''' libtesseract/ is a wrapper for google's Tesseract-OCR C API ( http://code.google.com/p/tesseract-ocr/ ). USAGE: > from PIL import Image > from pyocr.libtesseract import image_to_string > print(image_to_string(Image.open('test.png'))) > print(image_to_string(Image.open('test-european.jpg'), lang='fra')) COPYRIGHT: PyOCR is released under the GPL v3. Copyright (c) Jerome Flesch, 2011-2016 https://gitlab.gnome.org/World/OpenPaperwork/pyocr#readme ''' from os import devnull from .. import builders from . import tesseract_raw from ..error import TesseractError from ..util import digits_only import logging logger = logging.getLogger(__name__) __all__ = [ 'can_detect_orientation', 'detect_orientation', 'get_available_builders', 'get_available_languages', 'get_name', 'get_version', 'image_to_string', 'is_available', 'TesseractError', ] def can_detect_orientation(): langs = get_available_languages() return 'osd' in langs def detect_orientation(image, lang=None): # C-API with Tesseract 4 segfaults if running OSD_ONLY # psm mode with other than osd language # lang argument left purely for compatibility reasons # tested on 4.0.0-rc2 handle = tesseract_raw.init(lang='osd') try: tesseract_raw.set_page_seg_mode( handle, tesseract_raw.PageSegMode.OSD_ONLY ) tesseract_raw.set_image(handle, image) os = tesseract_raw.detect_os(handle) if os['confidence'] <= 0: raise TesseractError( "no script", "no script detected" ) orientation = { tesseract_raw.Orientation.PAGE_UP: 0, tesseract_raw.Orientation.PAGE_RIGHT: 90, tesseract_raw.Orientation.PAGE_DOWN: 180, tesseract_raw.Orientation.PAGE_LEFT: 270, }[os['orientation']] return { 'angle': orientation, 'confidence': os['confidence'] } finally: tesseract_raw.cleanup(handle) def get_name(): return "Tesseract (C-API)" def get_available_builders(): return [ builders.TextBuilder, builders.WordBoxBuilder, builders.DigitBuilder, builders.LineBoxBuilder, builders.DigitLineBoxBuilder, ] def _tess_box_to_pyocr_box(box): return ( (box[0], box[1]), (box[2], box[3]), ) def image_to_string(image, lang=None, builder=None): if builder is None: builder = builders.TextBuilder() handle = tesseract_raw.init(lang=lang) lvl_line = tesseract_raw.PageIteratorLevel.TEXTLINE lvl_word = tesseract_raw.PageIteratorLevel.WORD try: # XXX(Jflesch): Issue #51: # Tesseract TessBaseAPIRecognize() may segfault when the target # language is not available clang = lang if lang else "eng" for lang_item in clang.split("+"): if lang_item not in tesseract_raw.get_available_languages(handle): raise TesseractError( "no lang", "language {} is not available".format(lang_item) ) tesseract_raw.set_page_seg_mode( handle, builder.tesseract_layout ) tesseract_raw.set_debug_file(handle, devnull) tesseract_raw.set_image(handle, image) if "digits" in builder.tesseract_configs: tesseract_raw.set_is_numeric(handle, True) # XXX(JFlesch): PageIterator and ResultIterator are actually the # very same thing. If it changes, we are screwed. tesseract_raw.recognize(handle) res_iterator = tesseract_raw.get_iterator(handle) if res_iterator is None: raise TesseractError( "no script", "no script detected" ) page_iterator = tesseract_raw.result_iterator_get_page_iterator( res_iterator ) while True: if tesseract_raw.page_iterator_is_at_beginning_of( page_iterator, lvl_line): (r, box) = tesseract_raw.page_iterator_bounding_box( page_iterator, lvl_line ) assert r box = _tess_box_to_pyocr_box(box) builder.start_line(box) last_word_in_line = ( tesseract_raw.page_iterator_is_at_final_element( page_iterator, lvl_line, lvl_word ) ) word = tesseract_raw.result_iterator_get_utf8_text( res_iterator, lvl_word ) confidence = tesseract_raw.result_iterator_get_confidence( res_iterator, lvl_word ) if word is not None and confidence is not None and word != "": (r, box) = tesseract_raw.page_iterator_bounding_box( page_iterator, lvl_word ) assert r box = _tess_box_to_pyocr_box(box) builder.add_word(word, box, confidence) if last_word_in_line: builder.end_line() if not tesseract_raw.page_iterator_next(page_iterator, lvl_word): break finally: tesseract_raw.cleanup(handle) return builder.get_output() def image_to_pdf(image, output_file, lang=None, input_file="stdin", textonly=False): ''' Creates pdf file with embeded text based on OCR from an image Args: image: image to be converted output_file: path to the file that will be created, `.pdf` extension should not be specified lang: three letter language code. For available languages see https://github.com/tesseract-ocr/tesseract/blob/master/doc/tesseract.1.asc#languages. Defaults to None. input_file: path to the image file that should be beneath the text in output pdf. If not specified (stdin, incorrect file) output pdf is correct but tesseract writes some errors about not being able to open the file. Defaults to stdin. textonly: create pdf with only one invisible text layer. Defaults to False. ''' LibtesseractPdfBuilder()\ .set_lang(lang)\ .set_output_file(output_file)\ .set_text_only(textonly)\ .add_image(image)\ .build() class LibtesseractPdfBuilder(object): ''' Creates a pdf file with embeded text based on OCR from one or more images. ''' def __init__(self): self.images = [] self.output_file = None self.lang = None self.text_only = False def set_lang(self, lang): ''' Language to be used for ocr. :param lang: three letter language code. For available languages see https://github.com/tesseract-ocr/tesseract/blob/master/doc/tesseract.1.asc#languages. Defaults to None. ''' self.lang = lang return self def set_output_file(self, output_file): self.output_file = output_file return self def set_text_only(self, text_only): ''' :param text_only: create pdf with only one invisible text layer. Defaults to False. ''' self.text_only = text_only return self def add_image(self, img): ''' Add an image to be converted to a page in the pdf :param img: image to convert ''' self.images.append(img) # or something else return self def __validate(self): if len(self.images) < 1: raise ValueError( "At least one image is required to build the pdf!" ) if self.output_file is None: raise ValueError("An output-file is required to build the pdf!") def build(self): ''' Create and write PDF file. ''' self.__validate() handle = tesseract_raw.init(lang=self.lang) renderer = None try: tesseract_raw.set_page_seg_mode( handle, tesseract_raw.PageSegMode.AUTO_OSD ) renderer = tesseract_raw.init_pdf_renderer( handle, self.output_file, self.text_only ) assert renderer tesseract_raw.begin_document(renderer, "") for image in self.images: tesseract_raw.set_image(handle, image) # tesseract_raw.set_input_name(handle, input_file) tesseract_raw.recognize(handle) tesseract_raw.add_renderer_image(handle, renderer) tesseract_raw.end_document(renderer) finally: tesseract_raw.cleanup(handle) if renderer: tesseract_raw.cleanup(renderer) def is_available(): available = tesseract_raw.is_available() if not available: return False version = get_version() # C-API with Tesseract <= 3.02 segfaults sometimes # (seen with Debian stable + Paperwork) # not tested with 3.03 if (version[0] < 3 or (version[0] == 3 and version[1] < 4)): logger.warning("Unsupported version [%s]" % ".".join( [str(r) for r in version] )) return False return True def get_available_languages(): handle = tesseract_raw.init() try: return tesseract_raw.get_available_languages(handle) finally: tesseract_raw.cleanup(handle) def get_version(): version = tesseract_raw.get_version() version = version.split(" ", 1)[0] # cut off "dev" string if exists for proper int conversion index = version.find("dev") if index != -1: version = version[:index] version = version.split(".") major = digits_only(version[0]) minor = digits_only(version[1]) upd = 0 if len(version) >= 3: upd = digits_only(version[2]) return (major, minor, upd)