U Bb@sddlZddlZddlZddlmZddlmZmZmZm Z m Z m Z m Z m Z mZmZmZddlmZmZddlmZmZmZmZmZddlmZddlmZdd lm Z m!Z!dd l"m#Z#m$Z$m%Z%m&Z&m'Z'm(Z(m)Z)m*Z*m+Z+m,Z,m-Z-m.Z.ee/e e/e,d d d Z0ee/e e/e,d ddZ1ee/ee,e2fddddZ3ee/ee,e2fddddZ4ee/ddddZ5ee/ddddZ6e/e e/e7dddZ8e/e e/e7dddZ9Gdd d Z:Gd!d"d"e%Z;Gd#d$d$Zzz,_create_rectangle_accessor..cs t||Sr*)r/r&r6r'r2r3r?{r@cs t|Sr*)r:r=rBr2r3r?|r@)propertyr'r<r2r>r3_create_rectangle_accessorxs    rEcCstdt||S)NcreateRectangleAccessor)rrErDr2r2r3rFsrFc@seZdZdZdedddZeedddZe eed d d Z de e dd ddZ de e e e ddddZe ddddZedddZdS)Transformationu Specify a 2D transformation. The transformation between two coordinate systems is represented by a 3-by-3 transformation matrix written as follows:: a b 0 c d 0 e f 1 Because a transformation matrix has only six elements that can be changed, it is usually specified in PDF as the six-element array [ a b c d e f ]. Coordinate transformations are expressed as matrix multiplications:: a b 0 [ x′ y′ 1 ] = [ x y 1 ] × c d 0 e f 1 Usage ----- >>> from PyPDF2 import Transformation >>> op = Transformation().scale(sx=2, sy=3).translate(tx=10, ty=20) >>> page.add_transformation(op) rrrrrrctmcCs ||_dSr*rIr&rJr2r2r3__init__szTransformation.__init__r)cCs@|jd|jddf|jd|jddf|jd|jddffSNrrrIr=r2r2r3matrixszTransformation.matrix)rSr)cCs@|dd|dd|dd|dd|dd|ddfS)NrrrOr2)rSr2r2r3compresss      zTransformation.compressr)txtyr)cCs<|j}t|d|d|d|d|d||d|fdS)NrrrOrPrQrRrI)rJrG)r&rUrVmr2r2r3 translateszTransformation.translateNsxsyr)cCsx|dkr|dkrtd|dkr$|}|dkr0|}|dk sd?Z+dde e e e!dd@dAdBZ,dde e e!ddCdDdEZ-dde e e e!ddFdGdHZ.dde e e e e!ddIdJdKZ/de e'efe!ddLdMdNZ0eddOdPdQZ1e e ddRdSdTZ2e ddUdVdWZ3e ddUdXdYZ4e e ddZd[d\Z5e e ddZd]d^Z6ddd_d`Z7dddadbZ8deeedddedfZ9eddgdhZ:die;jde?e edpdqdrZ@deeedddsdtZAeeBeeBefddudvZCeDe;jEdwZFeGeHddxdyZIeIjJeHddzd{dyZIeDd|e;jEfZKeGeHdd}d~ZLeLjJeHddzdd~ZLeDdd|e;jEfZMeGeHdddZNeNjJeHddzddZNeDdd|e;jEfZOeGeHdddZPePjJeHddzddZPeDdd|e;jEfZQeGeHdddZReRjJeHddzddZReGeeSdddZTeTjJeeSddzddZTZUS) PageObjecta PageObject represents a single page within a PDF file. Typically this object will be created by accessing the :meth:`get_page()` method of the :class:`PdfReader` class, but it is also possible to create an empty page with the :meth:`create_blank_page()` static method. :param pdf: PDF file the page belongs to. :param indirect_ref: Stores the original indirect reference to this object in its source PDF N)r- indirect_refr)cCs&ddlm}t|||_||_dS)Nr) PdfReader)Z_readerrrrrLr-rq)r&r-rqrrr2r2r3rLs  zPageObject.__init__rMcst}|dt|7}|S)Ns%d)superhash_value_dataid)r&data __class__r2r3rts zPageObject.hash_value_data)r-widthheightr)cCst|}|ttjtd|ttjt|ttjt|dksV|dkr|dk rt |j dkr|j t |j d}|j j }|j j }nt|ttjtdd||f|S)a# Return a new blank page. If ``width`` or ``height`` is ``None``, try to get the page size from the last page of *pdf*. :param pdf: PDF file the page belongs to :param float width: The width of the new page expressed in default user space units. :param float height: The height of the new page expressed in default user space units. :return: the new blank page :raises PageSizeNotDefinedError: if ``pdf`` is ``None`` or contains no page z/PageNrr)rp __setitem__r PGZTYPEZPARENTr! RESOURCESrlenZpagesmediaboxryrzrMEDIABOXr#)r-ryrzpageZlastpager2r2r3create_blank_pages zPageObject.create_blank_pagecCstddt|||S)z\ .. deprecated:: 1.28.0 Use :meth:`create_blank_page` instead. createBlankPager)rrpr)r-ryrzr2r2r3r$s zPageObject.createBlankPage)angler)cCsR|ddkrtd|tjd}t|tr0|n|}t|||ttj<|S)z Rotate a page clockwise by increments of 90 degrees. :param int angle: Angle to rotate the page. Must be an increment of 90 deg. Zrz'Rotation angle must be a multiple of 90) r]r+r|ZROTATEr,intr.r"r )r&rZ rotate_objZ current_angler2r2r3rf2s zPageObject.rotatecCstdd||S)Nrotate_clockwiserfrrfr&rr2r2r3rBs zPageObject.rotate_clockwisecCstdd||S)z[ .. deprecated:: 1.28.0 Use :meth:`rotate_clockwise` instead. rotateClockwiserfrrr2r2r3rFs zPageObject.rotateClockwisecCstdd|| S)zt .. deprecated:: 1.28.0 Use :meth:`rotate_clockwise` with a negative argument instead. rotateCounterClockwiserfrrr2r2r3rOs z!PageObject.rotateCounterClockwise)res1res2resourcer)cCst}|||ttt||t}i}t|D]`}||kr||||krt|t t }|||<||||<qF||krF||||<qF||fSr*) rupdater+r.r listkeysZraw_getr rouuidZuuid4)rrrZnew_resZpage2resZ rename_reskeyZnewnamer2r2r3_merge_resourcesXszPageObject._merge_resources)streamrenamer-r)cCs|s|St||}|jD]\}}t|tr`tt|D]&}||}t|tr6|||||<q6qt|tr|D]&}||}t|trn|||||<qnqt dt |q|S)Nztype of operands is ) r operationsr,rranger~r r+dictKeyErrortype)rrr-operands _operatorir_r2r2r3_content_stream_renameks      z!PageObject._content_stream_rename)contentsr-r)cCs0t||}|jdgdf|jgdf|S)NrqQ)rrinsertappend)rr-rr2r2r3 _push_pop_gss zPageObject._push_pop_gs)rr-rJr)c CsT|\}}}}}}t||}|jdt|t|t|t|t|t|gdg|S)Nrz cm)rrrr) rr-rJabcr1efr2r2r3_add_transformation_matrixs z%PageObject._add_transformation_matrixcCs tj|kr|tjSdSdS)z Access the page contents. :return: the ``/Contents`` object, or ``None`` if it doesn't exist. ``/Contents`` is optional, as described in PDF Reference 7.7.3.3 N)r|CONTENTSr.r=r2r2r3 get_contentss zPageObject.get_contentscCstdd|S)zW .. deprecated:: 1.28.0 Use :meth:`get_contents` instead. getContentsr)rrr=r2r2r3rs zPageObject.getContentsF)page2expandr)cCs|j||ddS)a Merge the content streams of two pages into one. Resource references (i.e. fonts) are maintained from both pages. The mediabox/cropbox/etc of this page are not altered. The parameter page's content stream will be added to the end of this page's content stream, meaning that it will be drawn after, or "on top" of this page. :param PageObject page2: The page to be merged into this one. Should be an instance of :class:`PageObject`. :param bool expand: If true, the current page dimensions will be expanded to accommodate the dimensions of the page to be merged. )rN) _merge_page)r&rrr2r2r3 merge_pageszPageObject.merge_page)rr)cCstdd||S)zU .. deprecated:: 1.28.0 Use :meth:`merge_page` instead. mergePager)rr)r&rr2r2r3rs zPageObject.mergePage)rpage2transformationrJrr)c Cs(t}i}tt|tj}tt|tj}t} ||fD]6} tj| kr@| tj} t| tr@| D]} | | qfq@t j t j t j t j t jt jt jfD]0} t||| \}}|r||t| <||qtt|t jtt|t jt|tt j<t}|}|dk r6|t||j|}|dk rt||j}|j}|jdt t!|j"|j#|j$|j%gdf|jdgdf|jdgdf|dk r||}t&|||j}t||j}|||r|'||t||j|ttj(<||ttj<| |ttj<dS)NrrerWrOn))rr r|r}r.rZANNOTSr,rRESZ EXT_G_STATEZFONTZXOBJECTZ COLOR_SPACEZPATTERNZSHADINGZ PROPERTIESrprr r frozensetr+ZPROC_SETunionrrr-rtrimboxrrmaprleftbottomryrzr_expand_mediaboxr)r&rrrJrZ new_resourcesrZoriginal_resourcesZpage2resourcesZ new_annotsrZannotsrefresnewZ newrenameZnew_content_arrayZoriginal_contentZ page2contentZrectr2r2r3rs                zPageObject._merge_page)rrJr)c st|jj|jj|jj|jjf}|jj|jj|jj|jj|jj|jj|jj|jjfdk rtddDtfddtdddD}tfddtdddD}nddd}ddd}t|t|f}t |t |f}t|d|dt|d|df}t |d|dt |d |df}||j_ ||j_ dS) Ncss|]}t|VqdSr*rn.0xr2r2r3 Dsz.PageObject._expand_mediabox..c3s:|]2}d|d|ddVqdS)rrOrrQNr2rrZcorners2rJr2r3rEsrrOc3s:|]2}d|d|ddVqdS)rrPrRNr2rrr2r3rIsrrP) rr as_numericrrighttoptuplerminmax lower_left upper_right)r&rrJZcorners1new_xnew_y lowerleft upperrightr2rr3r0s@               $zPageObject._expand_mediabox)rrJrr)csBtddttrjtt|fdd|dS)ah mergeTransformedPage is similar to merge_page, but a transformation matrix is applied to the merged stream. :param PageObject page2: The page to be merged into this one. Should be an instance of :class:`PageObject`. :param tuple ctm: a 6-element tuple containing the operands of the transformation matrix :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. z%page.mergeTransformedPage(page2, ctm)z5page2.add_transformation(ctm); page.merge_page(page2)cst|jSr*)rprr-)Z page2ContentrJrr2r3r?ysz1PageObject.mergeTransformedPage..N)rr,rGrJr rr)r&rrJrr2rr3mergeTransformedPage[s   zPageObject.mergeTransformedPage)rr`rr)cCs*tddt||}||||dS)a2 mergeScaledPage is similar to merge_page, but the stream to be merged is scaled by applying a transformation matrix. :param PageObject page2: The page to be merged into this one. Should be an instance of :class:`PageObject`. :param float scale: The scaling factor :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. z*page.mergeScaledPage(page2, scale, expand)zWpage2.add_transformation(Transformation().scale(scale)); page.merge_page(page2, expand)N)rrGr`r)r&rr`rr_r2r2r3mergeScaledPages zPageObject.mergeScaledPage)rrarr)cCs(tddt|}||||dS)aJ mergeRotatedPage is similar to merge_page, but the stream to be merged is rotated by applying a transformation matrix. :param PageObject page2: the page to be merged into this one. Should be an instance of :class:`PageObject`. :param float rotation: The angle of the rotation, in degrees :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. z.page.mergeRotatedPage(page2, rotation, expand)z[page2.add_transformation(Transformation().rotate(rotation)); page.merge_page(page2, expand)N)rrGrfr)r&rrarr_r2r2r3mergeRotatedPages  zPageObject.mergeRotatedPage)rrUrVrr)cCs*tddt||}||||dS)aq mergeTranslatedPage is similar to merge_page, but the stream to be merged is translated by applying a transformation matrix. :param PageObject page2: the page to be merged into this one. Should be an instance of :class:`PageObject`. :param float tx: The translation on X axis :param float ty: The translation on Y axis :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. z/page.mergeTranslatedPage(page2, tx, ty, expand)z\page2.add_transformation(Transformation().translate(tx, ty)); page.merge_page(page2, expand)N)rrGrXr)r&rrUrVrr_r2r2r3mergeTranslatedPages zPageObject.mergeTranslatedPage)rrarUrVrr)cCs8tddt| | |||}||||S)a mergeRotatedTranslatedPage is similar to merge_page, but the stream to be merged is rotated and translated by applying a transformation matrix. :param PageObject page2: the page to be merged into this one. Should be an instance of :class:`PageObject`. :param float tx: The translation on X axis :param float ty: The translation on Y axis :param float rotation: The angle of the rotation, in degrees :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. z@page.mergeRotatedTranslatedPage(page2, rotation, tx, ty, expand)zmpage2.add_transformation(Transformation().rotate(rotation).translate(tx, ty)); page.merge_page(page2, expand))rrGrXrfr)r&rrarUrVrr_r2r2r3mergeRotatedTranslatedPages  z%PageObject.mergeRotatedTranslatedPage)rrar`rr)cCs0tddt|||}||||dS)a mergeRotatedScaledPage is similar to merge_page, but the stream to be merged is rotated and scaled by applying a transformation matrix. :param PageObject page2: the page to be merged into this one. Should be an instance of :class:`PageObject`. :param float rotation: The angle of the rotation, in degrees :param float scale: The scaling factor :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. z;page.mergeRotatedScaledPage(page2, rotation, scale, expand)hpage2.add_transformation(Transformation().rotate(rotation).scale(scale)); page.merge_page(page2, expand)N)rrGrfr`r)r&rrar`rr_r2r2r3mergeRotatedScaledPages z!PageObject.mergeRotatedScaledPage)rr`rUrVrr)cCs.tddt||||}||||S)a mergeScaledTranslatedPage is similar to merge_page, but the stream to be merged is translated and scaled by applying a transformation matrix. :param PageObject page2: the page to be merged into this one. Should be an instance of :class:`PageObject`. :param float scale: The scaling factor :param float tx: The translation on X axis :param float ty: The translation on Y axis :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. z`. :param float tx: The translation on X axis :param float ty: The translation on Y axis :param float rotation: The angle of the rotation, in degrees :param float scale: The scaling factor :param bool expand: Whether the page should be expanded to fit the dimensions of the page to be merged. .. deprecated:: 1.28.0 Use :meth:`add_transformation` and :meth:`merge_page` instead. zFpage.mergeRotatedScaledTranslatedPage(page2, rotation, tx, ty, expand)rN)rrGrfr`rXr)r&rrar`rUrVrr_r2r2r3 mergeRotatedScaledTranslatedPage s z+PageObject.mergeRotatedScaledTranslatedPage)rJrr)c slttrj|}|dk rLt||j}t||j}||tt j <|rh|j j |j j |j j |j j |j j |j j |j j |j j gtddDfddtdddD}fd dtdddD}t|t|f}t|t|f}td|dtd |d f}td|dtd |d f}||j _||j _dS) aQ Apply a transformation matrix to the page. :param tuple ctm: A 6-element tuple containing the operands of the transformation matrix. Alternatively, a :py:class:`Transformation` object can be passed. See :doc:`/user/cropping-and-transforming`. Ncss|]}t|VqdSr*rrr2r2r3resz0PageObject.add_transformation..cs8g|]0}d|d|ddqS)rrOrrQr2rZcornersrJr2r3 fsz1PageObject.add_transformation..rrrOcs8g|]0}d|d|ddqS)rrPrRr2rrr2r3rjsrrP)r,rGrJrrprr-rr r|rrrrrrrrrrrrr)r&rJrcontentrrrrr2rr3add_transformationBs@              $zPageObject.add_transformation)rJr)cCstdd||dS)z] .. deprecated:: 1.28.0 Use :meth:`add_transformation` instead. addTransformationrN)rrrKr2r2r3rzs zPageObject.addTransformationrYcCs"||dd|ddf|j|||_|j|||_|j|||_|j|||_|j|||_tj|kr|tj}t |t r|dd}n|d}t t |d|t |d|t |d|t |d|f}t |t r||t tjtdt d<n||t tjt d<dS)a` Scale a page by the given factors by applying a transformation matrix to its content and updating the page size. This updates the mediabox, the cropbox, and the contents of the page. :param float sx: The scaling factor on horizontal axis. :param float sy: The scaling factor on vertical axis. rz/BBoxrrOrPN)rrr`cropboxartboxbleedboxrr|ZVPr,rr#rnr r")r&rZr[ZviewportZbboxZ scaled_bboxr2r2r3r`s0     zPageObject.scale)factorr)cCs|||dS)z Scale a page by the given factor by applying a transformation matrix to its content and updating the page size. :param float factor: The scaling factor (for both X and Y axis). N)r`r&rr2r2r3scale_byszPageObject.scale_bycCstdd|||dS)zS .. deprecated:: 1.28.0 Use :meth:`scale_by` instead. scaleByrN)rr`rr2r2r3rs zPageObject.scaleBy)ryrzr)cCs0|t|jj}|t|jj}|||dS)z Scale a page to the specified dimensions by applying a transformation matrix to its content and updating the page size. :param float width: The new width. :param float height: The new height. N)rnrryrzr`)r&ryrzrZr[r2r2r3scale_toszPageObject.scale_tocCstdd|||dS)zS .. deprecated:: 1.28.0 Use :meth:`scale_to` instead. scaleTorN)rr)r&ryrzr2r2r3rs zPageObject.scaleTocCs<|}|dk r8t|ts&t||j}||ttj<dS)a Compress the size of this page by joining all content streams and applying a FlateDecode filter. However, it is possible that this function will perform no action if content stream compression becomes "automatic". N)rr,rr-Z flate_encoder r|r)r&rr2r2r3compress_content_streamss   z#PageObject.compress_content_streamscCstdd|dS)zc .. deprecated:: 1.28.0 Use :meth:`compress_content_streams` instead. compressContentStreamsrN)rrr=r2r2r3rs z!PageObject.compressContentStreams)Tj_sepTJ_sepr)c Csd}|tj}t|ts(t||j}d}|jD]\}}|dkrFq2|dkr`dt|d}q2|dkr|d}t|tr||7}||7}|d7}q2|dkr|d7}q2|d kr|d7}|d}t|tr||d7}q2|d kr|d }t|tr|d7}||7}q2|d kr2|dD]} t| tr8||7}|| 7}nrt| t t frt | | d krt |dksx|ddkr|d7}n(t |dkr|ddkr|dd}q|d7}q2|S)a Locate all text drawing commands, in the order they are provided in the content stream, and extract the text. This works well for some PDF files, but poorly for others, depending on the generator used. This will be refined in the future. Do not rely on the order of text coming out of this function, as it will change if this function is made more sophisticated. :return: a string object. r?)TfsTfsTcsThsTlsTmodeTwrTj T*'"rOTJ rN) r|rr.r,rr-rrnr$r"rrr~) r&rrtextr space_scaleroperatorZ_textrr2r2r3_extract_text_oldsR                zPageObject._extract_text_oldc Csd}t|d|jdjD]P\}}|dkr@dd|dD}ng}||dd d||d 7}q|d 7}z||d d D]j}||d 7}||d d |d 7}z(|d d |d}||d 7}Wqtk rYqXqWntk r|d7}YnX|S)Nrz /ContentsbytesrcSsg|]}t|tr|qSr2)r,rorr2r2r3r=s z1PageObject._debug_for_extract..rzutf-8rrz ============================= /Resources/Fontz /EncodingzNo Font ) rZ getObjectr-rdecodejoinrh Exceptionr)r&outZoper_sfoZenc_reprr2r2r3_debug_for_extract7s2 *  zPageObject._debug_for_extracti@)objr- space_width content_keyr)csfd d itt|d}d|krFtt|dD]}t||||<q0z4t|tr^||n|}t|tsxt||d}Wntk rYdSXddddddggddddddgddddddgdd dddttd d d }t tt tt td d d t tt ddd tdfdd t t dd fdd } |j D]\} } | dkr| dg| d| qh| dkr| d| dg| d| d g| dg| d| d!dqh| d"kr| d#| d  g| d$| qh| d%kr| dD]p} t| tt fr:| d| gt| t tt tfrtt| krt dkr d&d'kr| dd'gqn| d(krL 7 z d&d)kr d)7 Wntk rYnXzpz>|d*} | | dd+d,kr|| | d| 7 Wn,tk r>td-| dtYnXW5d Xn | | | qh 7 S).a  Locate all text drawing commands, in the order they are provided in the content stream, and extract the text. This works well for some PDF files, but poorly for others, depending on the generator used. This will be refined in the future. Do not rely on the order of text coming out of this function, as it will change if this function is made more sophisticated. :param float space_width: force default space width (if not extracted from font (default 200) :param Optional[str] content_key: indicate the default key where to extract data None = the object; this allow to reuse the function on XObject default = "/Content" :return: a string object. rrrrrg@@g(@)rr)cSs|dkr dSdS)Nrrrr2)rr2r2r3signsz&PageObject._extract_text..sign)rWrr)c Ss|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|d|dgSrNr2)rWrr2r2r3mults&&z&PageObject._extract_text..mult)rWr)cSs8|ddkrdS|ddkr dS|ddkr0dSdSdS) NrPgư>rgưrrr2)rWr2r2r3orients   z(PageObject._extract_text..orientrMcsdS)Ng@@r2r2) _space_widthr2r3current_spacewidthsz4PageObject._extract_text..current_spacewidthN)rrr)c s&d}|dkr,ddddddg 7 d dS|dkrD 7 d nf|dkrZnP|dkrz Wn$tk rddddddgYnXn|d kr 7 d t|d t|d t|d t|d t|dt|dgn|dkrt|d dn|dkr2dt|d  nx|dkrLt|d n^|dkr dkrh 7 d z:|d d |d d |d d |d fWn8tk rtd td td d|d fYnXzt|d Wntk rYnXn|dkrNd}dt|d 7<dt|d 7<n\|dkrd}t|d t|d t|d t|d t|dt|dgn |dkrd}d8<n|dkrd}t|d tr |d 7 nd}t|d tr t|d n|d }td trpz| d d}Wn4tk rl| d dkr`dndd}YnXnd fdd|D} d fd d|D7 ndS|r" } |}|dd}|dd}t t |d |d t |d |d } | } |z|d kr|d!| krT d"d#kr d#7 nDt || d$krt || d%kr d"d&kr d&7 nl|d'kr|d(| krЈ d"d#kr d#7 nDt || d$krt || d%kr d"d&kr d&7 n|d)kr|d(| krJ d"d#kr d#7 nDt || d$krt || d%kr d"d&kr d&7 nx|d*kr|d!| krĈ d"d#kr d#7 nDt || d$krt || d%kr d"d&kr d&7 Wntk r YnXdS)+NFsBTrrrsETqQscmrrrOrPrQrRsTzgY@rTLrz???TdTsTmrr surrogatepasscharmapz utf-16-becs2g|]*}|dkr d|n t|fqS)r)rrrcmapr2r3rszGPageObject._extract_text..process_operation..cs(g|] }|dkr d|n|qS)rr2rr!r2r3rsgrrg333333? rrg?rr)rpopr rnrrr,ror%rr rbsqrtabs) rrZcheck_crlf_spacetttrWoZdeltaXZdeltaYkrZTLrZ char_scaleZ cm_matrixZcm_stackr"ZcmapsrZ font_sizerroutputrrZ tm_matrixZtm_prevr2r3process_operations*                            .            z3PageObject._extract_text..process_operationrrrrrrrrrOsTDrrrrrsDorz/XObjectz/Subtypez/Imagez" impossible to decode XFormObject )r rrr,ror.rrrnrrrrr"rr&r~ IndexErrorextract_xform_textr warningswarnr)r&rr-rrZresources_dictrrrr-rrr_Zxobjr2r+r3 _extract_textQs   4,              zPageObject._extract_text)rrrr)cCs|||j|tjS)a/ Locate all text drawing commands, in the order they are provided in the content stream, and extract the text. This works well for some PDF files, but poorly for others, depending on the generator used. This will be refined in the future. Do not rely on the order of text coming out of this function, as it will change if this function is made more sophisticated. :param space_width : force default space width (if not extracted from font (default 200) :return: The extracted text )r2r-r|r)r&rrrr2r2r3 extract_textszPageObject.extract_text)xformrr)cCs|||j|dS)z Extract text from an XObject. space_width : float = force default space width (if not extracted from font (default 200) :return: The extracted text N)r2r-)r&r4rr2r2r3r/s zPageObject.extract_xform_textcCstdd|j||dS)zW .. deprecated:: 1.28.0 Use :meth:`extract_text` instead. extractTextr3)rr)rr3)r&rrr2r2r3r5s zPageObject.extractTextcCs<|}t|tstttt|d\}}||}||fS)z Get the names of embedded fonts and unembedded fonts. :return: (Set of embedded fonts, set of unembedded fonts) r)r.r,rr^_get_fonts_walkr )r&rZfontsZembeddedZ unembeddedr2r2r3 _get_fontss zPageObject._get_fontsr2cCstdd|jS)V .. deprecated:: 1.28.0 Use :py:attr:`mediabox` instead. mediaBoxrrrr=r2r2r3r9s zPageObject.mediaBox)r6r)cCstdd||_dS)r8r9rNr:rAr2r2r3r9s z/CropBoxcCstdd|jS)zU .. deprecated:: 1.28.0 Use :py:attr:`cropbox` instead. cropBoxrrrr=r2r2r3r;s zPageObject.cropBoxcCstdd||_dS)Nr;rr<rAr2r2r3r;s z /BleedBoxcCstdd|jS)zV .. deprecated:: 1.28.0 Use :py:attr:`bleedbox` instead. bleedBoxrrrr=r2r2r3r=s zPageObject.bleedBoxcCstdd||_dS)Nr=rr>rAr2r2r3r=s z/TrimBoxcCstdd|jS)zU .. deprecated:: 1.28.0 Use :py:attr:`trimbox` instead. trimBoxrrrr=r2r2r3r? s zPageObject.trimBoxcCstdd||_dS)Nr?rr@rAr2r2r3r?s z/ArtBoxcCstdd|jS)zT .. deprecated:: 1.28.0 Use :py:attr:`artbox` instead. artBoxrrrr=r2r2r3rA#s zPageObject.artBoxcCstdd||_dS)NrArrBrAr2r2r3rA-s cCsd|kr dStt|dSdS)N/Annots)r rr=r2r2r3 annotations2szPageObject.annotationscCs$|dkr|td=n ||td<dS)z Set the annotations array of the page. Typically you don't want to set this value, but append to it. If you append to it, don't forget to add the object first to the writer and only add the indirect object. NrC)r rAr2r2r3rD9s  )NN)NNN)NNN)F)NNF)F)F)F)F)F)F)F)F)F)rr)rrr)r)rr)Vrirjrkrlr rrrLrrtrmr rnrrrrrfrrrrr rrorrrrrrrrboolrrrrrrGrrrrrrrrrrr`rrrrrrrrr|rr2r3rr/r5r r7rErrrCr#r9setterrr;rr=rr?rrArrD __classcell__r2r2rwr3rps   '           _ /  &      & %  9 &      N 9           rpc@s`eZdZegefeegefddddZedddZeedd d Ze edd d Z dS) _VirtualListN)length_function get_functionr)cCs||_||_d|_dS)Nr)rIrJcurrent)r&rIrJr2r2r3rLIsz_VirtualList.__init__rMcCs|Sr*)rIr=r2r2r3__len__Rsz_VirtualList.__len__)indexr)cst|tr:t|tt}|jfddSt|tsLtdt}|dkrd||}|dkst||kr|t d |S)Ncs |Sr*r2)idxindicesr&r2r3r?Yr@z*_VirtualList.__getitem__..z!sequence indices must be integersrzsequence index out of range) r,slicerrPr~rrLr TypeErrorr.rJ)r&rMclsZlen_selfr2rOr3 __getitem__Us  z_VirtualList.__getitem__ccs tt|D]}||Vq dSr*)rr~)r&rr2r2r3__iter__dsz_VirtualList.__iter__) rirjrkrrrprLrLrTrrUr2r2r2r3rHHs   rH)rfntembr)cs|dkrt}|dkrt}tds2ttfSd}dkrR|ttddkrfdd|Dr|ttdD]}ttt|||q||fS)a_ If there is a key called 'BaseFont', that is a font that is used in the document. If there is a key called 'FontName' and another key in the same dictionary object that is called 'FontFilex' (where x is null, 2, or 3), then that fontname is embedded. We create and add to two sets, fnt = fonts used and emb = fonts embedded. Nr)z /FontFilez /FontFile2z /FontFile3z /BaseFontz /FontNamecsg|]}|kr|qSr2r2rrr2r3rsz#_get_fonts_walk..)sethasattraddr rorr6r)rrVrWZfontkeysrr2rXr3r6is    r6)NN)>rbrr0decimalrtypingrrrrrrr r r r r Z_cmaprr_utilsrrrrr constantsrr|rrerrorsrrZgenericrrrrrrr r!r"r#r$r%ror4r5rnr/r8r:r;rCrErFrGrprHr6r2r2r2r3sd 4  8     Vu#