Method for compressing monitoring data of FC network switch

Abstract

The invention provides a method for compressing monitoring data of an FC (Fiber Channel) network switch. The method comprises the following steps of: segmenting the monitoring data of the FC network switch into a fixed field H and a load P; a knowledge base L and a compiling table D are initialized according to the format defined by the FC frame protocol, the compiling table D is used for storing all roots in the load P, and the knowledge base L is used for converting subfields in the fixed field H into set coded values; compressing the fixed field H into a first compressed encoding stream according to the knowledge base L; compressing the load P into a second compressed encoding stream according to the compiling table D; and splicing the first compressed encoding stream and the second compressed encoding stream to generate compressed monitoring data. And through the established knowledge base L and compilation table D, a relatively long character string is represented by a relatively short code, so that the compression of the monitoring data of the FC network switch is realized.

Description

Technical Field

[0001] The invention relates to the technical field of data compression, and in particular to a method for compressing monitoring data of an FC network switch. Background Art

[0002] In airborne avionics systems, it is crucial to monitor the data frames transmitted in the FC network in real time and store them properly for data debugging and analysis. By monitoring and recording these flows, network fault diagnosis and fault location can be performed during ground debugging. However, due to the high communication rate and large data volume of the FC network, the storage space of the monitoring card is limited, so it is necessary to maximize storage utilization and reduce resource waste. In order to efficiently utilize limited bandwidth resources, efficient algorithms and data structures need to be used to compress FC network traffic. By reducing network data for compression, the volume occupied by data can be greatly reduced, thereby reducing storage costs.

[0003] However, there is currently no effective compression method to compress data frames. Summary of the invention

[0004] In view of this, the embodiment of this specification provides a method for compressing FC network switch monitoring data, so as to achieve adaptive improvement of the LZW algorithm based on the dictionary model compression technology, help reduce the amount of data for the complexity of large-scale deployment of high-speed networks, and also have scalability and adaptability, ensuring that the data transmission meets the needs of real-time communication.

[0005] The embodiments of this specification provide the following technical solutions:

[0006] A method for compressing monitoring data of a FC network switch, comprising:

[0007] The monitoring data of the FC network switch is divided into a fixed field H and a load P;

[0008] Initialize the knowledge base L and the compilation table D according to the format defined by the FC frame protocol, wherein the compilation table D is used to store all the roots in the payload P, and the knowledge base L is used to convert the subfields in the fixed field H into the set encoding value;

[0009] Compressing the fixed field H into a first compressed encoding stream according to the knowledge base L;

[0010] Compressing the payload P into a second compressed encoding stream according to the compilation table D;

[0011] After the first compressed coded stream and the second compressed coded stream are spliced ​​together, compressed monitoring data is generated.

[0012] Furthermore, the knowledge base L is initialized according to the format defined by the FC frame protocol, including:

[0013] The monitoring data is divided into multiple protocol fields according to the format defined by the FC frame protocol, wherein each protocol field includes multiple groups of field explanations, and each group of field explanations includes field content and a coding value corresponding to the field content;

[0014] After setting the corresponding coding value for each field content, the protocol field, field content and coding value are saved in the knowledge base L.

[0015] Furthermore, the initialization compilation table D includes:

[0016] The hexadecimal number converted from 0 to 255 is used as the initial root, and the sequence number of the initial root is used as the index number;

[0017] The initial root and the index corresponding to the initial root are saved in the compiled table D.

[0018] Further, the fixed field H is compressed into a first compression coding stream according to the knowledge base L, including:

[0019] According to the protocol format definition of the FC frame, the fixed field H is divided into multiple subfields;

[0020] All subfields are processed in a loop until all subfields are processed, and the compressed codes generated by each subfield are concatenated to generate the first compressed code stream:

[0021] According to the protocol format definition, determine the protocol field to which the subfield belongs;

[0022] In the knowledge base L, according to the protocol field, the field content that matches the subfield and the encoding value corresponding to the field content are queried. If the matching field content is queried, the encoding value is output as a compressed code.

[0023] Furthermore, it also includes:

[0024] If the matching field content cannot be found, set the set code value corresponding to the current subfield;

[0025] The protocol field, subfield, and set coding value corresponding to the current subfield are respectively used as the protocol field, field content, and coding value in the knowledge base L, and are appended to the knowledge base L.

[0026] Further, the payload P is compressed into a second compressed encoding stream according to the compilation table D, including:

[0027] Further, the payload P is compressed into a second compressed encoding stream according to the compilation table D, including:

[0028] Read the first character in the payload P, obtain the index number corresponding to the first character from the compilation table D, and use the index number as the second compression code;

[0029] All characters after the first character in the payload P are read in order until all characters in the payload P are converted into the second compression code, and all the second compression codes are concatenated in order to generate a second compression code stream:

[0030] The character read is regarded as the first character, and the character after the first character read is regarded as the second character;

[0031] Concatenate the first character and the second character into a middle character, and determine whether the middle character is in the compilation table D;

[0032] If the middle character is in the compilation table D, the index number corresponding to the middle character is obtained from the compilation table D, and the index number is used as the second compression code.

[0033] Furthermore, it also includes:

[0034] If the middle character is not in the compilation table D, add the middle character to the compilation table D as a new root and assign a new index number to the new root;

[0035] The new index number is used as the second compression code.

[0036] Furthermore, it also includes:

[0037] If the first character or the middle character is not in the compiled table D, add the first character or the middle character as a new root to the compiled table D, and assign a new index number to the new root;

[0038] The new index number is used as the second compression code.

[0039] Compared with the prior art, the at least one technical solution adopted in the embodiments of this specification can achieve the following beneficial effects:

[0040] By building a knowledge base L and a compilation table D, a shorter code is used to represent a longer string to achieve compression. The basic principle is to extract different characters from the original file data, create a compilation table based on these characters, and then use the character index in the compilation table to replace the corresponding characters in the original file data to reduce the size of the original data. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 The present invention is a flowchart of a method for compressing monitoring data of a FC network switch according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0044] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0045] The present invention aims to maximize storage utilization, utilizes the characteristic information of FC network data packets, takes into account the characteristics of the airborne environment and user application requirements, and adaptively modifies the dictionary-based compression algorithm LZW algorithm. In order to reduce the time of dynamically creating compilation tables for the traditional LZW algorithm and reduce the time delay during the compression process, the present invention separates the fixed fields and the payload part of the data frame during the compression process and processes them separately. Figure 1 As shown, a knowledge base is used for the compilation table of fixed fields of the data frame, and the LZW algorithm is used to dynamically construct the compilation table for the payload part, thereby quickly achieving the compression task.

[0046] Step 1: Initialize the knowledge base L according to the relevant definition of ICD. Initialize the compilation table D, which contains all possible roots of the load.

[0047] Step 2: Initialize the prefix I, suffix C, and encoding stream S to be empty.

[0048] Step 3: After the monitoring buffer receives the data frame, the data frame is divided into a fixed field H and a payload field P according to the format definition;

[0049] Step 4, process the fixed field H. Split it in sequence according to the format defined by the FC frame protocol. Split out field I, find the corresponding content in L according to the protocol field to which I belongs, and then query whether it is in the key-value pair corresponding to the protocol field according to the specific content of I. If it is, output the encoded value of I to S. Otherwise, add I to the key-value pair of the relevant protocol field.

[0050] Step 5: Determine whether the encoding of H is complete. If the encoding of the header is complete, clear the string I and start to process the compression of the payload P. Otherwise, repeat step 3.

[0051] Step 6: Let character C be the next character in the data stream.

[0052] Step 7: Determine whether I+C is in the compilation table D. If so, extend I with C, that is, I=I+C. If not, output the code corresponding to the current prefix I to S. Then add I+C to the compilation table D and set I=C;

[0053] Step 8: Determine whether the encoding of P is completed. If completed, output the compressed string S; otherwise, return to step 5.

[0054] In step 1, the structure of the knowledge base L is defined as:

[0055]

[0056] The structure of compile table D is defined as {root 1: index number, root 2: index number, ...}. Since network traffic is usually transmitted in hexadecimal format, in the initialization of compile table D, 256 bytes from 0x00 to 0xff are filled into the table according to the index. The index number is a sequential number starting from 1.

[0057] In step 2, the prefix I refers to the corresponding field in the current processing stream. The suffix C refers to the next character to be read. The code stream S refers to the output code, that is, the data after the compression operation.

[0058] In step 3, the data frame transmitted by the FC network mainly consists of a SOF delimiter, frame content and an EOF delimiter. The frame content consists of an FC frame header, payload and CRC. Generally speaking, SOF and EOF occupy 4 bytes respectively, the FC frame header occupies 24 bytes, the CRC occupies 4 bytes, and the payload occupies 0 to 2112 bytes. Among them, SOF, EOF, and FC frame headers have a fixed protocol format, and different fields represent different meanings, so these parts are defined as fixed fields H. CRC is used to check the completeness of the frame header and payload fields. It has uncertainty like the payload field, so it is classified as the payload field P.

[0059] Specific implementation methods of the embodiments of the present invention:

[0060] Assume there is an FC data stream, the hexadecimal representation of the frame is:

[0061] 0xBCB55656 04010003 0401000A 493A0018 (frame header)

[0062] 00000000013EFFFF0000000000052301 0000000000000000200001400000000

[0064] 00004DE7148F00000000000000000000(payload)

[0065] BC6DCC62BC957575 (frame tail).

[0066] The knowledge base L used is shown in Table 1.

[0067] Table 1 (Common Knowledge Base L Reference Values ​​for Protocol Fields)

[0068]

[0069]

[0070]

[0071]

[0072] Step 1: Initialize the knowledge base L according to the relevant definition of ICD. Initialize the compilation table D, which contains all possible roots of the load. The initialization knowledge base L is:

[0073] {

[0074] "SOF":{"OxBCB55656":0x0,"0xBCB53636":0x1},

[0075] "EOF":{"0xBC95D5D5":0x0,"0xBC957575":0x1}

[0076] "R_CTL":{"0x22":0x0,"0x23":0x1,"0x24":0x2,"0x04":0x3},

[0077] "D_ID":{"0x10000":0x100,"0x10001":0x101,"0x10002":0x102,"0x100 03":0x103,"0x10004":0x104,"0x10005":0x105,"0x10006":0x106,"0x1 0007":0x107,"0x10008":0x108,"0x10009":0x109,"0x1000a":0x10a,"0 x1000b":0x10b,"0x1000c":0x10c,"0x1000d":0x10d,"0x1000e":0x10e},

[0078] "CS_CTL / Priority":{"0x0":0x0,"0x2":0x1,"0x4":0x2,"0x8":0x3},

[0079] "S_ID":{"0x10000":0x100,"0x10001":0x101,"0x10002":0x102,"0x10003":0x103,"0x10004":0x104,"0x10005":0x105,"0x10006":0x106,"0x10007":0x107,"0x10008":0x108,"0x10009":0x109,"0x1000a":0x10a,"0x1000b":0x10b,"0x1000c":0x10c,"0x1000d":0x10d,"0x1000e":0x10e},

[0080] "TYPE":{"0x01":0x0,"0x49":0x1},

[0081] "F_CTL":{"0x3A0018":0x0,"0x320018":0x1},

[0082] "SEQ_ID":{"0x00":0x00},

[0083] "DF_CTL":{"0x00":0x0},

[0084] "SEQ_CNT":{"0x00":0x00},

[0085] "OX_ID":{"0xFFFF:0xFFFF},

[0086] "RX_ID":{"0xFFFF:0x0},

[0087] "Parameter":{"0x00000000":0x0,"0x830":0x1,"0x1060":0x2},

[0088] "Message ID":{"0x15012":0x0,"0x15013":0x1,"0x52301":0x2},

[0089] "Reserved":{"0x0000000000000000":0x0},

[0090] "L+Prority":{"0x0":0x0,"0x1":0x1,"0x2":0x2,"0x4":0x3,"0x80":0x4,"0x81":0x5,"0x82":0x6,"0x84":0x7},

[0091] "Message Payload Length":{},

[0092] "LS_CMD":{"0xE0000000":0x0,"0xE1000000":0x1,"0xE4000000":0x2}

[0093] "FunctionID":{"0x2F00E008":0x0,"0x9200E00B":0x1}

[0094] }

[0095] Step 2: Initialize prefix I, suffix C, and coded stream S (compressed monitoring data) to be empty. Let I be empty, C be empty, and S be empty.

[0096] Step 3: After the monitoring buffer receives the data frame, it divides the data frame into fixed field H and payload field P according to the format definition. The stream is parsed and the fixed field H is divided. H includes SOF, FC header and tail EOF field. The SOF and FC headers include:

[0097] 0xBCB55656

[0098] 0x040100030401000A493A001800000000013EFFFF00000000

[0099] The protocol field is 0x49, which is the ASM protocol. Based on this, the ASM header information is divided into:

[0100] 0x00052301000000000000000002000014

[0101] The trailing EOF includes: 0xBC957575

[0102] The payload P includes: 0x0000000000004DE7148F0000000000000000000000 and BC6DCC62 (BC6DCC62 is CRC code)

[0103] Step 4, process the fixed field H (using the initialized knowledge base L). Split in sequence according to the FC frame protocol definition format. H is split into the set {0xBCB55656, 0x04, 0x010003, 0x04, 0x01000A, 0x49, 0x3A0018, 0x00, 0x00, 0x0000, 0x013E, 0xFFFF, 0x000000000005, 0xBC957575} according to the format definition. Split out field I, find the corresponding content in L according to the protocol field to which I belongs, and then query whether it is in the key-value pair corresponding to the protocol field according to the specific content of I. If so, output the encoded value of I to S. Otherwise, add I to the key-value pair of the relevant protocol field. First, take out 0xBCB55656 from the set and assign it to I. At this time, the protocol field to which I belongs is "SOF". In the knowledge base L, SOF involves two values, namely {"OxBCB55656":0x0,"0xBCB53636":0x1}, so the code of I is 0x0, and the code value is output to S.

[0104] Step 5, determine whether the encoding of H is compressed. If the encoding of the header is completed, clear the string I and start processing the compression of the payload P. Otherwise, repeat step 3. The encoding of H is not completed, return to step 4. I = 0x04 (process the second field 0x04 of the fixed field H), the corresponding protocol field is "R_CTL", query the knowledge base L, find the corresponding encoding value "0x3", output to S, S = 0x03. Repeat the above steps until I = 0x BC957575, the protocol corresponding to this field is "EOF", query the knowledge base L to take out the encoding value 0x1 and add it to S, S = 0x03103210a1000000013E002022141. At this time, the fixed field processing is completed, and P is initialized to empty.

[0105] Step 6: Let character C be the next character in the data stream (character C = first character). Now start processing the payload field. Let C be 0x00.

[0106] Step 7: Determine whether I+C is in the compilation table D. At this time, I+C is 0x00. In the compilation table D, (I+C)I is extended with C, that is, I is 0x00 (I=I+C).

[0107] Step 8, determine whether the encoding of P is completed. If completed, output the compressed string S, otherwise, return to step 6. At this time, the encoding of P is not completed, and return to step 6. Let C be the next character 0x00, I+C=0x0000, which is not in the compilation table D. If it is not in the compilation table D, then output the encoding corresponding to the current prefix I to S. That is, output 0x00 to S. Then add I+C to the compilation table D, and let I=C, so a new key-value pair {"0x0000":"0x100"} is added to D, and I=0x00 at this time.

[0108] The encoding of P is not yet complete. C is the next character 0x00, I+C=0x0000, in the compilation table D, C is used to expand I, I=0x000000. C takes the next character 0x00, I+C=0x000000, which is not in the compilation table D, and outputs the corresponding encoding of I to S, that is, S=S+0x100100, and then adds a new key-value pair {"0x000000":"0x101"} in D, I=0x00.

[0109] Take C as the next character 0x00, I+C=0x0000, in the compilation table D, use C to extend I, I=0x0000; continue to take C as the next character 0x00, I+C=0x000000 in the compilation table D, use C to extend I, I=0x000000. Take C as the next character 0x00, I+C=0x00000000 is not in the compilation table D, output the code of I to S, that is, S=S+0x100100101, and then add a new key-value pair {"0x00000000":"0x102"} in D, I=0x00.

[0110] Take C as the next character 0x4D, I+C=0x004D is not in the compilation table D, output the code of I to S, that is, S=S+0x10010010100, and then add a new key-value pair {"0x004D":"0x103"} in D, I=0x4D.

[0111] Take C as the next character 0xE7, I+C=0x4D7E, no longer compile table D, output the encoding of I to S, S=S+0x100100101004D, and then add a new key-value pair {"0x4D7E":"0x104"} in D, I=0x7E.

[0112] Take C as the next character 0x14, I+C=0x7E14, no longer in the compilation table D, output the encoding of I to S, S=S+0x100100101004D7E, and then add a new key-value pair {"0x7E14":"0x105"} in D, I=0x14.

[0113] Take C as the next character 0x8F, I+C=0x148F, no longer in the compilation table D, output the encoding of I to S, S=S+0x100100101004D7E14, and then add a new key-value pair {"0x148F":"0x106"} in D, I=0x8F.

[0114] Take C as the next character 0x00, I+C=0x8F00, no longer in the compilation table D, output the encoding of I to S, S=S+0x100100101004D7E148F, and then add a new key-value pair {"0x8F00":"0x107"} in D, I=0x00.

[0115] Take C as the next character 0x00, I+C=0x0000, in compile table D, use C to extend I, I=0x0000; continue to take C as the next character 0x00, I+C=0x000000 in compile table D, use C to extend I, I=0x000000. Take C as the next character 0x00, I+C=0x00000000 in compile table D, use C to extend I, I=0x00000000. Take C as the next character 0x00, I+C=0x0000000000 is not in the compilation table D, output the encoding of I to S, that is, S=S+0x100100101004D7E148F102, and then add a new key-value pair {"0x0000000000":"0x108"} in D, I=0x00.

[0116] Take C as the next character 0x00, I+C=0x0000, in compile table D, use C to extend I, I=0x0000; continue to take C as the next character 0x00, I+C=0x000000 In compile table D, use C to extend I, I=0x000000. Take C as the next character 0x00, I+C=0x00000000 In compile table D, use C to extend I, I=0x00000000. Take C as the next character 0x00, I+C=0x00000000 In compile table D, use C to extend I, I=0x000000000. Take C as the next character 0x00, I+C=0x000000000000, which is not in the compilation table D, output the encoding of I to S, that is, S=S+0x100100101004D7E148F102108, and then add a new key-value pair {"0x000000000000":"0x109"} in D, I=0x00.

[0117] Take C as the next character 0xBC, I+C=0x00BC, and do not compile table D. Output the encoding of I to S, S=S+0x100100101004D7E148F10210800, and then add a new key-value pair {"0x00BC":"0x10a"} in D, I=0xBC.

[0118] Take C as the next character 0x6D, I+C=0xBC6D, no longer compile table D, output the encoding of I to S, S=S+0x100100101004D7E148F10210800BC, and then add a new key-value pair {"0xBC6D":"0x10b"} in D, I=0x6D.

[0119] Take C as the next character 0xCC, I+C=0x6DCC, no longer compile table D, output the encoding of I to S, S=S+0x100100101004D7E148F10210800BC6D, and then add a new key-value pair {"0x6DCC":"0x10c"} ​​in D, I=0xCC.

[0120] Take C as the next character 0x62, I+C=0xCC62, no longer in the compilation table D, output the encoding of I to S, S=S+0x100100101004D7E148F10210800BC6DCC, and then add a new key-value pair {"0xCC62":"0x10d"} in D, I=0x62.

[0121] Step 8: Determine whether the encoding of P is completed. If completed, output the compressed string S. At this time, the encoding of the payload P is completed, and the encoding corresponding to I is output to S. The final S (compressed monitoring data) = 0x03103210a1000000013E002022141100100101004D7E148F10210800BC6DCC ​​62.

[0122] Beneficial effects of the embodiments of the present invention:

[0123] The traditional LZW compression algorithm is improved. Considering the use of the inherent characteristic information of FC data packets, the data frames are processed separately according to the format definition. Because in the definition of the frame format of the FC protocol, different bytes represent different meanings. An FC protocol frame usually consists of a SOF delimiter (frame header), frame content and an EOF delimiter (frame end). The frame content consists of an FC frame header, a payload and a CRC. In FC communication, except for the data and CRC fields in the payload part, the other parts have strict content definitions. Each part has a corresponding semantic representation. These contents are clearly defined in the ICD (Interface Control Definition) and the probability of randomness is small. In order to reduce the time of dynamically creating the compilation table of the traditional LZW algorithm and reduce the time delay in the compression process, the fixed fields and the payload part of the data frame are separated and processed separately during the compression process. A knowledge base dictionary is used for the compilation table of the data frame header, and the LZW algorithm is used to dynamically build the compilation table for the payload part, so as to quickly achieve the compression task.

[0124] The above is only a specific embodiment of the present invention, and cannot be used to limit the scope of the invention. Therefore, the replacement of equivalent components, or equivalent changes and modifications made according to the scope of protection of the patent of the present invention, should still fall within the scope of this patent. In addition, the technical features of the present invention can be freely combined with each other, with each other, and with each other.

Claims

1. A method for compressing monitoring data of a FC network switch, characterized in that: include: The monitoring data of the FC network switch is divided into a fixed field H and a load P; Initialize the knowledge base L and the compilation table D according to the format defined by the FC frame protocol, wherein the compilation table D is used to store all roots in the payload P, and the knowledge base L is used to convert the subfields in the fixed field H into set encoding values; Compressing the fixed field H into a first compressed encoding stream according to the knowledge base L; Compressing the payload P into a second compressed encoding stream according to the compilation table D; After splicing the first compressed coded stream and the second compressed coded stream, compressed monitoring data is generated.

2. The method for compressing FC network switch monitoring data according to claim 1, characterized in that: Initialize the knowledge base L according to the format defined by the FC frame protocol, including: The monitoring data is divided into a plurality of protocol fields according to the format defined by the FC frame protocol, wherein each of the protocol fields includes a plurality of groups of field explanations, and each group of the field explanations includes field content and a coding value corresponding to the field content; After setting a corresponding coding value for each of the field contents, the protocol field, the field contents and the coding value are saved in the knowledge base L.

3. The method for compressing FC network switch monitoring data according to claim 1, characterized in that: Initialize and compile table D, including: The hexadecimal number converted from 0 to 255 is used as the initial root, and the sequence number of the initial root is used as the index number; The initial root and the index corresponding to the initial root are saved in a compilation table D.

4. The method for compressing FC network switch monitoring data according to claim 1, characterized in that: Compressing the fixed field H into a first compressed encoding stream according to the knowledge base L includes: The fixed field H is divided into a plurality of subfields according to the protocol format definition of the FC frame; All subfields are processed in a loop until all subfields are processed, and the compression codes generated by each subfield are concatenated to generate a first compression code stream: According to the protocol format definition, determining the protocol field to which the subfield belongs; In the knowledge base L, according to the protocol field, the field content matching the subfield and the encoding value corresponding to the field content are searched, and if the matching field content is found, the encoding value is output as the compressed code.

5. The method for compressing FC network switch monitoring data according to claim 4, characterized in that: Also includes: If the matching field content cannot be found, set the setting code value corresponding to the current subfield; The protocol field, the subfield, and the set code value corresponding to the current subfield are used as the protocol field, the field content, and the code value in the knowledge base L, respectively, and are appended to the knowledge base L.

6. The method for compressing FC network switch monitoring data according to claim 1, characterized in that: Compressing the payload P into a second compressed encoding stream according to the compilation table D includes: Read the first character in the payload P, obtain the index number corresponding to the first character from the compilation table D, and use the index number as the second compression code; Read all characters after the first character in the payload P in order until all characters in the payload P are converted into the second compression code, and concatenate all the second compression codes in order to generate a second compression code stream: The character read is taken as the first character, and the character after the first character read is taken as the second character; concatenate the first character and the second character into an intermediate character, and determine whether the intermediate character is in the compilation table D; If the middle character is in the compilation table D, the index number corresponding to the middle character is obtained from the compilation table D, and the index number is used as the second compression code.

7. The method for compressing FC network switch monitoring data according to claim 6, characterized in that: Also includes: If the middle character is not in the compilation table D, add the middle character as a new root to the compilation table D, and assign a new index number to the new root; The new index number is used as the second compression code.

Images