT_LIB_IMAGE, LibImage
 )
+from convert_stream.mod_types import get_void_metadata
+from convert_stream.mod_types.table_types import ColumnBody
 from ocr_stream.modules import (
-    LibOcr, ModuleOcr, DEFAULT_LIB_OCR, MOD_PYOCR, MOD_PYTESSERACT,
-    pytesseract, Output, pyocr,
+    LibOcr, DEFAULT_LIB_OCR, pytesseract, Output, pyocr,
 )
 from ocr_stream.bin_tess import BinTesseract
 
 try:
     import pyocr.tesseract
-except:
+except Exception as err:
+    print(err)
     pyocr = None
 
 
@@ -30,6 +31,7 @@ class TextRecognized(object):
     """
 
     def __init__(self, bytes_recognized: bytes):
+        self.metadata: dict[str, str | None] = get_void_metadata()
         self.bytes_recognized: bytes = bytes_recognized
         self.list_bad_char: list[str] = [
             ':', ',', ';', '$', '=',
@@ -51,21 +53,11 @@ class TextRecognized(object):
     def to_document(self) -> DocumentPdf:
         return DocumentPdf(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_dict(self, separator: str = '\n') -> dict[str, ColumnBody]:
+        return self.to_page_pdf().to_dict(separator=separator)
 
     def to_dataframe(self, separator='\n') -> DataFrame:
-        return self.to_map(separator).to_data()
+        return DataFrame.from_dict(self.to_dict(separator=separator))
 
 
 class ABCOcrTesseract(ABC):
@@ -166,8 +158,6 @@ class IPyOcr(ABCOcrTesseract):
             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:

{ocr_stream-2.3 → ocr_stream-2.3.2}/ocr_stream/recognize.py

@@ -1,8 +1,8 @@
 #!/usr/bin/env python3
 from __future__ import annotations
-from convert_stream import ImageObject, LibImage, DocumentPdf, PageDocumentPdf, ConvertPdfToImages, PdfStream
+from convert_stream import ImageObject, DocumentPdf, PageDocumentPdf, ConvertPdfToImages
 from ocr_stream.bin_tess import BinTesseract
-from ocr_stream.modules import LibOcr, DEFAULT_LIB_OCR, ModuleOcr
+from ocr_stream.modules import LibOcr, DEFAULT_LIB_OCR
 from ocr_stream.ocr import TesseractOcr, TextRecognized
 from pandas import DataFrame
 from soup_files import File, ProgressBarAdapter, CreatePbar

{ocr_stream-2.3 → ocr_stream-2.3.2}/ocr_stream.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ocr_stream
-Version: 2.3
+Version: 2.3.2
 Summary: Extrair textos de documentos digitalizados e imagens.
 Author: Bruno Chaves
 Requires-Python: >=3.11
@@ -12,7 +12,7 @@ Requires-Dist: pandas
 Requires-Dist: requests>=2.32.4
 Requires-Dist: opencv-python
 Requires-Dist: pytesseract
-Requires-Dist: convert_stream>=2.3
+Requires-Dist: convert_stream>=2.3.2
 Requires-Dist: soup_files>=1.4
 
 # ocrlib

{ocr_stream-2.3 → ocr_stream-2.3.2}/ocr_stream.egg-info/requires.txt

@@ -5,5 +5,5 @@ pandas
 requests>=2.32.4
 opencv-python
 pytesseract
-convert_stream>=2.3
+convert_stream>=2.3.2
 soup_files>=1.4

{ocr_stream-2.3 → ocr_stream-2.3.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "ocr_stream"
-version = "2.3"
+version = "2.3.2"
 authors = [
   { name="Bruno Chaves"},
 ]
@@ -21,7 +21,7 @@ dependencies = [
     "requests>=2.32.4",
     "opencv-python",
     "pytesseract",
-    "convert_stream>=2.3",
+    "convert_stream>=2.3.2",
     "soup_files>=1.4",
 ]
 

ocr_stream-2.3/ocr_stream/__init__.py

@@ -1,51 +0,0 @@
-#!/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
-from soup_files import File, Directory, InputFiles, LibraryDocs, CreatePbar, ProgressBarAdapter
-from convert_stream import ImageObject, PdfStream, DocumentPdf, ConvertPdfToImages
-
-
-def read_images(
-            images: list[ImageObject], *,
-            pbar: ProgressBarAdapter = CreatePbar().get()
-        ) -> DocumentPdf:
-    pbar.start()
-    print()
-    max_num = len(images)
-    pages = []
-    rec = RecognizeImage()
-    for idx, image in enumerate(images):
-        pbar.update(
-            ((idx + 1) / max_num) * 100,
-            f'[OCR] {idx + 1} / {max_num}',
-        )
-        out = rec.image_recognize(image).to_page_pdf()
-        pages.append(out)
-    tmp_doc = DocumentPdf.create_from_pages(pages)
-    print()
-    pbar.stop()
-    return tmp_doc
-
-
-def read_document(
-            document: DocumentPdf, *,
-            dpi: int = 300,
-            pbar: ProgressBarAdapter = CreatePbar().get()
-        ) -> DocumentPdf:
-    pbar.start()
-    print()
-    convert = ConvertPdfToImages.create(document)
-    convert.set_pbar(pbar)
-    images = convert.to_images(dpi=dpi)
-    print()
-    return read_images(images, pbar=pbar)
-
-
-
-
-
-
-