GB2527383A - Content transformation - Google Patents

Abstract

Content transformation includes transforming of content items 104 in a content management system (CMS) repository from a source format to a target format. The transformations can be performed using a transformation/computing node cluster 110 having multiple nodes, each of which is configured for a specific content transformation type (e.g. PDF transformations, image processing). Router nodes 112 can receive requests for content items and route content items to transformation nodes that are configured for a specific content transformation type to either transform a requested content item to the target format or perform an intermediate transformation as part of a transformation chain. The transformation node cluster 110 can be dynamically configurable based on estimates of expected loads for the various types of transformations. The transformation nodes maybe assigned among available defined transformations to increase efficiency if a threshold metric such as wait times are crossed. Load balancing can be accomplished automatically as the router nodes 112 direct content items for transformations simply based on their current and target content formats.

Description

CONTENT TRANSFORMATION

TECHNICAL FIELD

[OOfl The subject matter described herein relates to content management systems (CMS) and, so e unplementat on, ror spteficalJy to n UaticaHy scalable content transformation service for transforming eloudbased content from a source format to a destination format and to transforming of content items using a transformation node cluster.

BACKGROUND

j092J Fntemrise content management (ECM) covers a broad range of applications.

including document management (DM), Web content management (WCM), records management (RM, digital asset management (DAM, search of managed content, and the like.

A content rnanagerrent system CMS) sultable for managing the various content items (also referred to herein in some examples as files, documents, or the like) that an enterprise produces or generates, retains or otherwise stores, manipulates or modifies. ete, can support the requirements of one or more of such applications, and optionally other requirements, to provide a coherent solution in which content and management processes are capable of accessing content across a variety of applications subject to access controls, permissions, and the like, An "enterprise" can generally refer to an organization, such as for example a business or company, a foundation, a university, or the like, and can have content requirements related to one or more business processes, content uses, etc. [003] A CMS manages the actual digital binary content in a content item, the metadata that describes a context of each content uem, associations between a content item and other content or content items, a place and classification of a content item in a repository, indexes for imding and accessing content items, etc. The CN4S can also manage processes and lifecyeles of content items to ensure that this information is correct. The CMS can further manage one or more processes (c>g. business processes, workflows, or the like) thr creating, capturing, storing.

manipulating, editing, mid disn'ibuting content items, as well as all parts of a content item lifecycle (e.g, how long a content item is retained, what actions are taken after the end of a retention period, etc.), [004] A CMS for use in enterprise content management can include one or more of JoL urrent managcment tcols, apphauons, nid ntertacc to support gwcral office 1⁄2nrk, search anc Jiscosuy No know rarageinent capahiut es of a CJVS Lcn support ni nerou\ bL\1nes proce'scs opt onafly rncludln3, hut rot lirted to, ae mar ageninit, rsww anc, appiova and the like Co Ia orat on apohcauons and se ices of a (MS au uppott t e collaboiatn e Lkveloprlent of rn4orniatio i a d knowkdize ir the ceat m and retbemrent of co" en and documents V eb co ite I manageme i servit c of a CMS wh eh can he sa able, can supoort tic delivery and deployment of' content from the enterprise to its customers. Records management capabilities of a CMS can optionally capture and preserve records based upon government-approved or other standards. A standards-based platform can also provide access to applications that use these standards, such as publishing, image management, email management. etc. [005] Servic&s and contxds for managn content o, a can n. luec Feat.. res such as metadata management, version control, lifccycle management, worktlow, search, associations to other content taggi is, o nmnen n...g, cu mha alkw users of IN. CVS to F d deind c intent items among very large content collections that can span muitipic parts of an enterprise.

Accuracy and consistency of the information can also be ensured, even for very large content collections across an enterprise. Content and other information in a CMS can be presented.

n ihhshed ctc trrough the Web or an otner dun id anpropr at in allow jsu to access (ha information,

SUMMARY

10061 As discussed in greater detail below, features of the current subject matter can include capabilities relating to efficient trans-tbrnation of content items ftom a source format to a target format. A router node can receive a transformation request identifyinc. a content item and a target format for the content item, The router node can direct the request to one or more transformation nodes within a transformation node cluster, The one Or more transfbrmation nodes can be pre-configured to perform specific transformation required as at least part of a transfrirniation chain to convert a content item front its source format to the target format, [007J In one aspect a corputenirrnlr K ted mdhod nicluces routLr node o" a traristornia on iode clL stel that mchiics a phrra ny ol nodc rcccn mg a iequcst horn a chcnt mzchme for a content ik in toied a a epos my assonated wuh a n'er machine that s assocatcd t} the tramfornat,on node cLster He jar'sforg mtion modc i. ter inUuds a p uahty ol trcnstorrnciton noces, and each of the pLiralit of trrstorrnation no,cs Is configi. red to exccutc at least one out no all 01 a phuahty of o iert Lans!ormation t. es Tie mcthod further includes the router node identifying a current content format of the content item and a target content formaL of the content item where the. target content format differs from the current content format, Based on the identifying, the router determines one or more transformation types horn the plurality of content transformation types fbr transfbrmning the content item to the target format. The router node routes the content item to one of the plurality of transformation nodes in the transformation node cluster. The one of the plurality of transformation nodes is configured to execute one of the one or more content transformation Lynes determined for the content item.

908j In some variations, one or more of the following features can optionally he included in any feasible combination. Each of the plurality of transformation nodes can optionally be configured to execute only one of the plurality' of content transformation types.

The server machine can optionally be connected with the transformation node cluster by a communcation network. The plurality of transfbrmation types can include a transformation type that includes transforming the content item from the current content format (e.g. a portable doenrient Io"nat) no tie target content loi -nn, which irciudes the content item pht nb two or more scction e g in 11 e portahlc document foi nn) he pluraluy of tta3nAounrhon ypes can indud.e transthrm!ng the content item from the current content format in an editable text ocutnent tormat irto a tarbet contelit forrrat n ai image torirat k numbci of the ph raIny ot tran.forriatmn nodes in the trantonration note duster configured to txcci;c each of the rilu ahty ci concnt nnsformllon types is based on an es. nate a' trnstormat1or loads of different transtbrmation types to be handled by the transformation node cluster.

[009] The one or more transformation types determined by the muter can include a transformation chain comprising a plurality oltransformation types. The one of the plurality of transformation nodes in the transformation nod:e cluster can have a first transformation type to tiansi rn' thc content icm o a fir$ intirmediate cott it item ha\ ng a first inte'ec ate lormat The method can further include routing the first intermediate content item to a second node of the ph rality ot ransfonna Jon nodes n tie t'anc,.ormation node chitc having a scto transformation type to transform the first intemiediate content item to either a second intermediate content item hiving a second intermediate format or to a target content item having the target Ibnirat, The transihnnation chain can include a transfbrmat.ion tpe resulting in formation of mukiple intermediate content items, and each. of the muitinle intermediate content items 1equres a subsequent transfbrmation. The method can further include thc router routing the multiple intermediate content items to multiple transformation nodes in the. transformation node cluster, The multiple transtbrmation nodes can each he configured to execute a same one of a plurality of content transformation types such that the multiple intermediate content items are processed in paralleL (04)14)] The mcthod can also further include the router creating a transient handier that manages the request. The router also tags the request with a handier context and a request identifier, and maps a response received from the one of the plurality of transformation nodes and coma img the handier eoitcv aM mx ne ioe itfliei back to the.equ3st aid to the ttan'ienl hadkr IL t aniei randkt keep' a reord the response and notifies the rot,tci when ill transformations of a first transformation type of the one or more transformation types has been completed. The router can route intermediate content items resulting from the first transthrmation type to one or more additional nodes of the olurality of uansforrnation nodes in th transiorri a on rode cluster lie one or mom ruditionrl roocs of t1 o p <uahtv of LramRMmatl:n nodes can be confgured to trarsfo. ii the mtci n ed ate content Lern to o'ie or more next intermediate content items or to the target content format of the content item, [00111 In another aspect, a computerunpiemented method includes estimating, an expected load for each of a plurality of transformation types fin content item mquests from one ot more ci ent machines uelaung to contci c'us mainta ned u a co tent rnan dcnuent systcm repository. The estimating is based on transformation usage data, and can he peribrmed by at.

least ont comptnz sys-en't that irducc o puter haidwarc 3ach tiansfotniat on t\pe jf the

S

pluralit of transfbnnation types transtTorms a first content fbrrnat to a second content format differing from the first content ibrmat. The method further includes configuring a transfonnation node cluster that includes a plurality of nodes. The configuring includes designating each of two or more subsets of the plurality of nodes for executing one of a plurality of content tranStorrnaton types. Each of the two or more subsets has a designated number of nodes of the plurality ol> nodes, and the number of nodes is based on the estimated load for the one of the plurality of transtbrmation types for which that subset is designated. The method further includes assigning one or more router nodes within the plurality of nodes. The one or more router nodes are configured to receive a content item request from the one or more client machines and to route the content item to an arpropriate subset of the two or more subsets. The content transformation thr which the appropriate subset is designated matches a required transformation type of the one or more required transformation types.

1OOI2 H son e variauo Is, one o mote of the to org feat ires can optioflally K nuk ded in any teasibie comi. mat on At least one of the one or m e i outer r ode can also he included in one of the two or more subsets of the plurality of nodes. The method can further mc tide renf g.uing t e trc stniation ned ci is cr, thc e contigunng compi istng changing the designated number of nodes of the plurality of nodes for at least one of the two or more subsets based on a changed estimate of the expected load.

OO l3 ihe one o morc wutei uode can be further cot flgurcc to uenLt a eunent content format of a content tern referenced in a received request of the content item requests and a tin get ontent foi inst of tie conten era spccit'h d in ic recen ed \qucst a d r duc"m're o ic or more required transformation types for the content item from the plurality of content transformation types.

[0014J The ore or mere required transformation types can include a transformation chain that incudes a p!urality of transformation types. [he appropriate subset can include a first transformation type to transform the content item to a first intermediate content item having a first intermediate thrrnaL The one or more router nodes can he further configured to route the first intermediate content item to a second appropriate subset, which can have designated for it a second content transformation, The second content transformation can match a second required transformation type of the one or more required trans%rmation types br transforming the first intermediate content item to either a second intermediate content item having a second intermediate format or to a target content item having the target lbrmat.

[00151 The transftrmnatiori usage data can include historical data pethrining to content transformation requests requested from the transformation node cluster. The estimating can tOrther include use of at least one of a predictive model, machine learning, and a neural network to make predictions based on the historical data. The stored data can include stored calihacks provided by the one or more router nodes to client machines in response to previously completed content item requests. The stored cailbacks can include at. least one of references to requested content items, arrays of iransfonnod content item references referencing intermediate content items created in transformation chains to produce transfOrmed content items in response to the kompleted content item reouess, snng o2 optiors spec fed n tie co itent em rc4Jcts, htmgs ofopuatlon\ peiformed to convert content flows to tircz content tormars, and iderufieis [Or transformation nodes that participated in the transthrrnation chains. The transformation usage data can further include at least one of a type and number of business process applications supported by the content management system repositoiy, usage factors defined for content item transformations expected to he called by the business process applications, and an absolute or relathre amount ofmobllt;devht us rnpw by the Th estifttathig can mitt ilioludo aeThig on at k one huts input psnitot Slating to esp:ted trhsthtnisio flr*t.

tOO 1:61 current subj& mth ean ibàlOd; but arçrngt thnfte, tq cthSs *tát with: iw ckscripthnproYJ4S herein as well as Siólès that comprise; a tang3bIy cn*$!, in...cc4aS iqc,. opethk çue one cr m, macbines t;g,s COfliPutell, eteC) to flSUit Ifl Pq$: i$Jç e$ing one or p1ore of the krje4 features.

SiiñThl, riptiter systi mnreal$ 4q4 ffiflaayiSk oncot arc proesser&andont or mote meffi*ttes coupled to the one or more processors A morn cny, vttch can include a omp@tet-readable storage medium, may £nøude, enco4e *rc or the like c or moçe ptgtAs tháUóàUófOiSfdE mOre processors to perfonwqneqr nerrofthc op 4iSdespr!bed )fl. COMptitt hØleftstS Ehtht coh t whit one or more Irnplementh lens 01 the tthttiib$ct thâttEnEfl he itOflISóhWd by óñó or more 4am processors nwi4tg h a single tomnputthg ystefti otsittflttpie ffi*t qtit Su& muttipló computhg yp c connectedkand can exchange dam and/or commands otah mntructionsov the ke 4aow or more connet'ons, indudingbut nUt Iliultod tOA tóótlOñ OVer network (;g. the Jn1a whles t'4 c: rea etk,tIe1 aa netwok a wi atea ñCtWotk a itt network or the gko, via a4kect connection between ona ormore of the mu i puthtg tv (0917] The detaVkofonetnnorevathtions qi eiiWec motfcrøesithed herein are set forth in the çç yi.g *in p. ripu s'o Other iat resand Mvantages at the subject mat er 4cc*e4! 41 rppIt from the di fiàmand4mvóings and frca the Sims. While certain features Of the cunerKly 4i:t skject matter n dibS:c* i1Iustairt puiposes in rck4Ipn, to cntc$ :cc'* s an. enteprise. sc!ucce management system, a records management system, or other business software solution or architecture, it shou]d be readily understood that such features are not intended to be limiting.

The claims that Ihilow this disclosure are imended to define the scope of the protected subject matter.

DESCRIPTION OF DRAWINGS

[0018] The accompanying drawings, which are incorporated in and constitute a part of this specification, show certain aspects of the subject mattet' disclosed her&n and, together with the description, help explain sonic of the principles assoeiatcd with the disclosed tmplementations. In the drawings.

J0019j FIG. I shows a diagram illustrating a system having features consistent with implementations of the current subject matter; [0020] FIG. 2 shows a diagram illustrating aspects of a load balancing arrangement; [0021] FIG. 3 shows a process flow diagram of a method for elastic se.alabflity for cloud based content trarxsfbrnations consistent with implementations of the current subject matter; 100221 FIG. 4 shows a process flow diagram of a method for scaling cloudha.sed content transformations consistent with implementations of the current subject matter; and [0023] FIG. 5 shows an example of a transformation chain consistent with implementations of the current subject matter.

10024] When practical, similar reference numbers denote similar structures, features, or elements. C)

DEThULD DESJ: U PTIOIN 1OC25] Impteeisns of th ufrentthJet s S a Other SibIt advantages, prSáie ohugie &cattbThy apptoahes fbr tk u.based toitent; trantthffleiE teatur, oP which are thusntS b t diagram I® of EIQ. l, wbkb shows at example:of a computing framework or architecture conSatent; with such; ap: gg* A lttnt nStSw2 can bç a desktqp orl4top computcr or wpkkflaS,n a server, a mobile device (e.g. a tablet or nmphonecth;), rany other machine that can requost a contenr item 104 needed by MciIent machine tt the SMS item 104 can have aourrent content tbrmat if the cuwcnt twusm cnsiset*: w!th a target content fàrrnat needui 1iy the client machine ia the: content , 1Q4 c, be 4økw*l dfr,c4 to the client macithie IS, for example by acessing the cos,t il,,.: 14)4' frm a content management repS tory 108 where; the; content item Ió4 S !*$fl WTh eurrçnt c,t:1orma i& n4 conttt with;g. equivalent to) e srgeateflau, conSt Eranstbrmation can be'requfredto ptovlde the content fta UI to the client tathS I P, i,,*Se,:S, nt i, +ang the target content fbrmat A call or request for tcontent jpj qng!,,S$ Utmáthine from a webappiiea'ion a native applktitibn, a server app c(.. a;command line uWi*y amohikyappUcion (e.g.an Mipr) or hik C'S'uthaát the clii ntmach m 102, 14ö261 S:ióh transtonnatons can be Ecomp using a t ation node cluster it The transfórmatn node cluster tthhn inS,' a plural* ot aodcs (tg. Coff'flg nodes), *hith can include Øhsicai sndIor vhtuaV:niaclttes One ovnwnnodesofthep1uraJhy of s'fr. the cluster 10 ça be dcitgnatedira muter node liz whth can r%eWe requests frqm one or morcIicnt machines Y02 For eaóh requested content Itent 104the:rnuter node! t2 he,,cSit content format otthccontent item *)it an a target cqntat tbnnat of the

IG

content item 104 needed by the requesting client machine 102, Based on the identifying, the ot tet node 112 Lan drec he conten1 tern 104 to a transtorna# on noce of a plurality of transfoininiot nodes ir the nansfo' nation uodc LII sEer 1 10 Lach uch tr istotnuition noUc car, he configured to haitdle a specific type of content transformation, for example from a first content format to a second content fbrmat. Optionally, a transformation node within the tnnstbrmation node cluster 110 can be configured to handle more than one specific type of content transformations. 1--lowever, a transtormatjon node generally is not configured to perfbrm nil of the possible content transformations available to he performed by the transformation node cluster lift In FIG. 1, the transformation node cluster 110 includes four examples of transformation nodes: first transformation nodes I 14A con flgured to convert Microsoft Office firna (locumelt to an irnag tormat (e p JPEG Git etc), r secot c tranqfornaaon node 11 4B conligured to spin PDF documents into smaller groups of pages, third transformaton nodes 1 14C configured to convert PDF files to image flies., and fourth transformation nodes 1 14D co figined for image p'occssing ( a to ebaige an itrag swe, teo1utior et) A router node 112 can also optionally function as a transiormation node 110.

O027] The transformation node cluster I JO can be configurable in that transformation nodes can be assigned to any defined transfbrmation (current content format to target content format) An administrator can assign the transformation nudes among available defined transtbrmat.ions. Alternatively or in addition, router nodes can also configure themselves (e.g. without human interaction). For example, a router node can determine or otherwise obtain or receive information regarding) which transformation nodes are avai hthle within tile cluster and also which transformation types are supported on each translbrmation node, Load balancing can he accomplished automatically as the router nodes 112 direct content items for transformation

II

simply based on their current and target content formats. Nodes in the transformation node cluster 110 can also be reconfigurahie without human interaction. For example, a wait time or ohei mali c iiidic0nve of Sy'tc" otlit 1CflL Y Ocfl K irunitoted and if a tineshuld of U a niche is crossed, one or more of the lransform&ion nodes can be assigned a new defined transformation among the group of availaMe defined transformations.

[0028] A client machine can he configured to include a transfbrrnation chant library or other resource that can include addresses of one or more router nodes 112. Such router node addresses can be explicitly designated, or can he dynamically discoverable based or-an elastic search of available router nodes, virtual machines tagged in a cloud computing framework (eg.

Amazon Web Services available from Amazon, inc. of Seattle, WA), or the 111cc. A client library consistent with implementations of the current subject matter can ac as an application rogu. nming interface (API) to the nat sfor-nanun clus1e When called, tht chut I biaiy car cause the client machine 102 to pass the request, which can include a content item reference (eg.

idicating hich u nient item or items arc needed by the client irac inc 10Th and one ot more c' itena deu thing a ccired Lu get content foimrt thr the cotent em l (4 d hese cr'tcria ccn include one or more of a target MLMIS type, a number of pages. a target resolution. etc. The equest ako in ucla a allhack rout1nc whicli the mite" nouc 112 can exec dc up ii con-plato" of the t anstormatior to o ifv thL dLnt nach ic 102 0 the locauon of tie trarsioned content acm, ch un he stored in he (MS itpositoiy 106 Use of a callI'ack i OtJt(t e s ck snbed allows the client machine 02 to no' need to wait tor complctror o4' V e co itent fiji sturmanon Ratic, -it chont iachine 102 an pto css without comnic. on o' ftc content item transforrnanon. for example by continuing to render a page, a vIew, etc. within which the transfhrmed content item 104A will apnear. When the router node 112 executes the callback routine upon completion of the transfonnation or ftansforrnations rcqutred to produce the content item in the target format, the router node 112 passes a transformed content item reference to the client machine 102.. and the client machine inserts the transformed content item as necessary.

[0029] In the approach discussed above, and consistent with implementations of the current subject matter, the client machine 102 is not required to ever actually store the content item 104 or the transformed content item 104A. Rather. the client machine 102 only needs to have a link or other identifier or reference to the content kern 304 such that the router node 112 can command a transformation node 114 configured to perform at least part of the required transformation to access the content item 04 from the CN4S rcrosilory 106. to perform the transformation for which the transl'orniation node 114 is configured, and to store a transformed content item 104A. In some examples in which the content hem 104 can he converted to a target format in one transformation, the Lransthrmed content item I 04A can he referenced in the callback from the router node 112. to the client machine 102. in other examples, transtbrrning a content item 104 to a targetS *fcrtnat can require a chain of transformations, which can sequentially convert a content item 104 into one or more intermediate format transformed content items I 04A which arc in turn converted either to the target fonnat or to yet another intermediate format. The callback from the muter node 112 to the client machine can optionally inchide a reference only the transformed content item 104A having the target format, Optionally, additional m1o'mauoi cin he ?ncludcd in the callback, such as tot exampie a.cference to the onginal ontent tern (e untrarstonec an arriy of nionned content item etcrences e g to re&.cree all or at kast s me cf the ii terrree te lorniat tranCormed conte it acrrs stot ud in the CMS repository 106 a 1 surg of any options spec fled in the request, a ison o' oil cr listing of operations performed to convert the content item i0i to the target format and/or transthrmation node(s) invo'ved i-i the tranfortration o" than of tnns orrnaUo, etc 100301 In contrast to the approach described above with reference to FIG. 1, conventional approaches to content transformation generally resemble that illustrated in the diagram 200 of FIG. 2, A comuent transformation is more typically handled in a distributed manner using one or more load balancers 202. A load balancer 202 can assign transtbrmation operations among a group 204 of generic Drocessing nodes 206 according to a load balancing algorithm, The generic rnoecing nudes 200 ate ot preeonhgued to nerform a specdc Irrqfo nanon ipe, but arc instead configured to rcccive a content item reference and to perthnn whatever operations might be required to convert the content item 104 to the target format, [00311 FIG. 3 is a process flow diagram of a method 300 thr elastic scalahility for cloud-based content (runslornmtlons, At 302. a outer node of a transformation node cluster receives a request from a client machine fbr a content item stored in a repository associated with a server machine. The server machine is associated with the transfOrmation node cluster (which can optionally include being connected with the transformation node cluster by a communication network). The transfOrmation node cluster fbrthcr includes a number of transformation nodes, each of which are configured to execute at least one, hut not all, of a plurality of content iransfhrmation types. A number of each of the different transformation nodes can he based on an estimated load of the transformation type. For instance, ifa larger number of transformations are to convert the content type from a Microsofi() Word document to a Portable Document Format PDfl, then a corresponding percentage of the transformation nodes will be provided as Word docunient-to-PDF document transformation types.

[00321 At 104 the router node identities a current content format of the content item and.

a target content format of the content item, In some instances, the current content fbrmat and the target content format will be the same, in which instances the router node routes the content item or alternatively a content item reference (eg.. a link, address, or other location of the content item in the repository) dircefly to the resuesting client mach inc without any transformation, However, in other instances, the target content format differs from the current content format, md at iOó hasc ci on.he IL cut ying a tninfor na. on type tot the onten item de cm ned At 308 the router node routes a request to transthrm the content item to one or more of the number of tmansfbrmation nodes in the transformation node cluster that is configured to execute the content ¶ranstomnmuori type determined for the content item, In some examples, the content item can be identified by a content item reference, which can be part of the request to transthrm the content item. The transfbrmation node pcrfrms at least one operation on the content item required to convert the content item to the transformed content item. As discussed further below, transforming a. content item fbrm a current content hem format to a target item format requested by' a client machine can in sonic cases require two or more tran.sfonriations to occur in a transfhrrnation chain, The router node can be configured to track progress of the transformations occurring in such a transformation chain and, if necessary to reassemble multiple parts of a split or separated eonwnt item or intermediate content item to form the target content item, 0033] flu, 4 is a process flow diagram of a method 400 for scaling c.lou&based content triisfortnar otis ihese opeinti n' cn opona lv be pertorn'cd by onc ol mm; computing sytems hat ineludc' computer hardware bo" example, one oi more of the odes in tnt' nlurht of odes an execute Jc ollo\km& ope"atlors A / 0, an epcc ed I ad foi ea of a phir.U ty of transformation types is estimated thr content item requests from one or more client machines ic relating to content items maintained itt a content management system repository. Each transformation type of the plurality of transibrmation types transforms a first content format to a second content format diii ring from the first content format. For example, a transformation tyoc can include converston of a Microsoft WordTM fbrmatted content item to a portable Cot ument cu runt POF) mat on coin eisioi of a Pill eimat to a aer it PD] fornajtc a content items ench including "chunks" (e.g a subset of) the original number of pages in the PDF, coneNior f a PDF fo'-nar*ec content itt a a sepaiats. 2ortak nM"ork graphics (PNG) image fotmat for eac nage n the P01.or ci item, wnverion of a PN i fbi matted.oitc"t item to a joint photographic experts group (JPEG or JPG) formatted content item, conversion het\&ten highci cnd a lower ie'otuti tn mage torrnat in the JPG for natted content item, tc [0034] The estimating can include use of pred:ictive models, machine learning, neural networks. etc. to make predictions based on the historical data. The estimating can he performed dynamically and automatically, for example by a machine (eg. a system including at least one orogrammabie processor and having access to one or more data storage devices) that accesses stoTed data pertaining to content transformation requests requested from the transformation node cluster. This stored data can he retained, ontionaily wituin a content management system repository such as those discussed above. As noted above, the callback from the router node 112 to the client machine can include a variety ot information relating to the content request., how the transftirmation was performed, etc Additionally, human input can also be used. For example., a human input to the etirnatng in mcludc a rcquc-t op eontigon. dditional nodes tor a gi' en tin isforma: on Ppe hsed on Kno% let fC ihrt an upcoming protuct or pro ect, a iew sto net o' the like w 11 reqnc a arge ii rnhet of a &1'en we (o" mulupEe given tvpes ot trcnCcrmations to be performed.

10035] Consistent with implementations of the current subject matter, such infbrrnation ear be retained and used for ralurg lonaid lcokng predicuons of cxpcctcd use of nultipi: types of content transformations, inc stored. calihucks can include some or aH of previously requested content items, arrays of transformed content item references referencing intermediate content tems created in transformation chains to produce transfornied content items in response to the completed content item requests, listings of options specified in the content item requests, listings of operations performed to convert content items to target content formats, identifiers for transformation nodes that participated in the transformation chains, etc. [0036 Additional input data can include any or all of a type and number of business process applications supported by the content management system repository, usage factors defined for content tern transIbrmations expected to he called by the business process applications, an absolute or relative amount of mobile device usage supported by the content management system. repository, etc. The type and number of supported business applications can he predictive of the type of content transformations likely to be requested, For example, a contracts applications may frequentEy require generation of page images of a large document.

Knowledge of the types of business applications that may make content item requests and/or the type, characteristics, sizes, etc. of content items that may he called by such business applications can he useful in estimating potential usage. Changes in mobile device usage can, for example, relate to changes in a number of requests for content transformations relating to producing content items m tormats that ha' ura'kr tile sues o rcducc nandw dth usigc 100371 At 404, a transformation node cluster is configured, The transformation node duster mcludc s a plutahtv of noces. and me contigunne includes th sgt at'ng A( h or wo or more subsets of die plurality of nodes for executing one of a pluraEity of content transformation types. in other words, a first subset of the nodes in the plurality of nodes is designated throne of the plurality of content transformation types, a second subset of' the nodes in the plurality of nodes is designated for a second one of the plurality of content transformation types. etc. Each of the two or more subsets has a designated number of nodes of the plurality of nodes, and the number ci nodes in each subset is based on the estimated load for the one of the plurality of transthrmation types fOr which that subset is designated. the number of each transformation node ci be cv uarni.eally adjt sted haed on re's ted st rnatc, hitoneal in ormation,uc a rah + ic 1040 iriL'rn, rUor or uhe load analyt e or meast "ernent [O038 At 406, one or more router nodes are assigned with the plurality of nodes. The one or more router nodes are configured to perform operations that include receiving content item request. Fro the one or more client machines. identifying a current content format of a content item referenced ins received request of the content item requests and a target content tornuat of the content item specified in the received request, and determining one or more required transformation types ibr the content item from the plurality of content transformation types. Based on this infOrmation, a router node is also configured to route the content item to an apnropr Tc suhsct ot the two or more ubets An apprup"uate suh'et is one fir whic' tile content transformation type designated for that subset matches a required transformation type of the one or more required transformation types.

0039] As noted above, producing a content item having the target fOrmat may require more than one transtormatuotu Accordingy, the router node can he configured to determine a chain of transformations and to sequentially direct the execution of one or more intermediate transformation operations by routing the content item first to a cluster of the two or more clusters that includes nodes configured fOr a first content transformation type in the transformation chain and then routing a resuitant intermediate content item (or multiple intermediate content items if the first uansfoimron eoverts the eonten tern to minpie intermediate contert items e g b a phttwg ope'atio i or the ike to cnother d sate o, tie two or moVe dusters that rickides iodca configured for a second content transformation type in the transformation chain. Router nodes san also hc c nhgwcd o teute a ecrc ice tot th content item n the oct 1orm4 back o ire client machine that made the request. As noted above, the reference for the content item in the can take the ffirrn ofa calihack, which can include other items besides the reference to the content item in the target format.

[0040] At 408, a transformation type for the content item from the number of content transformation types is determined, and the router node routes the content item to one of' the number of transformation nodes in the transformation node duster. The translibrmatiori noce to which the content item is routed is configured to execute the content transformation type determined tbr the content item, The router node can he intelligent with regards to selection of a *rantoirnatron node to hch o route a 1⁄2ur,forrralsos request For example, in additicn o hoosmg a tiar's! rrnataor node based on tran'thrmatmn type, the router noe can also choose r least utilized (Fe. least busy) or at least a "less utilized" transformation node to which to route the vansfcrmauon request The selection or a less or least util ed tramfoimation node can be nased on reai nne load rnlorma on that is t[an.,li'j Red to the ro ter node fiorn eac o, the transformation nodes, 0O4 I] in a further implementation, a transformation node cluster can be capable of parallel transformation of a large content item An amount of time (or alternatively, a number of procesor ycles) iequiied to pe c'n' r given content item tarstorniation t mc is ge era iy related to cot ent item size. In other words, the larger the content item, the longer it generally will take to transform. The additional resources required to transform a larger content item can he provided in parallel (thereby increasing the speed: with which content items or transformed and optionally providing a more consistent time to transform content items that is less dependent on the size of the content item) by multiple transformation nodes in a transformation cluster similar to those discussed above, ía theilitate this paraliel processing, a transformation node cluster can include a content item splitting transformation type. Transformation nodes of this tyre t.a a or nt itt in into a numhtt 01 rnai or u ntcnt ito ns chi, nks) In operatu n a router node can route a request for transformation of a large content item first to a content item splitting node, which splits the content item into chunks, Then, the router node can route each chunk to a node configured for an appropriate transformation type, based on the target content format. As the transformation cluster can include multiple transfbrrnation nodes supporting any given transformation type, the chunks can be transformed in para! id. The router node can include logic or other intelligent processing routines that cause it to wait for the transformed chunks to become ready, and to assemble the transformed chunks into the final transformed content item, thereby ensuring the chunks are assembled in the correct order.

E9042] Other aspects, kml.ures, and operations of a router node consistent with implementations of the current subject rnattcr may he better understood in light of the additional detail ros ided heIei When t'anstorrmng vrntenl tern troni one for mat another f can useful to cencloer both ot n number oft-aristormatrn req ared to transform a content nem from the source format to the target format and a number of content items generated from a content item transformation.

O043] With regard to the number of transtormaflons required, in some cases a direct tra sfoi n carnot he race horn the o,nce ormat to th taIgc orrna1 Rather a chami of transformations may he required such that one or more mtermechate transformation.s 10 one oi more intermediate tbrmats occur in the process of transforming the content item from the source format to the target format. As an example, a transfbrmation of a content item from a DOCX format (e.g. a Microsoft WordiM doeunerit forniaO to a PNQ (e.g. a portable network gruphics image format), may occur by first transforming the content item from the D(i)CX source format to an intermediate format such as PDF (e.g. a portabe document format), foflowed by transfoiming from the POE' intermedirte format to the PNC *arget format Chains ircluidmg interrnca ate tot nat t axbformatlons can tvptca > rdude two to fou ç i optionally mtn) transformations, When a single transform is possible, this can be referred to as a direct transformation. If a chain of transformations is required, this can he referred to as an indirect transform eLi on.

[0044 \1⁄2 h rebard to the nu ibe u codet tenv genetated i'm a tiancformatior ii can he possible that a single source content item may need to he transtbrmed into marty target content items, For example, the transformation of a multiple page DOCX file to a PNG target format can involve generating multiple PNG target content items with PNG file representing a single page.

100451 To further complicate matters. within a transformation ciiain (e.g. in an indirect transformation), any number of the transform steps may need to produce multiple Conteni items.

As an example, a typical transtormatjon can include generation of thumbnail or preview images (one per page of the content item) from a content item having a source content of DOCX. in this case, the transformation chain for a multiple page DOCX source file could include a first transformation from DOCX to POF and a second transformation from PDF to multiple PNG files ($e per pagq) and thi third t$nsYqnnati*n btSd. the si$ óieach PWitI1e4o,a thumbnail (eg as a transformation from a PHG intermediate f*A!mat to a IN) target toxznat) 100461 Therefore, foktDOCX ito haIsgThrsjpges, th msfaa&tt 4hin 5; of FICt 5 <can occiit The &st transfbrnxafion 502 gonltmju4e tran.stbrming the 1)00 to a fl)J nttnmatc fárn$icftowod bya truiámauotsö4 ofthtPiP fileta t&ePNGffles xeprendbg, the, pages at the source DOCX file. A thfrd trsnstormation fln&forniingach efihedwee PNGs intermSiate t1es to the target format WCfltesa (00471 if the source content Item happened to have a Iarg number Of pages, (be ii t:auon èlîàìñ 5043 d fth: mcloe an additional thtediàte fra foithMien$oi to the PDF p1it the DOCX t muItiplt tt Ahethth'dy the U tthatiOn tiode nflgted to ë4ftt DCCX to PDF;fl be bUt fiUtd VbUtpithPDF I 1 &Ii to iii $Ui:ShtflUñibo) page chunks4 in tlt&ló oH 300 páj DOCX UEtEoht Item; the niatron chant *afi Thelud a ttsflt ibfltatftin to th$!= th4?4t PD? filSr: fOllow by: POE ifln PNQ flWtta fthtioii fbi Soft: of the PDF öhik éifdThä PSOYto JPG t'ransfonnatlons for each ofme:&OOtesultingPNG interrned1atet1Iet iapør possi pa i DX to PEW inmtaitffien Ic pqcç . $!QW! bL7!ck. hut k. : &41xti e:e!!Si1 c: to aPUt4I9fPPF ip chw:S 1tK4i the transfSiations in a trazstotmatk chahi can be rtfontd It pallt whe*enr pøssib1i For eample1 as soon sss source coatendtcmtär snytrans onnalion in the transkrmation: chain becomes available, a mquest to tmns form it is: made to iris the nun*er of parallel tçanst& is addkinat transfØmatIón nodes can be*444 t the du. tie par&lçl nature traSktms means thcr is no guarantcc on tim orkkr that they compIete 10050] A router node consistent with the descriptions herein can handle all of the above transformation chaining, and can take into account that a transformation can result in multipk contcnt items, and Ll'a sbsequc t transorwali ns in the trmlorma n hair an tesalt in creation of a larger number of intermediate or target content items out, and that subseciucnt s('nsiorm(tion slen' carnot tait i ntii ner predeeoi trans4aunat Ons ha-r take' place Such a router can manage these demands in some examples using asynchronous invocation of multiple ransfoimauon tquests fui a g cr contcxt as wcl' a. mdeved o'flrn. retciences [0051] Asynchrono s invocation or muit pie transtorn atioi eque'b Fm a g ver contet (in other words flu a given source content item) can occur with a nonbloeking wait flu completion of all transformation requests. At run4irne, when a nonbioeking wait for multiple transformation requests is required, the router can create a transient handier that maintains the state required fin sending the requests, waiting for their response, and handling errors (e.g. a response may never come hack). Each request can he tagged with a handler context and a request identifier, so that when a response is received (eg. containing the context and request identifier) it can bc mapped bdck to the requcat hat inhated t arid to the I andier hat is managing it. The handier then keeps a record of the responses received. When all responses are eceived,jr whaftc'-ordei hey.oinc bak the handler notries the torcr that these transloinations are ro done and t at the t uter Li no\ e to flu next I ansi ma wn in the transformation chain, if there is one. If no further transformations are required to complete the transformation chain, the router responds to the client.

10052] Indexed content references can assist in determining which segmcnt of a content item is being processed. A segment can refer to a page number or group of page numbers of a source content item. hut it could also represent paragraphs, time within video, or other r,arametcrs, An indexed content reference can indude both 1l'e location of the content item ann the index. A request to transform content includes the indexed content reference of the content item to transform. In a simple case where, fOr example, a DOCX file is transformed to PDF, a transformation request would he made including a cornent reference such as "itentdocx index I", and the response would include a content reference such as "item.pdf -index 1" This indicates to the router to process the. item.docx content item from page 1 (i.e. the beginning of the JOCXt In a rnore eonplcc case \J'ere a argo DOCA tie is split into multiple DI s c g a 00 ii mge IIO X phi into Ic) hu:ks) a ttanstormati 1 eque\t wot Id id..ide t'c r dexed content reference "iteni.doc index 1" as before, but instead of returning a single indexed content reference, multiple indexed content references can be returned for each PDF chunk (e.g. "item,pdf index 1". "IueTn.pdt index 30'', "iteni.pdf index 60" and so on fOr all 10 chunks, I his treat a that when tE ost re,inieu indcxec. conlcnt rctercnccs arc ued i jhscuuc it transformation steps in the chain, the transfOrmer node understands which part of' th.e org1naI content item it is working with. For example, the next transfOrmation in a transformation chain may he to transform each PDF chunk into a series ot'PNGs (one fOr each page). In this ease, one of the subsequent requests will include the indexed content reference "itentpdf -index 30", The transform type (PDF to PNG) will itself return an indexed content reference or indexed content references that is/are relative to its source indexed content reference. For example, in this ease, the returned indexed content reference would be "item.png index 30", "itcm.png index 31", "itempng -. index 32" and so on for the 10 pages of that PDF chunk, With indexed content references, the handler that is waiting for all the responses to be returned (and can be returned in any order) can now sort those quickly by content reference index to ensure that the transformed content items can he rcturned in the appropriate order to the client.

[0053] One or more aspects or features of the sulect matter described herein can he m cgital electronk c' cuhi, vrerated Lncwtiv, speciJl dcigned app! aion specific wtegr0tcd U ewh tASICs), field prgra mrah e gate arrays (1 kiAs) comp ster haidwa e, firmware, software, and/or combinations thereof. The-se various aspects or features can include npkmentat on n one o wore oir xttei ungi ams tha at exeu aHe and/c" interpretable on a programmable system including at least one programmable processor, which can he special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device, The programmable system or computing system may include client-s and servers, A client and -server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

-[0054] These computer programs, which can also he referred to programs, software, software applications, applications, con!ponents, or code, include machine instructions for a progianiniab C pro(ccsot, nd can Ct n-rplementcd in a lug!' le-el procedural aguge an object-oriented programming language, a functional programming language, a logical programming language, and/or in assembly/machine language. As used herein, the term "machine-readable medium" refers to any computer program product, apparams and/or device, such as for example magnetic discs, optical disks, memory, and Programmable Logic Devices (PLDs), used to provide machine instructions and/or data to a programmable processor, including a machine-readable-medium that receives machine instructions as a machine--readable signal. The tern "machine-readable signal" refers to any signal used to provide machine instructions and/or daLa to a programmable processor. The machine-readable medium can store qch!Th:fl flØfl4j! flsft*tdy suób ai ftrecampJe as would a nOfrtOøl., solid' eww cc.ww......Sha$ 4'ive or any eqgva ssc:!ac4k,. Th. SthwtJe suth jj fltions iii a fl*,* manner, MCh1s for:, lSwei*Ma proces,.. ae r ether nw4,wacccs mw. aoçiØ4 wjffi one ormore phyiibatprolessor &toEes, lOOSI to prnvide ter interac$on with a user one or more aspects or Eeatures d the aübJ matter desciThed hetein *an be4mØmentedwtcouipiner having a 4ispl&y device, such as *3r example a cattwde ray tube (titt) or a fluictorystal display (La)) or a 110 emitting (LED) monItor lbr &1ai,g 1 rmation to the user and a lèybqard and a ppintitig Øç,, suèh as fér exam p1* a niouse or a tzadkbali, by stch the usçr nitty piwide input to the (:ilei kinds of;de+Jces;canvbeusS to provIde r Mteraetlon Sa user as well. For ø1 *4back:pcS$4 t$e urçn b*anyfonofseñty*4$ack such as for enr vimJ fàtk Sitory fi or tac,tjlà fSbackjand hiput from the?user maybe rcccWed am' tc!!, t4pg b S llnilW acou$Mk sp* qr tSPtile h Other ! input devices I ck1e, but are not lint to, touch screens or other touch-seSitive 4cøs such as qi muitMpoint fl'*iyQ Or apacMve ta4pds voice eition hardware and, c c,,,,,,,,em op*$ pS,4jt1 I,ne cwe 4c*s S. jjf4 inteipretation sr andthe like.

(øJ fr the deSptlons abownnd in die alms3 hrases such as it ii. nnc oP' or "one or more 0?' may. pgcur foil owed by a conjunctive ilst of elemexns or features. The term may also pçcur in;t lilt of w or mor clcmexits or featurn. *Icss MherMe iirlhatt)' or explicidy wntradioted by tiw context in hiäh itused,, sueh aphrasc is intended to mean any f The ltst elements or fawns lndWkluallyor any of the Si elements ortSww 26: iniosmbSon with: *nP of the other reElS enS or tøtures. :Fer:exaniple, the pbres "at çqgatf4bpç prçç$A U!. ;"& A and/ur$P!fl to A: ox k a..4 B gcect' 4 hMW rp$ip is also frç$$ including threeqr t:i: For q k, the pfrn "at least oncyot& B, and C:;" "ei ar more of A3 B and or and "A 13, andt;V" are each intended to mean "A alone, 13 aióee; c alon; A ifid B tOgitet, Aand C together, B a$ C tg$%er3 or A: antE and CtogetheC ke ottholnn "bted sç' dbavo aiid in the cléhus is iüteSttio h*th, "buèd 4 Idást in $rt:Ort,' thât:i: (ffik t&l ftMIJS:ôi tha1S&ttiMthb!e v:eit The *Jk mat descri herein can bç en*odiea hi systems4, appantus gc dc*ed configurátian the ifrplomentat:ionwset fi,rth in the foregoing, 4S$iS do S represent all implernenfi tons; cuniistint with the suJèct a 4SflcS In$ thc meitly some ps*w* asncth rdid to the 4escth su aattcr.: Mthngh a few vations hye been described t detail aho other mJfIb or pcW Th f fl: w*, varlitions can be provided in addi ion t those set forth herEin. For e*ampl; e itnp1ons::4ec: tve?ean be *ect$ to tS con Sk and sub lnatk of tb distód fratnes and/or combln bits an4 $ Øt:*S Mthr (w! 4J* Th additiOn, the lbgi flows: dpiot. in ti:accompan$ng ügures4tor 4escribed hes do neosnj:qnio the Øafticuler order shown, or: sequentiat order; to achieve desirle:ct., OthetthithtkUfl3y tIilu th:g di' thtiblldtThg ólüth,

Claims (8)

1. CLAIMSA computerirnplemented method comprising: receiving, by a router node ofa transtormation node cluster comprising a plurality of nodes, a request from a client machine for a content item stored in a repository associated with a server machine that is associated with the transformation node cluster, the translbrmauon node cluster further comprising a plurality of transfbrmation nodes, each of the plurality of transformation nodes being conflgured to execute at least one, hut not all, of a plurality of content transfhrmation types; identify.ng, by the router node, a current content-format of the content item and a target content fomiat of the content item, the target content thrniat differing from the current content lormat; determining, based on tue identifying. one or more transforniatton types from the p1tnaht of content zransfonnatio types for transform ng the onicnt item to theta et co tnt format; routing, by the router node, the content item to one of the plurality of transformation nodes in the transformaiion node cluster, the one of the plurality of transformation nodes being configured to execute one of the one or more content transformation types determined for the coil tent item.
2. 2, The coniputer4rnplemented method in accordance with claim I, wherein the p' irality of tiansformat on types eomnnses a trrsformauoi typ that rnchds ransfoimrng u e content item from the current content format comprising a portable document format into the target content format, which comprises the content item split into two or more sec-tion in the portable document format.
3. The COnlpUtT-niplemcnttd method ir acoruac' wn clani I or c14 ii 2 wherein the plurality of transfbrrnation types includes transforming the content item from the current content format in an editable text document thrmat into a target content ibrmat of an image tormat.
4. 4, I he computer-implernentee method in accordance with any preceding ciami, wherein a number of the plurality of transformation nodes in the transformation node cluster configurea to execute erch of the plurality of content transformation types is based on an estimate of transformation loads of difibrent transformation types to be hancfled by the transformation node cluster.
5. 5. The computer-implemented method in accordance with any preceding claim, thrther comprising: estimating a load for each transformation type; and configuring a number of the pinraifty of transformation nodes in the transformation node cluster configured to execute each of the plurality of content transfbrmation types based on El-ic estimated load thr each transformation type.
6. 6, The computer-implemented method in accordance with aily preceding claim, wherein the one or more transibrination types determined by the router comprise a transfbrnanon chain comprising a plurality of transformaU-on types, and whercm the one of the plurality of transformatton nodes in the transformation node cluster has a first transformation type totrattsthrm th con m to lint it, eotent it $avIflga first tof"idt and wh' in?the method füEther cpmp4; iagth tinuintermediate content item to aseccnd node ofthj$uramy of u'gsfmatk" nodes in the transtormationnodtctutev;havingasecond;ransformation ypç to t*'foi'nnhe fks't±tted iátó coiiteántëS to eithót as'n?id thtethi ediatecontenfitem having aaecmid Intennediate lbrimivor;;th a targot cofltentitcth S'ri"g;theta",' 16th".
7. 7. The compstt&h"leniented i'd in ae"'''e with claith 6, whtrein the tthflsfbflflaSfl chaib "in$" a ti"tnn"lin typt'toswthi"g 1" Thth"ation " fl"üItjpIe hncrit'edIate content itenis ana c"' of the multi jño inteE"odi'" oontstt i"i "flS a sSenttansSS, L The t"mp"taplemc"tcd t"Thod fl"t"tdM" with d"lth 7. Wh"iii the method M'ronwS& røuth*g, by the tul::1"frrt"tedi" content sS"muwptn"tb"loi nedsth the transbntation node ciusterthemuttipIi tEansfan' on' notseath being figured to execute a same one *Da piwaiItyofiontdittanS'on"aS' types sut' that U* Lnulffitefntenne&ite content frents an"pees'" in trai"1., 9. The eotnputSimp1ettente4 meik" fri accordance with any precedIttg: thirn wherein each of S,pinrathyMiansfârntation nodes nStgurS tA execute S' one of the; p1urafh JpS,t transfàmation' types.10. The computer-pemented metlwd " aqçor4ançç whi "any preceding cft'in providing a ç, Iback4 by The rQuter node to th: cliont machine, the callback.ccrnpthing atef' neejo &anstàrnted content item havin the target; content ibtinat.ii. The cornputer-irnplernented method in accordance with claim ii, wherein the callback further comnrises at least one of a reference to the content item requested by the client mac uric n arm of transtormed cor tent item reerences refeiencmg ore or more riterrrcdiate content items created in a transformation chain to produce the transfonned content item, a listing ol options speulka in toe reques. a Ins on' o operanon pci ormed to corvert the conten Ic n to the target content format, and identifiers for one or more transformation nodes that participa ed in tne ranstorratk'n C iclU' 12. The computer-implemented method of any preceding claim, further comprising: creating, h the router, a transient handler, the transient handler managing the request; tagging the request with a handier context and a request identifier; mapping a response received from the one of the plurality of transformation nodes and containing the handler context and request identifier back to the request and to the transient hand].er, the transier..t handier keeping a record of the response and notifying the router when all transformations of a first transformation type of the one orniore transformation types has been completed; and routing, by the router, intermediate content items rcsultinn from the first transformation type to one or more additional nodes of the plurality of transtbrmation nodes in the transformation node cluster, the one or more additional nodes of the plurality oftransfbrmation nodes being configured to transform the intermediate content items to one or more next intermediate content items or to the target content format of the content item.13. A computer program product comprising a machine-readable medium storing instructions that, when executed by at least one programmable processor, cause the at least one programmable processor to perform operations comprising: ieciM,& by a rtuS dettrwt *rSfton no:,, chu*er,,,p,,,sp!w,'*' of nodes, a request *um a client machme fur a content item stored in a npositury associated with a server citt' w,,,,c, d!,,4 Ssfirmatien node clu$er further cpnipthjflg ptucaflty fla tgafio nd,øEe,,,,, n....iranslónnation nodes being configsSItcexecute at leastoneQ but not a!!, ala plurality of cohtent transfotmat3on types dntifft'g,flihe "Sñöde, aéunertt conteñVfoEt'M öfihó'Softttoü áñä a target htsnttssof'll* &ffi" IteM; thelt"n t" ft" difft'ti4 fr"t'th t":":"tht "ñtci" bEll: d"infrig has' On th'idstifithg, one dttñOtó th"ioi " "iOn t" froni the' pt":"!i tcol"nt Jj' fa"ttf6tht't'the;t"tenth"n:tI*cI"goH"ii"it': rDuting,by the mutentede the eoment Item tø one obthe plurality of tsftrtnatkm in a"s'rat"n''4std: theoite of the pknhit' Mathf;thai"sdes'b"thg.configuredto"execmeoneaf)tbe"øne;ot more;content trenaft' rnatlon'types determined fir the øopienthem 1k. Tcmiputcr program pd"t b;'""w1th dab" t3 WhtSiu aitui"therOf the iiSj. tt iSes: the transfonristn node cluster cothgurS to txettte ov the plurality of content transformation p is based on n Simatt of transibniiatión loads *4iftSnt sbmation type wbe arnie by *e*raSwmatton iSectàer the computer prgram pSu&in accøttm,j with claihi 1'or claim i4 weSh the operatiois' flztther'compthe: psthnathg'a loS tor'each transfoni ation typc; and n configuring a number of the olurality of transibrmation noces in the transformation node cluster configured to execute eac.h of the plurality of content transtbrmation types based on the estimated load for each transformation type.16. The computer program product in accordance with any of claims 13 to 15.wherein the one or more transformation. types determined by the router comprise a transformation chain comprising a plurality of transformation types, and wherein the one of the.plurality of transformation nodes in the transformation node cluster has a first transformation type to transform the content item to a first intermediate content item having a first intermediate format, and wherein the operations thrther comprise: routing the first intermediate content item to a second node of the plurality of transformation nodes in the transformation node cluster having a second transformation type to transform the first intermediate content item to either a second intermediate content item having a. second intermediate format or to a target content item having the target tbrmat.17. The computer program product in accordance. with cLaim 16, wherein the transformation chain comprises a iransthrmation type resulting in formation of multiple intennediate content items, and each of the multIple intermediate content items requires a subsequent transformation, and wherein the operations further comprise: routing, by the router, the multiple intermediate content items to multiple transfbr.mation nodes in the transformation node cluster, the multiple transformation nodes each being configured to execute a same one of a plurality of content transformation types such that the multiple, intermediate content items are processed in parallel.18. The computer program product in accordance with any of claims 13 to 1 7, wherein the operations fUrther comprise: creating, by the router, a transient handler, the transient handler managing the request; tagging the request with a handier context and a request identifier; mapping a response received from the one ofthe plurality of transformation nodes and containing the handler context and request identifier back to the request and to the transient andler the ttansient handler keep ng a icooro of the rcor,e and otity ng te routr when all transformations of a first transformation type of the one or more transformation types has been completed; and routing, by the router, intermediate content items resulting from the first transformation type to onto' nore acditio'ia iods of Le rluralit) of tEanNfol -n'on no les in U-n.transformation node cluster, the one or more additional nodes of the plurality of transformation nodes being configured to transform the intermediate content items to one or more next Intermediate content items or to the target content fhrmai oLthe content item, 19. A system comprising: computer hardware configured to perform operations comprising: receiving, by a router node of a transformation node cluster comprising a plurality of nodes, a request from a client machine for a content item stored in a repository associated with a server rnacie-that is associated with the transformation node cluster, the transformation node cluster further comprising a plurality of transfhrrnation nodes, each of the plurality of transformation nodes being configured to execute at least one, but not aH, of a plurality of content transformation types; identifying, by the router rode, a current content format of the content item and a target content format of the content itcm, the target content format differing from the current content format; determining, based on the identifying, one or more transformation types from the ph.Fraiity of content transfinnauon types for transforming the content item to the target formal; routing, by the router node, the content item to one of the plurality oftranstbrmation nodes in the transformation node cluster, the one of the plurality of transformation nodes being configured to execute one of the one or more content transformation types determined for the content item.20. A system as in claim T 9, wherein the computer hardware cornpnses a pmgrammaie processor; and a machinereadable medium storing instructions that, when executed by the processor, cause the programmable processor to perform at least some of the operations.21, A computer-implemented method for scaling cloud-based content transformations, the method comprising: estimating, by at least one computing system comprising computer hardware based on transformation usage data, an expected load for each of a plurality of transfommtion types for content item requests from one or more client machines relating to content items maintained in a content management system repository, each transfonnation type of the plurality of transformation types transforming a first content format to a second content format difforing from the first content format; configuring, by the at least one computing system, a transformation node cluster comprising a plurality of nodes, the configuring comprising designating each of two or more subsets of the plurality of nodes for executing one rita plurality of content transformation types.each of the two or more subsets having a designated number of nodes of the plurality of nodes, tue number of nodes based on the estimated load: for the one of the plurality of transtorrnation types for which that subset is designated; and assigning, by the at least one eomputng system. one or more router nodes within the plurality of nodes, the one or more router nodes being configured to receive a content item request from the one or more client machines. and to roule the content item request to an appropriate subset of the two or more subsets, the content transformation fbr which. the appropriate subset is designated matching a required transformation typo of the one or more required transformaion types.22, [he eornputerinipIernented method in accordance with claim 21, further comprising reconfiguring the transformation node cluster, the re-configuring comprising changing the designated number of nodes of the plurality of nodes for at least one of the two or more subsets based on a changed estimate of the expected load.23. The computer-implemented method in accordance with claim 21 or claim 22, wherein the one or more router nodes are further configured to: identifS' a current content format of a content item referenced in a received request of the content item requests and a target content thrmat of the content item specified in the received request; and determine one or more required transformation types for the cotflent hem from the plurality of content transformation types.24. The computerimpiemented method in accordance with claim 23, wherein the one or more required transformation types comprise a transformation chain comprising a plurality of u&sforrnatwn lypts and whLeui Ilk appropt ate subset ha a fir',t translormatma type to transform the. content item ton first intermediate content item having a first intemwdiatc format.25. The computer-implemented method in accordance with claim 24, wherein the one or more rouier nodes are further configured to: route the first intermediate content item to a second appropriate subset, a second content transformation for which the second appropriate subset is designated matching a second required transformation type of the one or more required &ansthrmation types for transforming the first intermediate content item to either a second intermediate content item having a second intermediate format or to a target content item having the target ftrrnaL 26. The computer-implemented method in accordance with any of claims 21 to 25, wherein at least one of the one or more router nodes is also included in one of the two or more subsets of the purality of nodes.27. The computer-imlemnented method in. accordance with any of claims 21 to 26, wherein the transformation usage data comprises historical data pertaining to contcot transformation requests requested from the transformation node cluster.28. The computer-implemented method in accordance with claim 27, wherein the ebnutmg turtl'er compnss ue o at least one of a prejtctive modeS, m&hmc leaning and a neural network to make nredietions based on the historical data.29. The computer-implemented method in accordance with claim 27 or claim 28, wherein the stored data comprise stored cailbacks provided by the one or more router nodes to client machines in response to previously completed content item requests.30, The computer-impemented method in accordance with claim 29. wherein the stored cal Ibacks coiriprise at least one of references to requested content items, arrays of transftrn-ied content item references referencing intermediate content created in transformation chains to produce transformed content items in response to the completed content item requests, listings of options specified in the content item requests, listings of operations performed to convert content items to target content formats, and identifiers for tral-!siormatjork nodes that participated in the transformation chains.31. The computer-implemented method in accordance with any of claims 21 to 30, wherein the transformation usage data further comprises at least one of a type and number of husrnec proces applications supeorted by tl'e co ent ma ageinent system cep site!) usage factors defined fur content item transformations expected to he called by the business process applications, and an absolute or relative amount of mobile device usage supported by the content management system repos tory.32. The computer-implemented method in accordance with any of claims 21 ui 31, whenin the estimating further comprises acting on at least one human input parameter relating to expected transformation demands.33. A computer program product comprising a machine-readable medium storing instructions that. when executed by at least one programmable processor, cause the at least one programmable processor to perform operations comprising; estimating, based on transformation usage data, an expected load for each of a plurality of rr'nforrnsuo pcs fo comert thm requests flow or irore client ra llnN relating to cur ter 1. items warn -iined.n a.o dent mana°cmenl cytem repo'itcr' each tranloi it aton tvpc of the plurality of transformation types transforming a first content format to a second content format differing from the flrst content format.; configuring, a transformation node cluster comprising a plurality of nodes, the contigunng comprising designating each of two or more subsets of the plurality of nodes for executing one of a plurality of content transformation types, each of the two or more subsets having a designated number of nodes of the plurality of nodes, the number of nodes based on the estimated load for the one of the plurality of transformation types for which that subset is designated and assigning, one or more router nodes within the plurality of nodes, the one or mere router nodes being configured to receive a content item request from the one or more client machines, and to route the content item request to an appropriate subset of the two or more subsets, the content transformation for which the appropriate subset is designated matching a required transfOrmation type of the one or more required transformation types.34. The computer program product in accordance with claim 33, wherein the ooera1ons flrtbu co npnse ic-configuring the translonna on node Juster, the re-conl'g "ing comprising changing the designated number of nodes of the plurality of nodes for at least one of the two or more subsets based on a changed estimate of the expected ioad.35. The computer program product in accordance with claim 33 or claim 34, wherein the transformation usage data comprises historical data pertaining to content transformation requests requested from the transformation node cluster, 36. The computer program product in accordance with claim 35, wherein the stored datr carnpise stowd cailbacks pio ided b rhe 0 K OL more ro itu nodes to c cut nak hrnes vi respor\o to prtviously complettu cnLent urn ru1ucsts 37 Tt'e computer progrrni produt in a&o d1nL \kilh claim 6, whereat Me ctond calibacks comprise at least oneS of references to requested content items, arrays of transformed content teir etcencc% ielbrncmg intenvedare content items eieated in traisfotmaton c1iain to produce transformed content items in response to the completed content item requests, listings of options specified in the content item requests, listings of operations pertbrmed to convert content items to target content thrmats, and identifiers fix transformation nodes that participated in the ransthrmation chains.38. The computer program product in accordance with any of claims 33 to 37, wherein the transfbrrnation usage dat1i further comprlses at least one of a type and number of jusmess p?ocess a jlicut'ons s upporkd I y (he content niaraganent systeri "epository usage fttctors defined fin content item transtorrnatons exoeetcd to he called by the business process applications, and an absolute or relative amount of inobik device usage supported by the content management system repository.39. A system comprising: estimating, based on transformation usage data, an expected load for each of a plurality of Lransfrmaticm types lbr content item requests from one or more client. machines relating to conknt vein mairtaned n a content man( gement sytern repusito each tr.1n'JcMm flon t_pe of te p malay o't'ar sfomation t'pes Uars'orm -g a firs eortert foirnat to a second content format differing from the tim content format confiudng, atransfrmitiun okiS ream gap:lurality eno4e; the nwjigçprkiig deèjgtèt1éig ea*oflw. mo*sjc4 otthejtijr4fty ofSks fbv executhgone dfa plurality qtçppç tsftrmn frpç ., ç oçJe$'oqç haviñga4eSignat:öd punib., fm4, efTheJ*4calitflf nó4es, the twa, otn4s,,wkn the 1àe4 fpc Øe ce qfb ti' asignisg, qpe orqr rQu$r noØçs, wt.b th. plum,lity of nodes, the one or more ru nodes bóingconfigured torecèive a content item request from thétone or thóStlièñt machines, and to roUte the4oAtentheMtequett an apprnptiatd sUbset of the two tmonnub: , the obAtóñt frañnnitiUtt foi*hkh the 4$t$fà sübsóti ?désiëdthEtóhkAflEtiS thffltS*hitThñ ty$ Of Th oio orMoo*quIYod tias%thaThn;typ.4t A tth as in E fin 39, *hin,tcoi4utE,;hEd*aotthØiss ad ttiheSadäbIethedhin st*i flk tiEñ,tbàt3 cause the programmable prtcessttterfotm:at last srne of the operaftans Amendments to the claims have been made as follows:CLAIMS1. A system comprising: a content management system repositoiy configured to maintain content items; and a transformation node cluster comprising a plurality of router nodes and being configurable to have each of two or more subsets of the plurality of nodes designated for executing one of a plurality of content transformation types, each of the two or more subsets having a designated number of nodes of the plurality of nodes, wherein one or more assigned nodes within the plurality of nodes are configured to: receive, from one or more client machines, requests relating to maintained content items; li') estimate, based on transformation usage data, an expected load for each of a plurality of transformation types for a content item request, each transformation type of the plurality of transformation types transforming a first content format to a second content format U') 0 differing from the first content format; and route the content item request to an approptiate subset of the two or more subsets, the content transformation for which the appropriate subset is designated matching a required transformation type of the one or more required transformation types, wherein the number of designated nodes of each of the two or more subsets is based on the estimated load for the one of the plurality of transformation types for which that subset is designated.2. The system in accordance with claim I, wherein the transformation node cluster is re-configurable to change the designated number of nodes of the plurality of nodes for at least one of the two or more subsets based on a changed estimate of the expected load.3, The system in accordance with claim or claim 2, wherein the assigned one or more router nodes are further configured to: identif' a curent content format of a content item referenced in a received request of the content item requests and a target content format of the content item specified in the received request; and determine one or more required transformation types for the content item from the plurality of content transformation types. IC)4. The system in accordance with claim 3, wherein the assigned one or more r router nodes are frirther configured to route the content item to one of the plurality of transformation nodes in the transformation node cluster, the one of the plurality of transformation nodes being configured to execute one of the one or more content transformation types determined for the content item.5. The system in accordance with claim 4, wherein the one or more required transformation types comprise a transformation chain comprising a plurality of transformation types, and wherein the appropriate subset has a first transformation type to transform the content item to a first intermediate content item having a first intermediate format.6. The system in accordance with claim 5, wherein the one or more assigned router nodes are further configured to: route the first intermediate content item to a second appropriate subset, a second content transformation for which the second appropriate subset is designated matching a second required transformation type of the one or more required transformation types for transforming the first intermediate content item to either a second intermediate content item having a second intermediate format or to a target content item having the target format, 7. The system in accordance with claim 5 or claim 6, wherein the transformation chain comprises a transformation type resulting in formation of multiple intermediate content items, and each of the multiple intermediate content items requires a subsequent transformation.IC) 8. The system in accordance with claim 7, wherein the one or more assigned router nodes are further configured to route the multiple intermediate content items to multiple r transformation nodes in the transformation node duster, the multiple transformation nodes each being configured to execute a same one of a plurality of content transformation types such that the multiple intermediate content items are processed in parallel.9. The system in accordance with any preceding claim, wherein at least one of the one or more assigned router nodes is also included in one of the two or more subsets of the plurality of nodes, 10. The system in accordance with any preceding daim, wherein each of the plurality of transformation nodes is configured to execute only one of the plurality of content transformation types.H, The system in accordance with any of claims 8 to 10, wherein at least one of the one or more assigned router nodes is further arranged to: create a transient handler arranged to manage the request; tag the request with a handler context and a request identifier; map a response received from the one of the plurality of transformation nodes and containing the handler context and request identifier back to the request and to the transient handler; and route intermediate content items resulting from the first transformation type to one or more additional nodes of the plurality of transformation nodes in the transformation node cluster, the one or more additional nodes of the plurality of transformation nodes being U) configured to transform the intermediate content items to one or more next intermediate content items or to the target content format of the content item, r and wherein the transient handler is configured to keep a record of the response and LI) notify the router when all transformations of a first transformation type of the one or more transformation types have been completed.2. The system in accordance with any preceding claim, wherein the transformation usage data comprises historical data pertaining to content transformation requests requested from client machines, 13. The system in accordance with claim 12, wherein the historical data comprise stored callbacks provided by the one or more router nodes to client machines in response to previously completed content item requests.N, The system in accordance with claim 13, wherein the stored callbacks comprise at least one of references to requested content items, arrays of transformed content item references referencing intermediate content items created in transformation chains to produce transformed content items in response to the completed content item requests, listings of options specified in the content item requests, listings of operations performed to convert content items to target content formats, and identifiers for transformation nodes that participated in the transformation chains.15. The system in accordance with any preceding claim, wherein the transformation usage data further comprises at least one of a type and number of business process applications supported by the content management system repository, usage factors LI') defined for content item transformations expected to be called by the business process applications, and an absolute or relative amount of mobile device usage supported by the r content management system repository. L()6. The system in accordance with any preceding claim, wherein the one or more assigned nodes are further configured to estimate using at least one of a predictive model, machine learning, and a neural network to make predictions based on the historical data.17. The system in accordance with any proceeding claim, wherein the one or more assigned nodes are further configured to estimate by acting on at least one human input parameter relating to expected transformation demands.
8. 8. The system in accordance with any preceding claim, wherein the content management system repository and transformation node cluster are implemented by one or more programmable processors; and one or more machine-readable mediums storing instructions that, when executed by the one or more processors, cause the programmable processors to carry out at least some of the instructions.

   9, The system in accordance with any preceding claim, wherein the router nodes of the appropriate subset are configured to transform the content item from a source format to a destination format and provide the transformed content item to a client machine that requested the item.20. A computer-implemented method for scaling cloud-based content transformations, the method comprising: LI') estimating, by at least one computing system comprising computer hardware, based on transformation usage data, an expected load for each of a plurality of transformation types for r content item requests from one or more client machines relating to content items maintained in a content management system repository, each transformation type of the plurality of transformation types transforming a first content format to a second content format differing from the first content format; configuring, by the at least one computing system, a transformation node cluster comprising a plurality of nodes, the configuring comprising designating each of two or more subsets of the plurality of nodes for executing one of a plurality of content transformation types, each of the two or more subsets having a designated number of nodes of the plurality of nodes, the number of nodes based on the estimated load for the one of the plurality of transformation types for which that subset is designated; and assigning, by the at least one computing system, one or more router nodes within the plurality of nodes, the one or more router nodes being configured to receive a content item request from the one or more client machines, and to route the content item request to an appropriate subset of the two or more subsets, the content transformation for which the appropriate subset is designated matching a required transformation type of the one or more required transformation types.21. The computer-implemented method in accordance with claim 20, further comprising re-configuring the transformation node cluster, the re-configuring comprising changing the designated number of nodes of the plurality of nodes for at least one of the two or more subsets based on a changed estimate of the expected load.22. The computer-implemented method in accordance with claim 20 or claim 21, IC) wherein the one or more router nodes are further configured to: identify a current content format of a content item referenced in a received request of r the content item requests and a target content format of the content item specified in the received request; and detemine one or more required transformation types for the content item from the plurality of content transformation types.23. The computer-implemented method in accordance with claim 22, further comprising routing, by the one or more router nodes, the content item to one of the plurality of transformation nodes in the transformation node cluster, the one of the plurality of transformation nodes being configured to execute one of the one or more content transformation types determined for the content item.24. The computer-implemented method in accordance with claim 23, wherein the one or more required transformation types comprise a transformation chain comprising a plurality of transformation types, and wherein the appropriate subset has a first transformation type to tnmsform the content item to a first intermediate content item having a first intermediate format.25. The computer-implemented method in accordance with claim 24, wherein the one or more router nodes are frirther configured to: route the first intermediate content item to a second appropriate subset, a second content transformation for which the second appropriate subset is designated matching a second required transformation type of the one or more required transformation types for transforming LI') the first intermediate content item to either a second intermediate content item having a second intermediate format or to a target content item having the target format. r r26. The computer-implemented method in accordance with claim 25, wherein the transformation chain comprises a transformation type resulting in formation of multiple intermediate content items, and each of the multiple intermediate content items requires a subsequent transformation.27. The computer-implemented method in accordance with claim 26, wherein the method further comprises: routing, by the router, the multiple intermediate content items to multiple transformation nodes in the transformation node cluster, the multiple transformation nodes each being configured to execute a same one of a plurality of content transformation types such that the multiple intermediate content items are processed in parallel.28. The computer-implemented method in accordance with any of claims 20 to 27, wherein at least one of the one or more router nodes is also included in one of the two or more subsets of the plurality of nodes.29. The computer-implemented method in accordance with any of claims 20 to 28, wherein each of the plurality of transformation nodes is configured to execute only one of the plurality of content transformation types.30, The computer-implemented method of any of claims 20 to 29, further comprising: IC) creating a transient handler, the transient handler managing the request; tagging the request with a handler context and a request identifier; r mapping a response received from the one of the plurality of transformation nodes and containing the handler context and request identifier back to the request and to the transient handler, the transient handler keeping a record of the response and notifying the router when all transformations of a first transformation type of the one or more transformation types has been completed; and routing intermediate content items resulting from the first transformation type to one or more additional nodes of the plurality of transformation nodes in the transformation node cluster, the one or more additional nodes of the plurality of transformation nodes being configured to transform the intermediate content items to one or more next intermediate content items or to the target content format of the content item.3 The computer-implemented method in accordance with any of claims 20 to 30, wherein the transformation usage data comprises historical data pertaining to content transformation requests requested from client machines.32. The computer-implemented method in accordance with claim 3], wherein the estimating further comprises use of at least one of a predictive model, machine learning, and a neural network to make predictions based on the historical data.33, The computer-implemented method in accordance with claim 31 or claim 32, wherein the stored data comprise stored callbacks provided by the one or more router nodes to client machines in response to previously completed content item requests. IC)34. The computer-implemented method in accordance with claim 33, wherein the r stored callbacks comprise at least one of references to requested content items, arrays of transformed content item references referencing intermediate content items created in transformation chains to produce transformed content items in response to the completed content item requests, listings of options specified in the content item requests, listings of operations performed to convert content items to target content formats, and identifiers for transformation nodes that participated in the transformation chains.35. The computer-implemented method in accordance with any of claims 20 to 33, wherein the transformation usage data further comprises at least one of a type and number of business process applications supported by the content management system repository, usage factors defined for content item transformations expected to be called by the business process applications, and an absolute or relative amount of mobile device usage supported by the content management system repository.36. The computer-implemented method in accordance with any of claims 20 to 35, wherein the estimating further comprises acting on at least one human input parameter relating to expected transformation demands.37. The computer-implemented method in accordance with any of claims 20 to 36, further comprising transforming the content item from a source format to a destination format and providing the transformed content item to a client machine that requested the item.IC) 38. The system or computer-implemented method in accordance with any preceding claim, wherein the plurality of transformation types comprises a transformation type r that includes transforming the content item from the current content format comprising a portable document format into the target content format, which comprises the content item split into two or more sections in the portable document format.39, The system or computer-implemented method in accordance with any preceding claim, wherein the plurality of transformation types includes transforming the content item from the current content format in an editable text document format into a target content format of an image format.40. A computer program product comprising a machine-readable medium storing instmctions that, when executed by at least one programmable processor, cause the at least one programmable processor to perform operations in accordance with the computer-implemented method of any of claims 20 to 39. IC) r r IC)