File: //usr/lib/python3/dist-packages/ocrmypdf/__pycache__/_validation.cpython-38.pyc
U
��Z^p> � @ sZ d dl Z d dlZd dlZd dlZd dlmZ d dlmZ d dlZddl m
Z
ddl m Z m
Z
mZmZ ddlmZmZmZmZmZmZmZ ddlmZmZmZm Z e d d
d d d
g�Z d Z e� e!�Z"e
� dd� Z#dd� Z$dd� Z%dd� Z&dd� Z'dd� Z(dd� Z)d d � Z*d d � Z+d d!� Z,d"d#� Z-d$d%� Z.d&d'� Z/d(d)� Z0d*d+� Z1d,d-� Z2d.d/� Z3d0d1� Z4dS )2� N)�Path)� copyfileobj� )�verify_python3_env)� BadArgsError�InputFileError�MissingDependencyError�OutputFileAccessError)�check_external_program� ghostscript�jbig2enc�pngquant�qpdf� tesseract�unpaper)�is_file_writable�is_iterable_notstr� monotonic� safe_symlinkZengZdeuZspaZitaZporc C s" t jdkr tjdkr t�d� d S )N�ntl zmYou are running OCRmyPDF in a 32-bit (x86) Python interpreter.Please use a 64-bit (x86-64) version of Python.)�os�name�sys�maxsize�log�error� r r �6/usr/lib/python3/dist-packages/ocrmypdf/_validation.py�check_platform= s �r c C s� | j s4tg| _ t�� d }|r4|�d�s4t�dt� d| j d krT| j d �d�| _ t| j �}|� t
� � �s�d}|t
� � D ]}||d 7 }q|t |��d S )Nr Zenz-No language specified; assuming --language %s�+zgThe installed version of tesseract does not have language data for the following requested languages:
�
)
�language�DEFAULT_LANGUAGE�localeZ getlocale�
startswithr �debug�split�set�issubsetr � languagesr )�optionsZ system_langr) �msgZlangr r r �check_options_languagesF s
�r, c C s
t | j�}|�t�}| jdkr0|s0d}t�|� t�� dk rj| j dkrj|sjd}|dt�� � �7 }t�|� | jdkrzd| _| jdkr�t
� | j |�s�t
d ��| j d
kr�d | _ | j d kr�t�� d
k r�t
d��d}t| j| j| j| jf�s�d}|| _| j�s| j�rtd��d S )NZhocrz�The 'hocr' PDF renderer is known to cause problems with one or more of the languages in your document. Use --pdf-renderer auto (the default) to avoid this issue.z9.20Zpdfz�The installed version of Ghostscript does not work correctly with the OCR languages you specified. Use --output-type pdf or upgrade to Ghostscript 9.20 or later to avoid this issue.zFound Ghostscript �autoZsandwichz�You are using an alpha version of Tesseract 4.0 that does not support the textonly_pdf parameter. We don't support versions this old.�pdfazpdfa-2zpdfa-3z9.19z7--output-type pdfa-3 requires Ghostscript 9.19 or laterFTz\--redo-ocr is not currently compatible with --deskew, --clean-final, and --remove-background)r' r! r( �
HOCR_OK_LANGSZ pdf_rendererr �warningr �version� output_typer Zhas_textonly_pdf�
tesseract_envr �any�deskew� clean_final� force_ocr�remove_background�lossless_reconstruction�redo_ocrr )r* r) Zis_latinr+ r9 r r r �check_options_output\ sP
�
�
��
����r; c C s, | j dkr(| jdkr td��| jd | _ d S )N� �-z@--sidecar filename must be specified when output file is stdout.z.txt)�sidecar� output_filer �r* r r r �check_options_sidecar� s
�rA c
C s� | j r d| _| jr | js td��| jr�tddtjddgd� z| jrRt�| j�| _W n. tk
r� } ztt |���W 5 d }~X Y nX d S )NTz&--clean is required for --unpaper-argsr z6.1z--clean, --clean-final��program�package�version_checker� need_version� required_for)
r6 ZcleanZ unpaper_argsr r
r r1 Zvalidate_custom_args� Exception�str)r* �er r r �check_options_preprocessing� s& ��
rK c C s� t | �rt| �S g }| �dd��d�}|D ]�}|s4q*z|�d�\}}W n& tk
rl |�t|�d � Y q*X z |�tt|�d t|��� W q* tk
r� t d��Y q*X q*t
|�s�t � d� t
dd � |D ��r�t d
��t �d |� t|�S ) N� � �,r= r zinvalid page rangezQList of pages to process contains duplicate pages, or pages that are out of orderc s s | ]}|d k V qdS )r Nr )�.0�pager r r � <genexpr>� s z%_pages_from_ranges.<locals>.<genexpr>z)pages refers to a page number less than 1zOCRing only these pages: %s)r r' �replacer& �
ValueError�append�int�extend�ranger r r r0 r4 r% )Zranges�pagesZ page_groups�g�start�endr r r �_pages_from_ranges� s. � r\ c C sX t dd� | j| j| jfD ��}|dkr.td��| jrB| jrBtd��| jrTt| j�| _d S )Nc S s g | ]}|rd nd�qS )r r r )rO Zoptr r r �
<listcomp>� s �z.check_options_ocr_behavior.<locals>.<listcomp>� z8Choose only one of --force-ocr, --skip-text, --redo-ocr.z,--pages and --sidecar are mutually exclusive)�sumr7 Z skip_textr: r rX r> r\ )r* Zexclusive_optionsr r r �check_options_ocr_behavior� s �� r` c C st | j dkr tddtjddd� | j dkrHtddtjdd | js@d
nd d � | j d
krpt| j| j| jg�rpt �
d� d S )Nr^ r
z2.0.1z--optimize {2,3}rB �jbig2r z0.28z --optimize {2,3} | --jbig2-lossyTF)rC rD rE rF rG Z recommendedr zdThe arguments --jbig2-lossy, --png-quality, and --jpeg-quality will be ignored because --optimize=0.) �optimizer
r
r1 r Z jbig2_lossyr4 Z png_qualityZ jpeg_qualityr r0 r@ r r r �check_options_optimizing� s.
�
� ��rc c C sF | j dkr | j�d�r t�d� t�| j�sB| js8| j rBt�d� d S )Nr- r. zg--pdfa-image-compression argument has no effect when --output-type is not 'pdfa', 'pdfa-1', or 'pdfa-2'zZTesseract 4.0 ignores --user-words and --user-patterns, so these arguments have no effect.)
Zpdfa_image_compressionr2 r$ r r0 r Zhas_user_wordsr3 Z
user_wordsZ
user_patternsr@ r r r �check_options_advanced s �� ���rd c C s| dd l }| j| j| j| jg}dd� |D �D ]L}|D ]B}|�|�dksPt|�dkr2td�|t t|��dd � �
� ���q2q*d S )Nr c s s | ]}|r|V qd S �Nr )rO �mr r r rQ s z)check_options_metadata.<locals>.<genexpr>ZCoi zROne of the metadata strings contains an unsupported Unicode character: '{}' (U+{})r^ ) � unicodedata�titleZauthor�keywordsZsubject�category�ordrS �format�hex�upper)r* rg Zdocinfo�s�cr r r �check_options_metadata s ��rq c C s* t | jd �tj_tjjdkr&d tj_d S )Ni@B r )rU Zmax_image_mpixels�PILZImageZMAX_IMAGE_PIXELSr@ r r r �check_options_pillow) s rs c C sZ t � t| � t| � t| � t| � t| � t| � t| � t| � t | � t
| � d S re ) r r, rq r; rA rK r` rc rd rs �check_dependency_versionsr@ r r r �
check_options/ s ru c C s� t jdd� dkrdS t jdkr.ttjd�t _t jdkr^| jdkrPt� d� d S ttjd
�t _t j
dkr�| j dkr�t� d � d S ttjd�t _
dS ) a� Work around Python issue with multiprocessing forking on closed streams
https://bugs.python.org/issue28326
Attempting to a fork/exec a new Python process when any of std{in,out,err}
are closed or not flushable for some reason may raise an exception.
Fix this by opening devnull if the handle seems to be closed. Do this
globally to avoid tracking places all places that fork.
Seems to be specific to multiprocessing.Process not all Python process
forkers.
The error actually occurs when the stream object is not flushable,
but replacing an open stream object that is not flushable with
/dev/null is a bad idea since it will create a silent failure. Replacing
a closed handle with /dev/null seems safe.
r � )rv � � TN�wr= z0Trying to read from stdin but stdin seems closedF�rzyOutput was set to stdout '-' but the stream attached to stdout does not support the flush() system call. This will fail.) r � version_info�stderr�openr �devnull�stdin�
input_filer r �stdoutr? r@ r r r �check_closed_streams= s"
�r� c
C sn ddddd�}g }t | �D ]8\}}|jp*d}|dkr|�d�|d |�|d ��� q|rjt�d
d �|�� d S ) N�nrJ ro ry )r �Z � i r z{0}{1}r rM z Page orientations detected: %srL )� enumerateZrotationrT rl �getr �info�join)Zpdfinfo� directionZ orientationsr� rP Zangler r r �log_page_orientationsm s
r� c C s� | j dkrNt�d� tj�|d�}t|d��}ttj j
|� W 5 Q R X |dfS z,tj�|d�}t | j |� |t� | j �fW S t
k
r� td| j � ���Y nX d S )Nr= z reading file from standard inputr �wbz<stdin>�originzFile not found - )r� r r� r �pathr� r} r r r �bufferr �fspath�FileNotFoundErrorr )r* Z work_folder�targetZ
stream_bufferr r r �create_input_filex s
r� c C s> | j dkr tj�� r:td��n t| j �s:td| j � d���d S )Nr= ztOutput was set to stdout '-' but it looks like stdout is connected to a terminal. Please redirect stdout to a file.zOutput file location (z) is not a writable file.)r? r r� �isattyr r r r@ r r r �check_requested_output_file� s
�
�r� c
C s z t |��� j}t |��� j}W n tk
r6 Y d S X || }|dk sP|dk rTd S g }dddddh}|D ]*}t| |d�rj|�d |�d
d �� d �� qj| jd
kr�|�d� n:t� � t
� � d�} | � � D ] \}
} | s�|�d|
� d�� q�|�r dd� |� d } nd} t
�d|d�d| � �� d S )Ng�������?i�a r5 r6 r8 Z
oversampler7 FzThe argument --�_r= z! was issued, causing transcoding.r zOptimization was disabled.)ra r
zThe optional dependency 'zD' was not found, so some image optimizations could not be attempted.z#Possible reasons for this include:
r z@No reason for this increase is known. Please report this issue.zThe output file size is z.2fu × larger than the input file.
)r �stat�st_sizer� �getattrrT rR rb r � availabler
�itemsr� r r0 )
r* r� r? Z output_sizeZ
input_sizeZratioZreasonsZ
image_preproc�argZimage_optimizersr r� Z explanationr r r �report_output_file_size� sH � �
�
��r� c C sR t ddditjdd� t ddtjdd� t�� d kr<td
��t d d tjd d� d S )
Nr Zlinuxz
tesseract-ocrz4.0.0)rC rD rE rF Zgsr z9.15z9.24zGhostscript 9.24 contains serious regressions and is not supported. Please upgrade to Ghostscript 9.25 or use an older version.r z8.0.2)r
r r1 r r r r@ r r r rt � s, �� ��rt )5r# Zloggingr r �pathlibr Zshutilr rr Z _unicodefunr �
exceptionsr r r r �execr
r r r
r r r Zhelpersr r r r � frozensetr/ r" Z getLogger�__name__r r r, r; rA rK r\ r` rc rd rq rs ru r� r� r� r� r� rt r r r r �<module> s@ $
B 0 1