CN103488754A - Method for transparent interoperability of heterogeneous document cooperation cloud services - Google Patents

Abstract

The invention belongs to the technical field of cloud computing and computer-supported collaborative work, and specifically relates to a method for transparent interoperability among heterogeneous document collaboration cloud services. The method provides transparent cross-cloud document collaboration services for users using heterogeneous document collaboration cloud services through an intermediate agent, and ensures the consistency of shared documents. The beneficial effects of the present invention are: on the one hand, users can use their familiar services to participate in cross-cloud document collaboration without learning or paying for new services, and without installing any software or plug-ins; on the other hand, service providers Interoperability with other similar services can be achieved without any additional work, such as developing standards or modifying code.

Description

A Method for Transparent Interoperability among Heterogeneous Document Collaboration Cloud Services

technical field

The invention belongs to the technical field of cloud computing and computer-supported collaborative work, and in particular relates to a method for transparent interoperability among heterogeneous document collaboration services.

Background technique

1. Cloud Collaborative Editing Service

In today's economic globalization, cloud services have become an ideal online collaboration platform. Geographically dispersed enterprises, organizations and individuals can access cloud services anytime and anywhere through any device for online collaboration. More and more users are turning to cloud services because of their high availability, elasticity, and low maintenance costs. Document collaboration cloud service is a representative type of online collaboration cloud service, such as Google Docs, Microsoft Office Web Apps, Zoho Office Suite, etc. They can support geographically dispersed users to carry out real-time online document collaboration. Users can open and edit online documents through browsers, and the updates of documents by each collaborative user can be synchronized in real time and ensure the final consistency of each user's document view. Document collaboration cloud services are usually tightly integrated with cloud services such as document storage, sharing, e-mail, and instant messaging to provide end users with a complete online collaboration solution.

2. Real-time collaborative group editing technology

Real-time collaborative group editing technology is one of the most important technologies for realizing document collaboration cloud services. Address Space Transformation AST (Address Space Transformation) and Operation Transformation OT (Operation Transformation) are the two most representative real-time collaborative group editing technologies. They perform operations correctly by converting remote operations or local document address spaces into the correct form, automatically resolve conflicts between concurrent operations, and achieve consistent maintenance of document state.

The basic idea of AST is: the shared document is copied on each user collaboration site, and the user's local operation can be immediately executed and sent to other sites; when the remote operation is received, the impact of concurrent operations on the document state is first removed, and the The document (that is, the address space) is converted to the state when the operation was generated. After the remote operation is executed in its original form, the effect of the concurrent operation is restored to convert the document to the latest state. In order to realize the address space conversion, each node has a valid status flag; the executed operation is stored in the corresponding document space node; the deleted node will be kept in the address space. The removal and recovery of the effect of concurrent operations can be achieved by traversing the address space and correctly setting the effective status flags of the nodes.

3. Interoperability of cloud services

Cloud services usually adopt a pay-as-you-go deployment model, and enterprise users can choose appropriate cloud services according to their preferences and needs to cope with changes in market demand. On the one hand, users usually do not want to be bound by one service provider; on the other hand, users are usually unwilling to give up the services and user interface they are already familiar with and turn to other service providers and thus pay extra fees. Therefore, users are usually more inclined to choose cloud services that support better interoperability.

Standardization is a way to achieve interoperability between cloud services. However, because it is difficult for major service providers to reach an agreement due to their own interests, the process of service standardization is often very slow. Cloud Service Brokerage (Cloud Service Brokerage) is a more effective solution to achieve interoperability between cloud services before service standardization. For example, Jolidriver is a typical cloud storage proxy service, which shields the differences between heterogeneous cloud storage services by encapsulating the APIs of various cloud storage services, and provides users with a unified cloud storage service interface.

However, document collaboration cloud services are multi-user interactive applications. Since multiple users are involved in the simultaneous operation of shared documents, it is much more difficult to achieve interoperability between heterogeneous document collaboration cloud services than cloud storage services. First, existing document collaboration services only provide APIs for document storage and sharing, and do not directly provide APIs that support shared document editing; second, concurrent operations on shared documents by multiple users on different cloud services may lead to Operational conflicts and eventually lead to inconsistencies in document status. The challenges of operational conflict resolution and data consistency maintenance in traditional real-time collaborative group editing systems still exist in the interoperability of heterogeneous document collaboration cloud services; third, support heterogeneous Interoperable proxy services between document collaboration services cannot change the experience of users using existing services, for example, they cannot require users to install new software or browser plug-ins, and cannot interfere with the collaboration between users.

4. Key issues to be addressed

A key problem to be solved in the present invention is how to realize interoperability between document collaboration cloud services in a transparent manner? Transparency here mainly includes two meanings: one means that the service provider does not need to pay any work, such as formulating standards or modifying codes, to achieve interoperability with other similar services; the other means that It means that users do not need to install any software or plug-ins, and can use their familiar services to conduct real-time online document collaboration with the user interface and other users.

Another key issue that needs to be solved is how to ensure document consistency in cross-cloud document collaboration? As users collaborate on cross-cloud documents using their respective familiar services, concurrent document operations may conflict or cause shared documents to be in an inconsistent state. The difficulties in achieving consistency maintenance in document collaboration mainly exist in two aspects: on the one hand, we cannot accurately know the consistency maintenance technology, conflict resolution method, document synchronization protocol, etc. used by each document collaboration service, and each service provides On the other hand, the existing real-time collaborative group editing technology relies on the acquisition of user operation history. Under the premise of ensuring the above-mentioned interoperability transparency, we cannot obtain the user's operation history. Therefore, we can only treat each document collaboration cloud service as a "black box" to maintain document consistency in cross-cloud document collaboration.

Contents of the invention

In order to solve the problems in the prior art, the present invention proposes a method for transparent interoperability among heterogeneous document collaboration cloud services.

In a method for transparent interoperability between heterogeneous document collaboration cloud services provided by the present invention, the cross-cloud document collaboration is provided in the form of an intermediate proxy service, and a robot is introduced into the intermediate proxy service, which communicates between the robot and the document collaboration cloud service to which it belongs. The automatic interaction between robots and the document interaction between robots realizes the transparent interoperability between heterogeneous document collaboration cloud services; the details are as follows:

Using web browser automation technology to read the document status, convert the view representation of the document into a unified document model representation, and apply the received remote operation to synchronize the local document with the shared document in the group and realize the collaboration between the robot and the document Automatic interaction between services.

The robots of the user groups participating in the cross-cloud document collaboration are connected to each other through the collaboration engine to form a P2P document collaborative editing network. The update operation of the user group users to which the robot belongs is derived through inter-document diff (document comparison), and the document update is detected. Utilize the address space conversion technology AST to convert the received remote operations of robots from other user groups into the correct form, and execute the converted operations, so that the shared documents in all document collaboration services are in a consistent state, realizing the robot-to-robot interaction document interaction.

1. Main idea and intermediate proxy service design

和

，他们在不同的文档协作云服务上独立地对同一份文档进行编辑。对于组内的任一用户，例如

，

可以看到

中其他用户对文档的做出的任何更新。因此，对于任意两个用户

和

，如果他们能告诉对方所属组内文档的变化，并将对方告知的文档变化正确的应用于自己做所属组的文档上，那么

和

中的文档即可通过

和之间的“对话”实现同步。 Document collaboration services allow geographically dispersed users to edit a document at the same time, and the documents and views between users will be kept automatically synchronized. Consider the two groups of users in Figure 1

and

, they independently edit the same document on different document collaboration cloud services. For any user in the group, e.g.

,

can be seen

Any updates made to the document by other users in the . Therefore, for any two users

and

, if they can tell the other party about the changes in the documents in the group they belong to, and apply the document changes informed by the other party to the documents in their own group, then

and

Documentation in the

and The "dialogue" between them is synchronized.

Based on the above idea, the present invention introduces a robot into the intermediate proxy service, and automates the interaction between the robot and the document collaboration cloud service and the document synchronization dialogue between the robots, so as to realize transparent interoperability among heterogeneous document collaboration cloud services. From the user's point of view, they do not need to install any software or plug-ins, and can use familiar cloud services to participate in cross-cloud document collaboration according to their own habits. Users only need to regard the robot user as an ordinary member of the same group. Shared documents can be shared with the robot in the proxy service. From the perspective of service providers, robot users are no different from other ordinary users. They do not need to make any changes to the source code, and do not even need to provide any API to support interoperability with other similar services.

中的文档更新可以通过如下步骤同步到组

中： Robots in the intermediate proxy service work in the same way as ordinary users, using web browsers to participate in document collaboration in the group they belong to. Group

Document updates in can be synchronized to the group by the following steps

middle:

将其本地文档与所属文档协作服务器进行同步以获得组内其他用户对文档的更新； Step 1 :

Synchronize their local documents with their document collaboration server to obtain updates from other users in the group;

读取当前的文档状态； Step 2 :

Read the current document state;

将最新的文档状态与之前读取的文档状态进行比较； Step 3 :

Compare the latest document state with the previously read document state;

Step 4 : If any updates are detected, Send these updates as operations to ; Otherwise, end this document synchronization;

Step 5 : Apply the received document update to the local document;

将其本地文档与所属文档协作服务器进行同步以更新组内的共享文档状态。 Step 6 :

Synchronize their local documents with their own document collaboration server to update the status of shared documents within the group.

Figure 2 shows the architecture of the intermediate proxy service. Each robot user consists of three core components: CES User Interface (CUI: CES User Interface), CES Adapter (CA: CES Adapter), and Collaboration Engine (CE: Collaboration Engine).

The CES user interface is used to participate in the document collaboration of the belonging group. The user interface used by the robot is the same as that of ordinary users, that is, the web user interface opened in the web browser.

The CES adapter realizes the automated interaction between the robot and the document collaboration service it belongs to, including: reading the document status, applying the received remote operation, and realizing the synchronization of the local document and the shared document in the group it belongs to. Intuitively, the CES adapter can be regarded as the "eye" and "hand" of the robot. The implementation of the CES adapter is related to the document collaboration service, but it implements the following public interface:

interface CESAdaptor {

ReadDocument() : Document

Apply(Operation[] operations) : void

Sync() : void

}

The ReadDocument method returns the document object displayed in the user interface. A document object corresponds to one or more pieces of HTML code in a web user interface. Robots need to "talk" on the basis of the same document model.

The Apply method is called to apply editing operations to the robot's local document. These operations come

From other bots, that is, from user groups using other document collaboration cloud services, and also defined on the unified document model. The call to the Apply method must ensure that the incoming operations are defined on the document currently displayed in the robot's user interface.

The Sync method is called to realize the synchronization between the robot's local document and the shared document in the cloud service.

step. Specifically, when the Sync method is called, other ordinary users in the same group will see changes made to the document by users in other groups. At the same time, if other ordinary users in the group make changes to the document, the local document of the robot will also will be updated.

The collaborative engine is responsible for detecting document updates and achieving document synchronization with other robots. In cross-cloud document collaboration, the robot will periodically synchronize with the server and check for document updates, that is, changes made to documents by other ordinary users in the group it belongs to. If any changes are detected, the robots will send these updates as actions to the collaborative engines of other robots. When receiving remote operations from other robots, due to consistency issues, these operations cannot be directly applied to the robot's local file, but first need to convert these operations into the correct form.

The collaboration engine is applicable to all document collaboration cloud services, it only needs to be implemented once, and it will be reused when integrating new document collaboration cloud services. It is worth noting that only two robots are shown in Figure 2. In fact, the number of robots depends on the number of user groups participating in cross-cloud document collaboration. These robots are connected to each other through the collaboration engine in the intermediate agent service, forming a P2P document collaboration network.

2. Adaptation method of document collaboration cloud service

Since almost all document collaboration cloud services do not provide document editing APIs, a more general method—Web browser automation technology is adopted in the present invention to realize automatic interaction between robots and document collaboration services. Selenium is a set of tools for automating web applications. It uses JavaScript injection to interact with UI elements in web pages and provides client APIs in Java, C#, Ruby, and Python.

Before adapting the document collaboration service, we need to define a unified document model and operations. The choice of document model depends on the document formats supported in document collaboration, such as Office Open XML for describing formatted documents, SVG DOM for describing graphic documents, etc. Here we take a simplified formatted document as an example, and use Selenium to implement the adaptation of Google Document and Microsoft Word Web App to illustrate how to implement the methods in the CES adapter interface through Web browser automation technology. In the simplified formatted document model, a text document is constructed as a sequence of formatting characters. The operations supported on the document include Insert(pos, ch) and Delete(pos, ch), which represent inserting a formatted character at a specified position and deleting a character at a specified position, respectively. The modification of the character format can be realized through the combination of Delete and Insert operations.

Reading the document state ( ReadDocument ) is actually a reverse engineering of converting the view representation of the document into the model representation. In a web user interface, a document corresponds to one or more code fragments in an HTML view. For example, Figure 3 (a) and (b) show a sample text document and its corresponding HTML code snippet in the Microsoft Word Web App user interface, respectively.

In Microsoft Word Web App, the document content is located in the DOM element <div id="PageContentContainer" ...>. Selenium is able to target the element using the element's CSS selector div#PageContentContainer. Each paragraph is located in the element p.Paragraph; consecutive characters with the same format are located in the element span.NormalTextRun, and the formatting of the text is specified by the element span.TextRun. Figure 3 (c) and (d) respectively show the hierarchical structure of document display related Web elements in Microsoft Word Web App and Google Document. Selenium supports locating web elements through identifiers, tags, CSS selectors, etc., and obtaining the attributes and internal text of elements. Thus, the content of a document can be extracted by traversing the document to reveal a hierarchy of related elements.

Similar to the interaction between ordinary users and the Web user interface, the application remote operation ( Apply ) can be completed by sending a series of mouse or keyboard messages to the user interface. Before applying an editing operation, the cursor needs to be moved to the editing position specified by the operation parameter pos. We adopt two strategies to optimize the efficiency of cursor positioning: (1) start positioning from the beginning of the paragraph. The number k of the paragraph where the editing position is located and the characters before the paragraph can be easily calculated. Therefore, the cursor positioning can be positioned at the beginning of the paragraph by one mouse operation and realized by pos-k right-click operations; (2) record the previous edit Location. For the editing of the same paragraph, the cursor positioning can be completed by calculating the relative position to the previous editing position and moving the cursor to the right or to the left. The second strategy works well for sequential insertion and deletion operations. After setting the cursor to the correct editing position, the delete operation can be completed directly by sending a "DELETE" keyboard message to the web user interface. When inserting formatting characters, if the format of the character to be inserted is the same as the current format, just send a keyboard message directly; otherwise, you need to interact with UI controls related to font settings to set the character format. Figure 4 shows the UI controls related to font settings in Google Document and Microsoft Word Web App. For example, bolding or canceling bolding of fonts can be achieved through the following code snippets:

// Enable/disable the bold font setting in Google Document

driver.findElement(By.cssSelector("div#boldButton")).click()

// Enable/disable the bold font setting in Microsoft Word Web App

driver.findElement(By.cssSelector("a#fsbcBold-Small")).click()

The Sync method in the CES adapter interface provides the collaborative engine with the document synchronization programming interface within the group. In fact, it is an extension of the document collaboration service to support the synchronization of asynchronous documents. The document synchronization in Microsoft Word Web App is itself asynchronous, and the user needs to click the "Save" button or the "Ctrl+S" button to explicitly save the document to trigger the document synchronization. So, the Sync method of the Microsoft Word Web App adapter can be done by sending a mouse click event to the "Save" button (a #qatSave-Small) control. However, the documents in Google Document are synchronized in real time, user documents are automatically saved and changes made to documents by other users in the group can be seen in real time. Fortunately, Google Document can support offline document editing. Offline operations are saved locally and synchronized to the server when the network is available. Therefore, the asynchronous document synchronization of Google Document can be realized by switching between online and offline working modes.

3. General collaborative engine

和

的原始文档均为

。在文档同步之前，

和

分别检测到文档状态被组内用户更改为

和

，如果他们彼此交换更新操作

和

，并将操作直接应用于本地文档上，

的文档状态将变为

，而的文档状态将变为

，其中字符

被错误的删除了。 Since in cross-cloud document collaboration, users in different document collaboration cloud services modify the shared document concurrently, the received remote operations from other robots will result in inconsistent document status. For example, as shown in Figure 5(a), suppose the robot user

and

The original documents for

. Before the document is synced,

and

Respectively detects that the document status has been changed by a user in the group to

and

, if they exchange update operations with each other

and

, and apply the operation directly on the local document,

The document status for will change to

,and The document status for will change to

, where the characters

was deleted by mistake.

在执行之前被转换成了

；类似的，

在执行之前被转换成

，尽管其转换后的形式和其原有形式相同。 Therefore, the present invention adopts the address space conversion technology AST, and before executing the received remote operation, first converts it into the correct form, and then applies the converted operation to the local file of the robot through the Apply method in the CES adapter, As shown in Figure 5(b), the remote operation

before execution is converted to

;akin,

is converted to

, even though its transformed form is the same as its original form.

In the intermediate proxy service, each robot user forms a P2P document collaboration network. For a robot user, its local document corresponds to the shared document in the document collaboration service to which it belongs, and the editing operations on the document by other ordinary users in the group can be regarded as operations performed by the robot user on its local document.

The implementation of the collaboration engine can be applied to all document collaboration cloud services, including two core methods DetectLocalChanges and ExecuteRemoteChanges , which are used to detect the update of the local document and execute the received remote operation respectively. After all user actions have been performed on each robot's local document, the shared document across all document collaboration services will be in a consistent state.

The DetectLocalChanges method can be implemented as follows:

Step 1 : Call the Sync method in the CES adapter to update the local document status;

Step 2 : Call the ReadDocument method in the CES adapter to read and record the current document status;

Step 3 : Perform a diff operation with the previously recorded document status;

Step 4 : If an update is detected, send the update action to other robot sites.

In a typical real-time collaborative group editing, the operations performed by users will be captured and recorded. However, in the interoperation between document collaboration services, robot users cannot know the operation history of other ordinary users in the group on documents. Therefore, in this invention, we use the way of diff between documents to derive the update operation on the document.

，接收到的远程操作，以及操作产生时文档状态的向量时间戳

，ExecuteRemoteChanges方法可以按照如下步骤实现： Given remote robot user identifier

, the received remote operation , and a vector timestamp of the document state when the operation occurred

, the ExecuteRemoteChanges method can be implemented as follows:

Step 1 : If the operation is in the non-causal ready state, add it to the waiting queue and return; otherwise, skip to step 2;

Step 2 : Convert the operation opts into the correct form through the AST method;

Step 3 : Call the Apply method in the CES adapter to apply the converted operation to the local document;

Step 4 : Check the waiting queue, repeat steps 2 and 3 to execute the causal ready operation;

Step 5 : Call the Sync method in the CES adapter to update the shared document of the group to which it belongs.

The beneficial effects of the present invention are: on the one hand, users can use their familiar services to participate in cross-cloud document collaboration without learning or paying for new services, and without installing any software or plug-ins; on the other hand, service providers Interoperability with other similar services can be achieved without any additional work, such as developing standards or modifying code.

Description of drawings

Figure 1: The basic idea of the present invention: realize the transparent interoperability among heterogeneous document collaboration cloud services through the interaction between the automation robot and the associated document collaboration cloud service and the document synchronization between the robots.

Figure 2: Architecture of the intermediate proxy service.

Figure 3: (a) a sample text document; (b) the HTML code snippet corresponding to the sample document in the Microsoft Word Web App user interface; (c)–(d) Microsoft Word Web App and Google Document related to the document display Web Element hierarchy.

Figure 4: UI controls related to font settings and their CSS selectors in (a) Google Document and (b) Microsoft Word Web App.

Figure 5: (a) Inconsistencies in document synchronization between robots; (b) Operational transitions.

Detailed ways

The intermediate proxy service for cross-cloud document collaboration can be implemented in the following ways:

(1) Realize the general collaboration engine according to the content of the invention, where DetectLocalChanges and

For the pseudocode of the ExecuteRemoteChanges method, see Appendix 2 and Appendix 3, respectively. The operation conversion method in the collaborative engine can refer to the article "Consistency maintenance based on the mark & retrace technique in groupware systems" published by Gu Ning et al. on ACM Group'05 ("Consistency maintenance based on the mark & retrace technique in groupware systems") method"). The realized collaboration engine can be reused, as shown in Figure 2, the robots are connected through the collaboration engine to form a P2P document collaboration network. Therefore, it is also necessary to implement a corresponding message sending and receiving module in the collaborative engine for broadcasting local operations and receiving remote operations.

(2 The adaptation method of the document collaboration cloud service as described in the summary of the invention, for each document collaboration cloud service that needs to be accessed (for example: Google Document or Microsoft Word Web App), independently implement the three methods in the adapter interface CESAdaptor : ReadDocument , Apply , Sync . During the implementation process, it is necessary to analyze the document structure, document elements and corresponding interface elements, and use the Selenium framework to access and interact with them. It is worth noting that the document status returned by the ReadDocument method must be defined in the unified On top of the document model of .Apply method implementation example can be found in Appendix 1, the pseudo-code gives an example of performing formatting character insertion and deletion operations.

(3) As shown in Figure 2, each robot consists of a CES Web user interface, a CES adapter and a general protocol

The same engine is composed of three components. Each robot can run in a separate process. In order to achieve cross-cloud document collaboration, it is also necessary to uniformly schedule each robot process. The specific scheduling method can be determined according to the actual application process. For example, cross-cloud document collaboration can be started in the following manner: start the robot process→start the web browser→open the shared document→start the collaboration engine. In order to ensure that the document states in heterogeneous cloud services can be correctly synchronized, before starting the collaborative engine, it is necessary to ensure that the document states read by each robot are in a consistent state.

Appendix 1. Actions applied to local documents for robots

Appendix 2. Detecting Document Status Changes

* Among them, the Diff method can use Mayes Diff to return the sequence of Insert and Delete operations that generate newdoc from the document state doc.

Appendix 3. Performing remote operations

* Operation conversion can be implemented using AST technology. For the implementation method of this technology, please refer to the following references: Gu, Ning, Jiangming Yang, and Qiwei Zhang. "Consistency maintenance based on the mark & retrace technique in groupware systems." In Proceedings of the 2005 international ACM SIGGROUP conference on Supporting group work , pp. 264-273. ACM, 2005.

Claims (6)

1. A method for transparent interoperability between heterogeneous document collaboration cloud services, characterized in that the cross-cloud document collaboration is provided in the form of an intermediate proxy service, and a robot is introduced into the intermediate proxy service, which communicates between the robot and the document collaboration cloud service to which it belongs. Automatic interaction between robots and document interaction between robots to realize transparent interoperability between heterogeneous document collaboration cloud services; details are as follows: Use web browser automation technology to realize the automatic interaction between the robot and its document collaboration service: read the document status, convert the view representation of the document into a unified document model; apply remote operations from other robots; convert local documents Synchronize with shared documents in the group; Robots of user groups participating in cross-cloud document collaboration are connected to each other through a collaborative engine to form a P2P document collaborative editing network: deduce the update operation of the user group users to which the robot belongs through the diff between documents, and detect document updates; then use address space conversion Technology AST converts the received remote operations of robots from other user groups into the correct form and executes them, so that the shared documents in all document collaboration services are in a consistent state and realize document interaction between robots. 2. The method for transparent interoperability between heterogeneous document collaboration cloud services according to claim 1, characterized in that: the robot is composed of three core components: CES user interface, CES adapter and collaboration engine; The CES user interface is a web user interface opened by a web browser, and is used to participate in the document collaboration of the group to which it belongs; The CES adapter is used to realize the automatic interaction between the robot and the associated document collaboration service; The collaborative engine is responsible for detecting document updates and realizing document synchronization with other robots. 3. The method for transparent interoperability between heterogeneous document collaboration cloud services according to claim 2, wherein the CES adapter implements the following public interface: interface CESAdaptor { ReadDocument() : Document Apply(Operation[] operations) : void Sync() : void } in: The ReadDocument method returns the document object displayed in the user interface; The Apply method is called to apply the editing operation to the robot's local document; The Sync method is called to synchronize the local document of the robot with the shared document in the cloud service to which it belongs. 4. The method for transparent interoperability among heterogeneous document collaboration cloud services according to claim 1, characterized in that: the Selenium framework is adopted in the web browser automation technology to realize the automatic interaction between the robot and the associated document collaboration service . 5. The method for transparent interoperability between heterogeneous document collaboration cloud services according to claim 1, characterized in that: the method for detecting document updates in the collaboration engine is implemented in the following steps: Step 1: Call the Sync method in the CES adapter to update the local document state; Step 2: Call the ReadDocument method in the CES adapter to read and record the current document status; Step 3: Perform a diff operation with the previously recorded document status; Step 4: If an update is detected, send the update action to other robot sites. 6. The method for transparent interoperability between heterogeneous document collaboration cloud services according to claim 1, characterized in that: synchronization with other robots is realized in the collaboration engine, and is realized according to the following steps: Step 1: If the operation is in the non-causal ready state, add it to the waiting queue and return; otherwise, skip to step 2; Step 2: Convert the operation sequence into the correct form through the AST method; Step 3: Call the Apply method in the CES adapter to apply the converted operation to the local document; Step 4: Check the waiting queue, repeat step 2 and step 3 to execute the causal ready operation; Step 5: Call the Sync method in the CES adapter to update the shared document of the group to which it belongs.

Images