Transformer Configuration

purple importer summary the purple transformer analyses input documents ingested via purple importer, allows users to select configurations tailored to their needs, performs transformations, validates the results and generates the desired output formats its range of applications includes document conversion, customisation, access settings, workflow automation and content management these configurations are set in the config properties file that is imported with the content in purple importer transformer configuration transformer configurations support subdirectory specific config properties files that selectively override properties from a parent directory, allowing different import settings per subfolder while sharing a common base configuration purple ingest ingest automatically merges config properties files based on the folder hierarchy — no explicit parent reference is needed a subfolder configuration inherits all properties from parent directories and overrides only what it explicitly sets teamid/config yml ← team root only teamid/upload/config properties ← base config teamid/upload/subfolder/config properties ← automatically inherits from above legacy importer subdirectory configurations must explicitly declare their parent using a parent= property pointing to the parent config file teamid/upload/config properties teamid/upload/subfolder/config properties ← parent= /config properties the parent= key is not recognized in purple ingest when migrating configurations from the legacy importer, these references can be safely removed pdf transformer parameter type default description pagesperstage int (>0) 2 number of pages except for start and end (see below), for single page pdfs use 1, usually it's a double sided pdf, so 2 is default pagesperstagefromstart int (>0) 1 number of pages in the pdf that belong to a start group, which has the pagesperstagefromstart number of pages on each stage, for double sided pdfs it's 1, because there is one cover page numberofpagesfromstart int (>=0) 1 number of pages per stage for the start group, for double sided pdfs it's 1 because there is a single page at the beginning pagesperstagefromend int (>0) 1 number of pages in the pdf that belong to an end group, which has the pagesperstagefromend number of pages on each stage, for double sided pdfs it's 1, because there is one final page numberofpagesfromend int (>=0) 1 number of pages per stage for the end group, for double sided pdfs it's 1 because there is a single page at the end removetitlepage bool false if true the title page is removed this might be useful for pdfs that are starting with a double page and are processed by x cago, as they will prepend a blank single page in order to process everything correctly preview bool false flag indicating whether the transformation is made for a preview version or for the complete version maxpagesofpreviewversion int (> 0) maximum number of pages of a preview version of the issue (including the numberofnonepreviewstages), if not declared all pages are included, please be aware that the purchase suggestion will be only shown on stages that are part of the preview but go beyond the numberofnonepreviewstages for example for a magazine with a cover page a maxpagesofpreviewversion=9 and a numberofnonepreviewstages=4 will show the first 7 pages (i e 4 stages) normally and pages 8 9 (i e stage 5) blurred and having a purchase suggestion numberofnonepreviewstages int (>=0) 0 the number of stages that are displayed completely (not blurred) numberofnonepreviewpages int (>=0) 0 the number of pages that are displayed completely (not blurred), this is to be preferred over numberofnonepreviewstages previewpages string if dedicated preview pages shall be rendered, here one can specify which pages are part of the preview it's given like this 1,2 3,7,9 if set, maxpagesofpreviewversion and numberofnonepreviewstages/numberofnonepreviewpages are ignored and numberofblurredpages, numberofpurchasesuggestionpages and blurredpagesincludedinpreviewpages is used to calculate the total number of pages (these pages plus blurred pages at the end if blurredpagesincludedinpreviewpages is false) usually this is set by the corresponding csv column for each issue this value is a default if no value is set in the importer csv numberofblurredpages int (>=0) 2 if previewpages is set, this value specifies the number of blurred pages rendered either after the preview pages ( blurredpagesincludedinpreviewpages = false ) or from the number of blurred pages is rendered from the end of the declared preview pages ( blurredpagesincludedinpreviewpages = true ) these blurred pages will show the purchase suggestion if numberofpurchasesuggestionpages is not set to specify a different number of pages with purchase suggestion blurredpagesincludedinpreviewpages bool false if previewpages is set, this flag indicates whether the numberofblurredpages are rendered additionally to the declared preview pages (if false ) or from the end of the declared preview pages (if true ) numberofpurchasesuggestionpages int (>=0) usually the numberofblurredpages also defines which pages show the purchase suggestion if this should differ (e g there are no blurred pages but there needs to be a purchase suggestion at the end) this value specifies the number of pages from the end that show a purchase suggestion minzoom decimal (>0 0) 1 0 minimum zoom level, which is 1 0 for original size, if a four times ratio zoom out should be enabled it would be 0 25 maxzoom decimal (> minzoom ) 4 0 maximum zoom level (for zooming in) stagewidth int (>0) 768 this is the width of the stage, which is a virtual number of points not related to any resolution of images the height is calculated from the aspect ration of the pdf normalresourceoutputtype string (pdf/jpg/png) pdf determined the format of the normal page output forcergbprofile bool false if set to true all pages with cmyk profile will be forced to rgb output this may result in inappropriate colours due to cmyk profiles not correctly transformed to rgb pdfdownsamplingenabled bool false downsample images when the normalresourceoutputtype is pdf normalresourceoutputdensity int (>0) 100 density of the normal page output this is usually only used for images (jpg/png) but also for pdf when downsampling is enabled (see pdfdownsamplingenabled ) normalresourceoutputantialiasing int (>=0) 1 anti aliasing for normal page output of images are used instead of pdf (0 = off), higher values lead to longer transformation time , normally 1 is sufficient normalresourceoutputimagequality decimal (0 0 100 0) 100 0 the quality for normal resource images, relevant only for jpg image output thumbnailactive bool true determines if as a background of the page a thumbnail is rendered which is shown until the normal page output is rendered pageimageheight int (>0) 200 height of the page thumbnail that is rendered in the background of a page, this number determines the quality of the loading image especially relevant for preview quality where no normal page is displayed, adjust this to a lower value if the preview is not blurred enough thumbnailoutputtype string (jpg/png) jpg format of thumbnail images (usually jpg as they are blurry anyway) tocthumbnailheight int (>0) 200 height of the thumbnail image displayed in the table of contents (the higher the better quality) tocthumbnailoutputtype string (jpg/png) jpg format of the image displayed in the table of contents extractcoverpath string if in addition to the pkar a cover shall be extracted, this determines the file path to save the cover (including the file type jpg/png) thumbnailheight int (>0) 512 height of the cover thumbnail that is rendered in the background of a page, this number determines the quality of the cover image adjust this to a lower value if the preview is not blurred enough transparentstagebackground bool false if true, the background of the stage, the pdf is rendered on top, will be set to transparent instead of white this way the surface of the app/web background will shine through transparent pdf pages splitlandscapepages bool false if true all landscape pdf pages will be split into half horizontally usetrimbox bool false cut pdf to trim box before processing pdfsinglepageuploadenabled bool fase if enabled the pdf is also saved in single pages under the pdf output path with a suffix for each page and a json file besides describing the pages splitintosmallfiles bool false if pdf single pages within the pkar are very big, so that the overall pkar is too large, this parameter may help to minify the single pdf pages ignorepdflinks bool false if set to true, all links in the pdf are ignored by default links are transformed to touch areas opening the referenced page or external link ignorepdflinksinpreview bool false if set to true, all links in the pdf are ignored for the preview version of the issue by default links are transformed to touch areas opening the referenced page or external link if ignorepdflinks is set to true, links are also ignored for the preview version of the issue ignorepdftoc bool false if set to true, the table of contents that is given as part of the pdf is ignored by default the toc of the pdf is used to create the toc with corresponding labels for the transformed content sectionlabelintoc bool false include the section label in table of contents entries transformfulladvertisementpagestoimage bool false transforms all full advertisements pages to jpg images ( only if pdf is normalresourceoutputtype and an epub is provided to recognise advertisement pages ) transformpartialadvertisementpagestoimage bool false transforms all partial advertisements pages to jpg images ( only if pdf is normalresourceoutputtype and an epub is provided to recognise advertisement pages ) renderhtmltoc bool false renders an article based read mode styled table of contents besides the cover page ( only if an epub is provided, so an article toc is available, which can also be styled with custom css and custom js from dynamic resources ), it rendered on a separate page if renderhtmltocafterpage is set renderhtmltocafterpage int if set (and renderhtmltoc=true ) the html page is not rendered besides the cover page but after the given page as a standalone page it has an entry in the issue toc if renderhtmltoctitle is also set renderhtmltoctitle string if set (and renderhtmltoc=true and renderhtmltocafterpage is also set) this is the title of the toc entry (e g "inhaltsverzeichnis" or "table of contents") issuesourcenametemplate string a template that defines how the issue source name is compiled the issue source name is a custom property used in purple apps, purple manager and purple hub to tag every content piece that is related to the same issue the template contains parameters that match columns of the csv in braces, e g {publication} or {issue number} also custom properties from the csv may be included as {custom property name} multipagepdf bool false transform all pdf pages on one stage into a single multi page pdf this is relevant for printing feature as printing works only with a multi page pdf this does only work with pdf output type, not jpg linkwithalignment bool false transforms links to action urls that link to the stage and scroll to the correct page (e g on double side pdf pages in portrait mode) htmlpurchasesuggestionurl string if the purchase suggestion should be rendered as html this specifies the url to the html page htmlpurchasesuggestioncoordinates string if htmlpurchasesuggestionurl is given this is the set of coordinates the specify the relative position of the html page rendered on top of the pdf the coordinated are given as x1,y1,y2,y2 where each value represents the factor of the coordinate multiplied by the x or y dimension for example does 0 2,0 3,0 8,0 7 on a 500x1000 page render to (100|300) top left and (400|700) bottom right linkhighlightcolor string the highlight color of the link specified as #rrggbbaa (a=alpha) linkhighlightanimationperiods string if linkhighlightcolor is given this defines the fade in, stay visible and fade out period of highlighting in ms for example does 200,500,300 fade in for 200ms, let it stay visible for 500ms and fade out for 300 ms if fade in is 0, it immediately appears, if stay visible is 0 it stays visible for ever, if fade out is 0 it fades out immediately (only if stay visible is > 0) puzzlelinkhighlightcolor string the highlight color of the link specified as #rrggbbaa (a=alpha) puzzlelinkhighlightanimationperiods string if linkhighlightcolor is given this defines the fade in, stay visible and fade out period of highlighting in ms for example does 200,500,300 fade in for 200ms, let it stay visible for 500ms and fade out for 300 ms if fade in is 0, it immediately appears, if stay visible is 0 it stays visible for ever, if fade out is 0 it fades out immediately (only if stay visible is > 0) readmodelinkhighlightcolor string the highlight color of the read mode touch areas specified as #rrggbbaa (a=alpha) readmodelinkhighlightanimationperiods string if readmodelinkhighlightanimationperiods is given this defines the fade in, stay visible and fade out period of highlighting in ms for example does 200,500,300 fade in for 200ms, let it stay visible for 500ms and fade out for 300 ms if fade in is 0, it immediately appears, if stay visible is 0 it stays visible for ever, if fade out is 0 it fades out immediately (only if stay visible is > 0) readmodeexternallinkhighlightcolor string the highlight color of the read mode external links specified as #rrggbbaa (a=alpha) readmodeexternallinkhighlightanimationperiods string if readmodeexternallinkhighlightcolor is given this defines the fade in, stay visible and fade out period of highlighting in ms for example does 200,500,300 fade in for 200ms, let it stay visible for 500ms and fade out for 300 ms if fade in is 0, it immediately appears, if stay visible is 0 it stays visible for ever, if fade out is 0 it fades out immediately (only if stay visible is > 0) dynamicadvertisingconfig \<id> string | json a dynamic ad config ingests html pages from dynamic app resources at specific pages the id is referenced from the csv file via the "ads" column it may contain a comma separated list of dynamic ad configuration ids example for the json format for a dynamic ad configuration { "\<page number>" { "view" "\<relative path to file in dynamic resources>", "zoom" true|false (allow zoom), "toc" true|false (show in toc), }, } contentsfitstrategy string pdf determines if the page is automatically zoomed to fit the height of the screen ("pdf" value) or the width ("magazine" value) hubinterstitials bool false import hub interstitials (ad/content pages) into the issue (additionally requires hub url and other access parameters from hub import configuration) tochubinterstitials bool false include hub interstitials in the table of contents zoomhubinterstitials bool true allow zooming of hub interstitials eveningpublishingtime string evening publishing time ( hh\ mm x ) used to compute the publishing date of the issue (also affects the interstitial display window) read mode transformer (epub based) see also our latest epub format specification (word doc and pdf below) parameter type default description noreadmodeinpreview bool false if true the read mode will not be created for preview versions, also if skiparticlecreation in combination with externallinkschema and webviewurlschema is used for some read mode configurations, no touch areas will be generated readmodepackage string path to the read mode styling package (deprecated as read mode styling can also be added to dynamic resources) puzzlepackage string path to the puzzle styling package puzzlesolutionurl string url for the puzzle solution action button (which is not displayed if empty) readmodeversion string 1 use 2 for modern read mode, especially for tts read mode readmodestyle string a specific class name that is added to the body and container div of each read mode article for individual styling readmodetoc bool false display table of contents in read mode ignoreadvertisement bool false do not create read mode articles for advertisement articles (false=show advertisement articles) ignoreadvertisementintts bool false if ignoreadvertisement is false (advertisements are imported as articles) and tts is active this ignores those articles in tts optimizetouchareas bool false if set to true the touch areas leading to articles are combined to optimal fitting rectangles not intersecting different articles on one page this is particularly important if the new read mode is used, where articles from purple hub are pushed to the app as these touch areas are web views the less web views there are on a page the better it loads pages forceonetouchareaperarticle bool false force at most one touch area per article (results usually in more overlapping touch regions) ordertouchareas bool false in some cases, the epub may contain article block definitions where an inner rectangle is followed by an outer, which results in the bigger touch area overlying the smaller one if ordertouchareas is set to true, touch areas are sorted in a way that inner rectangles always follow outer ones addemailwwwlinks bool false auto detect e mail and www addresses in article text and turn them into links stripcolors bool false strips all individual colors from read mode content awspollylanguagecode string language code for aws polly tts, see also awspollyvoices string list of voices for amazon aws polly tts separated with a comma with it's gender (male or female ) separated with the pipe character | (e g "hans|male,marlene|female"), see also aws tts voices https //docs aws amazon com/polly/latest/dg/voicelist html additionally the "#" separator may be used to opt for another speech engine like "neural" example "brian#neural|male,emma#neural|female,amy#standard|female" awspollyaccesskey remove for ingest docid\ uepzjnme7gszocvwsa8ox string access key for aws polly tts voice output generation this access key needs to belong to user that is configured in aws iam service with permission to create files on the s3 bucket (see awspollybucket) and also has permissions to fully access aws polly awspollysecretkey remove for ingest docid\ uepzjnme7gszocvwsa8ox string secret key for aws polly tts voice output generation this secret key needs to belong to user that is configured in aws iam service with permission to create files on the s3 bucket (see awspollybucket) and also has permissions to fully access aws polly awspollyregion remove for ingest docid\ uepzjnme7gszocvwsa8ox string aws region for the s3 bucket to store for aws polly tts voice output generation awspollybucket remove for ingest docid\ uepzjnme7gszocvwsa8ox string s3 bucket to store for aws polly tts voice output generation awspollybucketkeyprefix string prefix for all paths on the s3 bucket, where aws polly tts voice output generation is stored (optional) awspollysamplerate int 22050 sample rate of the tts sound files awspollydelay int 0 delay between polly tasks in milliseconds, this can be tweaked to avoid polly api exceeding limits awspollyreadtitle bool false flag indicating whether the title should be read prior to reading the content includepollyclasses string (all) comma separated whitelist of css class names to include in aws polly tts generation excludepollyclasses string comma separated css class names to exclude from aws polly tts generation ignoreadvertisementintts bool false flag indicating whether advertisements should be ignored for tts skiparticlecreation bool false skip creation of read mode articles, used only for purple hub based article model on top of pdfs externallinkschema string schema for creating external links instead of read mode links for article touch areas (e g purple //kiosk/issue/by external id/{epub} {page} {seq}), used only for purple hub based article model on top of pdfs webviewurlschema string schema for url of web views that are rendered instead of article touch areas (e g resource //dynamic/storefront/dispatcher html?issue={epub}\&article={epub} {page} {seq}), used only for purple hub based article model on top of pdfs noimagescaling bool false per default all images that have more than 2048 width or height are scaled down to 2048 if this flag is set this downscaling is switched off, so that each image is put into read mode with its original dimensions indexarticlecontent bool false if this is switched on a search index is created from the epub article content this index is better than the standard content extracted from the pdf as this is a new feature, it is not active per default but highly recommended to be switched on customreadmodefilename string if empty, the custom read mode files from dynamic app resources are referenced by resource //dynamic/readmode/custom css and resource //dynamic/readmode/custom js if this parameter is set, the resource name "custom" is replaced with this value e g setting this parameter to "readmode" will result in resource //dynamic/readmode/readmode css and resource //dynamic/readmode/readmode js use this parameter to migrate from a deprecated custom read mode implementation to a new one enablearticleinviewdata bool false adds data about articles in view to content in order to support tracking of articles in view this requires the setting for articleidschema properly set, as this defines how articles are identified uniquely articleidschema string {source} template to define how articles are identified the template may use these variables {issue} "source name" of the issue, see issuesourcenametemplate {source} original file name of the article in the epub file {target} name of the article file in the pkar file {page} page number (starting with 1) the article starts on {seq} sequence number of the article on the starting page {title} title of the article {author} author of the article (if given in the meta information within the epub article) {section} section of the article (if known) phonetic pronunciation for tts if for tts pronunciation shall be optimized by adding a mapping for phonetic pronunciation, it may be added to the custom readmodepackage zip file referenced in the importer configuration mappings can be added with different phonetic alphabets like described in the aws polly documentation https //docs aws amazon com/polly/latest/dg/supportedtags html#phoneme tag the phonetic mapping file must be named phonemes ipa for ipa alphabet, phonemes x amazon pinyin for x amazon pinyin alphabet or phonemes x sampa for x sampa alphabet the mapping file consists of words mapped to phonemes like this pecan=pɪˈkɑːn pecans=pɪˈkɑːns please make sure you add every grammatical for of each word as the replacement of each word will be as one whole word, not as parts of a word or multiple words purple hub article importer (epub based) see also our latest epub format specification (word doc and pdf below) parameter type default description importhub bool false import articles into purple hub importhubonly bool false import articles only into the purple hub, i e do not create the normal issue output for that publication importpdfpages bool false import the pdf pages to the purple hub to support article modification with a pdf preview importpdftohub bool false import the pdf itself to the purple hub, so it can be downloaded from the web presenter baseurl string the purple hub base url username remove for ingest docid\ uepzjnme7gszocvwsa8ox string username used to import articles to the purple hub (e g "importer" or "importmanagerhub\@sprylab com" if direct publishing to app is used ( publishlive or publishpreview ) , for this use case the user needs to be added to the purple manager team as an editor with access to publications that the content is published to) password remove for ingest docid\ uepzjnme7gszocvwsa8ox string password of the user used to import articles to the purple hub (for importermanagerhub\@sprylab com look ) purpleteams string the team id for the posts publishliveweb bool false publish to web channel publishlive bool false publish to app channel into a live version publishpreview bool false publish to app channel into a preview version accountid string team of the app/publication the live content is published to (required if publishlive or publishpreview is active) appid string app the live content is published to (required if publishlive or publishpreview is active) publicationid string publication the live content is published to (required if publishlive or publishpreview is active) isissue bool false define whether content should be posted as issue or as simple articles posttype string post name of the post type that is created for each article during the import includecss bool false include css classes from epub by mapping ( this should be set to true for proper styling options ) author int the author of the posts (wp user id) not required covercategory string cover default category for cover (if isissue=true the cover is also imported as a post, so it may have a category) issueaccesslevel remove for ingest docid\ uepzjnme7gszocvwsa8ox int 0 issue access level as defined in the purple hub paywall options settings (0=first entry, 1=second entry, 2=third entry) articleaccesslevel int 0 article access level as defined in the purple hub paywall options settings (0=first entry, 1=second entry, 2=third entry) accesstype string paid used for ingest imports using the name scheme collector available options free, paid, locked adsfreeaccess bool false if issueaccesslevel is not 0=free and adsfeeaccess is true, ads will be imported as free articles ignoreimages bool false ignore all images in articles (like in some special configuration's archives) ignoreadvertisement bool false do not import advertisement articles (false=import advertisement articles) ignoresections bool false ignore sections (i e do not set them as category), per default the section is added as a category to each post skipwithtag string skip all articles with one of these tags (separator |), tags are coming from the div container within the body and may be "coverstory", "cover", etc includewithtag string include only articles with one of these tags (separator |), tags are coming from the div container within the body and may be "coverstory", "cover", etc skipwithsection string skip all articles with one of these sections (separator |) skipwithtitle string skip all articles with one of these titles (separator |) sectionastext bool false put section into content as text as a first element mappedclasses string mapping of epub classes to purple hub classes mappings are separated with | and defined by epubclass=hubclass, example article pull quote=quote|article standfirst=standfirst|article callout=callout articleheadercontainer bool false put article header in a header container articleheadingcontent bool false put article heading into content as text dropcaps bool false automatically add drop caps to the first paragraph dropcapsminletters int minimum number of letters of a drop cap paragraph (e g 300) featuredimage bool false set featured image (thumbnail) for each article if available gallery bool false import multiple images in a row (i e without text inbetween except image captions) as gallery (slider gallery) imagegrid bool false import galleries with small images (< 768 width) as image grid (masonary component) hero bool false import first big image as hero image heroparallax bool false set parallax effect on hero image imagemaxsize int maximum size of the images (all images with a height or width larger than this will be downscaled) imageminsize int minimum size of the images (all images with a height or width smaller than this will be ignored) stripcolors bool false strips all individual colors from epub input stripdropcaps bool false strips all drop caps from epub input groupboxouts bool false import boxouts as core group blocks rather than individual blocks usebylineasexcerpt bool false use the article byline as excerpt instead of the subheadline ignoretitlenode bool false ignore the in content title node and use the metadata title instead trimwhitespaces bool false trim excessive whitespace from epub content standfirstacf string name of the acf field the standfirst/deck paragraph maps to mediaissue bool false add the issue as a taxonomy term on media/images specialissue bool false mark the issue as a special edition hubarticleidattribute string article data attribute holding the exported hub article id from the print system skipsortorder bool false this needs to be set to "true" for importing into the new purple hub (like c01 purplsdshub com, demo purpledshub com or news purpledshub com) otherwise the import will fail! for older instances of the purple hub this setting must not be used as otherwise the import also will fail updatemode string default is that articles in purple hub are always created as new ones if this is set to update all existing articles are updated if is set to fail , the import fails if any article is already present in the purple hub config string standardmag | fullmag | fckoeln presets for certain types of import configs config flags standardmag isissue=true includecss=true importpdfpages=true featuredimage=true dropcaps=true dropcapsminletters=300 gallery=true imagegrid=true articleheadingcontent=true fullmag isissue=true includecss=true importpdfpages=true featuredimage=true articleheadercontainer=true hero=true heroparallax=true dropcaps=true dropcapsminletters=300 gallery=true imagegrid=true articleheadingcontent=true isissue=true includecss=true importpdfpages=true featuredimage=true gallery=true imagegrid=true articleheadingcontent=true sectionastext=true you may overwrite other flags or one of the flags predefined by the config by declaring those additionally aws polly credentials in purple ingest , the aws polly credentials and bucket configuration ( awspollyaccesskey , awspollysecretkey , awspollyregion , awspollybucket ) are managed by the platform via the team's bucket configuration in purple manager any values set in config properties for these properties are silently overridden remove them — they are provisioned automatically other polly settings such as voice selection can still be configured in config properties hub article importer properties in purple ingest, hub authentication uses jwt bearer tokens issued by the platform explicit username and password credentials in config properties are not accepted and will not authenticate remove both properties — authentication is handled automatically via platform issued jwt tokens properties specific to other transformers funkinform dialog parameter type default description moveimagestotop bool false move article images to the top of the article boxoutstarter string comma separated box names that start a boxout section boxoutcontinuation string comma separated box names that continue a boxout section ignoredboxnames string comma separated exact box names to ignore ignoredboxnameprefices string comma separated box name prefixes to ignore (prefix match) alfapeiq parameter type default description zoomads bool false allow zooming of advertisements tochasads bool false include advertisements in the table of contents ignorewndata string alfa\&peiq only comma separated publication/file names whose peiq data is ignored editions {editioncode} string alfa\&peiq only dynamic mapping of an ad edition code to a publication edition (e g editions voz ) peiqsection {sectioncode} string alfa\&peiq only dynamic mapping of a peiq section code (uppercase) to an internal section name peiq parameter type default description skippeiqstandaloneimage bool false skip creating articles for standalone images wwprevion parameter type default description tocfromdigdump bool false build the table of contents from the wwprevion digdump file edition ww {value} string dynamic per edition mapping ({value} = article edition id)