JPH04129429A - Dictionary retrieval system for data compressor - Google Patents

Abstract

PURPOSE:To attain quick dictionary retrieval by forming a link list utilizing the external hash method with a consecutive address in the dictionary retrieval of LZW coding. CONSTITUTION:The system is provided with a dictionary memory 20 in which a link address of an external has address based on input data consists of consecutive address of a next memory 200 partially and with a dictionary retrieval means 16 generating continuously an address of the next memory 200 based on the input data and retrieving an object data of an expansion memory 300 coincident with the input data. That is, the index address of the dictionary memory 20 having a list structure based on the external hash method in the dictionary retrieval of the LZW coding is formed by a consecutive address. Thus, since a succeeding address is predicted when one hash address is decided, the retrieval address of an object character is more quickened and the coding by high speed read for the dictionary memory 20 is attained and the coding processing time is reduced.

Description

[Detailed description of the invention] 【overview】

Regarding a dictionary search method for a data compression device using LZW encoding as an improvement of the incremental decomposition type, which is a type of universal encoding, the dictionary search time is reduced by enabling high-speed reading of the dictionary memory using the list structure of the external hash method. With the aim of reducing The system is configured to use consecutive addresses, and after the initial search based on input characters, a search using consecutive addresses is performed to speed up the search.

[Industrial application field]

The present invention relates to a dictionary search method for a data compression device using LZW encoding as an improvement on the incremental decomposition type, which is a type of universal encoding. In recent years, computers have come to handle various types of data such as character codes, vector information, and images.
The amount of data handled is also rapidly increasing. When handling large amounts of data, by compressing the amount of data by eliminating redundant parts, you can reduce storage capacity and speed up transmission. Universal encoding has been proposed as a method that can compress such various data using one method. Here, the field of the present invention is not limited to character code compression.
Although it can be applied to a variety of data, in the following we will follow the nomenclature used in information theory, and refer to a single word unit of data as a character, and data that does not contain multiple words as a character string. . A typical universal code is the Ziv-Lempel code (for details, see Munakata 1 "Ziv-Lempel Data Compression Method", Information Processing, Vol. 26. No. 1.19f1)
(See Year 5). In Zifflempel codes, ■ Universal type ■ Incremental parsin type
Two algorithms have been proposed (g). Furthermore, as an improvement of the universal algorithm, LZ
There are SS codes (T, C, Be1l, “BeNe
r OPM/L Text Compression
, IEEE Trans.
, VOl, C0M-34, NO, 12, I! EC
, 1986). Moreover, as an improvement of the incremental decomposition type algorithm, L Z
There is a W (Lempel-2iv-Welch) code (T, A, WeIch, ``A Technique
e tarHigh-Performance Dat
aComputation, Computer,
(See June 1984). Among these codes, the LZW code has come to be used for file compression in storage devices because of its high-speed processing capability and simple algorithm.

[Conventional technology]

FIG. 7 shows an encoding processing flow using a conventional LZW code, and FIG. 8 shows a decoding processing flow. First, LZW encoding processing has a rewritable dictionary,
Divide the input character string into different character strings (substrings) and register these character strings in the dictionary with reference numbers in the order in which they appear, and also register the currently input character string in the dictionary. The longest-dispersed character string is represented by a reference number and encoded. FIG. 9 shows an explanatory diagram of LZW encoding, FIG. 10 shows an explanatory diagram of LZW decoding, and FIG. 11 shows an example of a dictionary structure created during decoding. In addition, in Figure 9.10.11, to simplify the explanation, a
An example of compressing and restoring data consisting of a combination of three characters bc is taken up. In the LZW encoding process in FIG. 7, first step 81, (
(hereinafter, "step" is omitted), a character string consisting of one character for each character is registered in the dictionary as an initial value, and then encoding is started. To encode Sl, a dictionary is searched using the input first character K to obtain a reference number ω, and this is used as the initial character string. Next, in S2, the next character K of the input data is read, and after checking whether character input is completed in S3, the process proceeds to S4, and the character K read in in 82 is added to the initial character string ω obtained in Sl. Search whether the added extended character string (ωK) exists in the dictionary. If the character string (ωK) is not in the dictionary in S4, proceed to S6 and use the code word code (
ω), add a new reference number to the character string (ωK), register it in the dictionary, replace the input character K in S2 with the reference number ω, increment the dictionary address n, and return to S2. Read the next character K. On the other hand, if the character string (ωK) is in the dictionary in S4, the character string (ωK) is replaced with the reference number ω in S5, the process returns to S2, and the maximum match length is increased in S4 until the character string (ωK) cannot be found in the dictionary. Continue searching. LZW encoding will be specifically explained as follows with reference to FIGS. 9 and 10. First, the input data 1nput in FIG. 9 is read from left to right. When you enter the first letter a, there are no matching strings in the dictionary other than the letter a, so 0UTPUT C0DEl(
The reference number ω) is output as a code word. Then, let the character a be the initial character string ω. Next, if the second character S is input, the extended character string ωKab obtained by adding this input character to the initial character string ω is not found in the dictionary, so the character S 0UTPUT C0DE 2 is output as a code word. Then, the extended character string ω is added with reference number 4 to =ab and registered in the dictionary. In actual dictionary registration, the character string 1b is registered as shown on the right side of FIG. Then, the character S becomes the word-initial character string ω. If you then input the third character a, the expanded character string ω, which is the initial character string ω added to the character S, =ba=2a is not in the dictionary, so the character a is 0UTPUTCOI)E 1
After outputting as a code word, =ba is added to the extended string ω by 2
It is represented by a and is registered in the dictionary with the reference number 5. Then, the letter a becomes a new initial character string ω. For the fourth input character, the extended character string ω = ab is already registered in the dictionary as code word 4 of 1b, so
Let the character string ωK be a new initial character string ω, and the fifth character C
Input and create the expanded character string ω =4 c=a b c. Since =abc is not registered in the dictionary for this extended character string ω, 0UTPUT C of character string a b = 1 b
Output 0DE 4 as a code word, and add =a to the extended character string ω
bc is registered in the dictionary as code word 6 in the form of 40. This process continues in the same manner. The decoding process shown in FIG. 8 performs the reverse operation of the encoding process shown in FIG. 7. In the LZW decoding shown in FIG. 8, decoding is started after a character string consisting of one character for every character is registered in the dictionary as an initial value in the same way as during encoding. First, read the first code (reference number) with Sl, and
Since 0DE is set as 0LDcode and the first code corresponds to one of the reference numbers of one character already registered in the dictionary, the character code (K) matching the input code C0DH is searched and the character K is output. Note that the output characters are FINcha for later exception handling.
Set it to r. Next, go to 82, read the next code, and set it to C0DE.INc
Set as ode. In S3, it is checked whether there is a new code, that is, whether the code input has ended, and the process proceeds to S4, where it is checked whether the code C0DE inputted in S3 is defined (registered) in the dictionary. Normally, the input code word has been registered in the dictionary in the previous processing, so the process advances to S5 and the character string code (ωK) corresponding to the code C0DH is read from the dictionary, and the character K is temporarily stacked in S6. , the reference number C0DE(ω) is changed to a new code C0DE, and the process returns to S5 again. S6
Repeat the steps recursively until the reference number ω reaches one character K, and finally proceed to 87 and change the stacked character to L in S6.
Pop up and output in IFO (Last In FartOuj) format. At the same time, at 87, a new reference number is added to the character string representing the set (ωK) consisting of the previously used code ω and the first character K of the character string restored this time, and the character string is registered in the dictionary. The LZW decoding process will be specifically explained as follows with reference to FIG. First, in Figure 11, the first input code word (INPUT C0D
E) is 1, - for the letters a, b, c already the reference number 1. 2. 3 is registered in the dictionary as shown in FIG. 10, so by referring to the dictionary, the character string a having the reference number matching code word 1 is replaced and output. Similarly, the next code word 2 is replaced with the character S and output. At this time, the character string ω that is a combination of the code word 1 processed last time and the first character of the character string just decoded is =
Add a new reference number 4 to 1b and register it in the dictionary. The third code word 4 is the character string 1b found by dictionary search.
is replaced with the character string ab and the character string ab is output. At the same time, the character string ω that is the combination of code word 2 processed last time and the first character a of the character string just decoded is = 2a (=ba)
is added with a new reference number 5 and registered in the dictionary. This process is repeated in the same manner. The LZW decoding shown in FIG. 11 involves the following exception handling. This exception handling occurs in the decoding of the sixth input codeword 8. Code word 8 is not defined in the dictionary at the time of decoding and cannot be decoded. In this case, the character string 5 is obtained by adding the first character of the previously decoded character string ba to the previously processed code word 5.
Exception processing is performed to find b and then replace it with 5 b=2 a b=b a b and output it. After outputting the character string, a reference number 8 is added to a character string 5b obtained by adding the first character of the currently decoded character string to the previous code word 5, and the result is registered in the dictionary. This exception handling is performed at 84 and S in the decoding process flow in FIG.
Finally, in step 87, the character string is output and the new character string is registered in the dictionary with a reference number added.
It will be held at 7. In addition, in the LZW decoding shown in Figure 8.11, we have explained the case where the dictionary is created in real time while decoding the code on the decoding side, but the dictionary created during encoding can be used as a copy on the decoding side as is. It may be encoded by doing this. In this case, exception handling on the decoding side becomes unnecessary. If LZW encoding is performed according to the procedure shown in the processing flow diagram in Figure 7, in the worst case, the entire dictionary will have to be searched every time a dictionary is searched for one character string, and the dictionary search will take time. There was such a problem. Therefore, in conventional dictionary search methods, open hashing or chain
ng) to increase processing speed. First, in a dictionary search using the intra-membrane hash method, when considering a set S consisting of multiple character strings, the storage position of the character string X in the set S can be determined from the address of the character string X itself. It has a mechanism that allows direct calculation, and allows for high-speed dictionary searches. Assuming that the storage location of the character string, that is, the hash table, is assigned addresses from 0 to -1, in the hash method,
Define one function h: s → (0, 1, ..., m-1),
Find the address of character string X in set S as h(x). This function is called a hash function, and the value h (x) is called a hash address of the character string X. The hash method is normally used when the size of the set S is much larger than the number m of addresses. However, no matter how the hash function is selected, a case may occur in which the hash addresses for different character strings xi and x2 of the set S match h (xi)=h (x2). This is called a collision, and one of the countermeasures against collision is the external hashing method (open hashing, or c
haining) is used. As shown in Figure 12, in the external hash method, a linked list is prepared for each hash address i indicated by an index (directory), and the hash address h(x) =
The character strings X of i are stored in order from the beginning of the linked list. Each linked list with the same hash address h(x) is called a packet. LZ using list structure of external hash method for dictionary search
A processing flow diagram of W encoding is shown in FIG. Also the 14th
The figure shows the structure of a dictionary memory according to the external hash method, and specifically shows the search procedure and registration procedure of LZW encoding using the tree structure of the encoded character string shown in FIG. 15 as an example. First, in FIG. 14, the dictionary memories include a first memory 100, a next memory 200, and an extension memory of the next memory 200.
mory) Consists of 300. Here, the first memory 100 has an index of the external hash method shown in FIG.
directory), and the next memory 200 is the first
This corresponds to rnextj in the linked list in FIG. 2, and the extended memory 300 corresponds to rnameJ in FIG. Also, the tree structure in Figure 15 is for characters. , K2. . K2゜2,..., 4. has already been registered, and the broken line 42 indicates a case where it is newly registered. In the process of FIG. 13, the hierarchy in this tree structure is represented by an i counter, and the number of characters in the same hierarchy is represented by a j counter. Therefore, the registered address of each character is expressed as ω. Now, the LZW encoding and registration by dictionary search according to the processing flow in Figure 13 when the character string "K10/K22. K32/K42" included in the registered tree structure in Figure 15 is input will be explained as follows. become that way. In FIG. 13, first, the following initialization process is performed in Sl. ■ Initialize the dictionary to include the first character. For example, if there are 26 alphabetic characters, the character code is directly registered as a hash address in the first memory shown in FIG. 14. In the case of FIG. 15, it means that the character KIO at the top of the tree is registered at address ω,0. (2) Set the current number of characters registered in the dictionary n to the number of characters registered in (2) above. In the case of 26 alphabetic characters, n
=26. ■Let the first character K input be the initial character string i. In the case of Figure 15, the first input character is . Therefore, the word initial character string i=1. In the following processing flow, the initial character string i will be described as a j counter. ■Initialize the dictionary search array to 0. That is, the search array for first, next, and extended memories is Ii+sl.
[1, Nmax], next [1, Nmax], E
Since it is expressed as XT [1, Nmax], set it to 0
Initialize to . When the initialization process of Sl is completed, the process advances to S2 and the next character "K2□" is read. Next, at 83, a check is made to see if there are any unprocessed characters. When all the processing is completed, the process proceeds to S16, where the code word code (ω) is output, and the processing ends. At this time, since there are unprocessed characters, the process proceeds to dictionary search steps 85 to S9. The dictionary search step begins with address ω in S5. Set the value of the initial word string i=1 at that time, and j
Set the counter to j=0. As a result, the first memory address ω1. =ω, 0 is generated. Next, at 86, the address ω of the first memory 100 is determined. If you read the contents of the address ω1. = ω21 is obtained, so
Set the i counter to i=2. Next, the process proceeds to S7, where it is checked whether or not i=0, and since i=2 at this time, the process proceeds to S8, where the address ω2. extended memory 3
00, the character rK2+J is read out, and it is determined whether it matches the input character "K22" obtained in S2. in this case,
Since the two do not match, proceed to 89, and at this time i
The counter value i=2 is set in the j counter to make j=2, and the address ω2. of the next memory 200 is set. j of address ω, = ω2□ stored in j counter i
=2. Therefore, the new address ω,=
ω2□ is created. Next, return to S7, check i=0, and at this time i=
2, the process goes to S8 again to read the registered character "K2□" from the extended memory 300 at the address ω2□ and determine whether it matches the input character "K22". At this time, since the two match, the process returns to 82 and the next character "K3□" is read. Thereafter, by repeating the processes 85 to S9 in the same manner, dictionary searches are performed in the order shown by the solid arrows in FIG. 14, and the search processing up to the already registered characters rK4+J is performed. When the search for the registered character "K4□" is completed and it is determined in S8 that there is a mismatch between the last input character 1 and 4□, i=2 is set in S9, and the address ω4. Since the content of the next memory 200 is 0, i=0 is set. Therefore, when returning to S7, it is determined that i=0,
Skip the dictionary search step and proceed to SIO, and retrieve the address ω indicating the character string ``K, ., 2□, 3゜J''.
3□ is output as the code word C0de (ω), and SL1~
Proceed to step 14 of dictionary registration. In the dictionary registration step, first, the currently registered character string n is set to n=i, that is, n=4, and n is further incremented by one. Then, the character "K4□" is registered at the address ω,=ω42 of the extended memory 300. Next, in 812, it is checked whether j=0 or not, and at this time, j=
Since it is 2, proceed to 814 and next memory 200
address ω4. Write the address ω4□ in which the characters "K42" are registered. On the other hand, if j=0 in S12, that is, if the state has shifted to registration in the first memory 100, the address ω of the first memory 100 in FIG.
01. As shown in ω2□ and ω32, the extended memory 300
Stores the character registration address. By registering the character "K42" in this character registration step, the next memory 200 and expansion memory 300 in FIG. 14 are stored at the address ω41.
ω42 is now in the registered state, and the address ω42 of the new character “K42” has been added to the tree structure shown in FIG. In addition, in FIG. 14, the address ω4. For convenience of explanation, these are shown redundantly for search and registration. When the dictionary registration steps SL1 to S14 are completed, S1
Set the character "K4□" registered in step 5 to the new initial character string 11, that is, the value of the i counter, and return to S2 again and use the characters rK and s2J as the top of the tree to perform a dictionary search for subsequent character strings. Transition.

[Problem to be solved by the invention]

As described above, in conventional LZW encoding, when encoding is performed by software by executing the processing flow shown in FIG. 7, dictionary search processing takes a lot of time.
The processing flow shown in FIG. 13 uses the external hash method to speed up the dictionary search. However, in dictionary searches using the external hash method, candidate characters are generated one after another, candidate characters are compared with input characters, and matches and mismatches are determined in a sequential manner. This accounts for 80% of the time, and there is a need for even higher speeds. In addition, since a list structure using an external hash method is used to read out candidate characters, there is little correlation between the storage address of the current candidate character and the storage address of the next candidate character, and they can be read out at any time. Therefore, it was not possible to read addresses first, and it was not possible to make the most of the performance of the elements that made up the dictionary memory. For example, when using DRAM as a dictionary memory, there is no continuity in addresses, so for example, column addresses (Row
Fix the row address ((Cotum
It has been difficult to perform high-speed reading such as in a page mode in which only AddressJ is changed. For example, in the case of Fig. 14, there is no continuity of addresses in the addresses ω3□ and ω33 of the next memory 200, so as shown in Fig. 16, the normal glued mode in which the column address and row address are specified individually each time is used. Therefore, there was a problem that speeding up could not be achieved. The present invention has been made in view of such conventional problems, and provides a dictionary search method for a data compression device that can shorten dictionary search time by enabling high-speed reading of dictionary memory using the list structure of the external hash method. The purpose is to provide

[Means for Solving the Problem] FIG. 1 is a diagram illustrating the principle of the present invention. First, the present invention divides encoded data into different subsequences, adds a different reference number to each subsequence, and registers it in a dictionary. The present invention is directed to a data compression apparatus that specifies and encodes matching subsequences using reference numbers, such as a data compression apparatus that performs LzW encoding. As a dictionary search method for such a data compression device, the present invention includes a first memory 100 according to the list structure of the external hash method and a next memory 200 having an extended memory 300, and stores an external hash address based on input data. a dictionary memory 20 whose concatenated addresses are partially composed of consecutive addresses of the next memory 200;
Expansion memory 3 that continuously generates addresses of the next memory 200 based on input data to match the input data.
The present invention is characterized in that it is provided with a dictionary search means 16 for searching candidate data of 00. Here, the dictionary search means 16 includes a pipeline control means 26 that performs a match check between input data and candidate data, presence or absence of candidate data, and reading of the next candidate data in parallel. Furthermore, the high speed page mode is used as the access mode for the dictionary memory 20. [Operation] According to the dictionary search method of the data compression device according to the present invention having such a configuration, in the dictionary search of LZW encoding, the index address of the dictionary memory having a list structure based on the external hash method is composed of continuous addresses. By doing this, once one hash address is determined, the next address can be predicted, which speeds up search access for candidate characters, enables encoding by high-speed reading of dictionary memory, and reduces encoding processing time. can.

【Example】

FIG. 3 is an embodiment configuration diagram showing an embodiment of a data compression device (encoding device) equipped with the dictionary search method of the present invention shown in FIG. 2; In FIG. 2, the original data 10 to be processed is a DMA
(Direct Memor7 Access) Input via the control circuit 12. MPU as a control means
I4 sets the input original data 10, the 1-character and the reference number of the previous character string in the multiple character reading circuit 18 of the dictionary search circuit 16, and then activates the dictionary search circuit 16. Thereafter, the dictionary search circuit 16 reads the candidate character of the character string extended by one character from the dictionary memory 20, the match check circuit 22 performs a match check (verification) between the input character and the candidate character, and the concatenation detection circuit 24 reads the candidate character. Detects the presence or absence of. The pipeline control circuit 26 reads the next candidate character from the dictionary memory 20 in parallel with the match checking circuit 22 collating the input character with the candidate character and the connection detection circuit 24 detecting the presence or absence of the candidate character. By performing pipeline processing in the pipeline control circuit 26 in this manner, search and collation processing for each of a plurality of candidate characters can be executed within the cycle time of the dictionary memory 20. Furthermore, the dictionary search circuit 16 is provided with a continuous address circuit 28, which generates continuous addresses.
A plurality of character reading circuit 18 is made to read out one-touch addresses and candidate characters registered at consecutive addresses in a dictionary memory 20. In encoding with the LZW code, a character string in the dictionary memory 20 that matches the maximum length is obtained. Therefore, by adding input characters, the character string is successively extended by one character, and when there are no more candidate characters, it is known that the character string has the maximum matching length. At this time, the character string up to the maximum match length is represented by a reference number using the address ω, and that reference number ω is transferred from the input/output boat 30 to the external compressed code word code (ω
). FIG. 3 is an embodiment configuration diagram showing the detailed configuration of the dictionary search circuit 16 of the present invention shown in FIG. 2 together with the dictionary memory 20. In FIG. 3, address register 18-1, register 18-2, and register 18-3 correspond to the multiple character reading circuit 18 of FIG. 2, and registers 22-1, . Comparator 22-
2 corresponds to the coincidence check circuit 22 in FIG.
4-1 corresponds to the connection detection circuit 24 of FIG. 2, and counter 28-1 corresponds to the continuous address circuit 28 of FIG. Next, dictionary search according to the embodiment of FIG. 3 will be explained with reference to FIG. 4, which is an explanatory diagram of the search procedure and registration procedure, and FIG. 5, which is an explanatory diagram of the tree structure showing the registration state of the dictionary memory 20. In the following explanation, memory address ω is the upper address 1.
ω1 by the 1st lower address j. shall be expressed as . Assume that the character string "K, ., 2°, K32. K4□" included in the tree structure of FIG. 5 is input as the original data 10. First, the MPU 14 sets the reference number ω1o corresponding to one character of 1o to the first character of the first input character string in the address register 18-1 that specifies the upper address.
Set 2° to the second input character in register 18-2. Next, the pipeline control circuit 26 is instructed to start up the dictionary search circuit 16. The pipeline control circuit 26 first sets the counter 28-1, which generates consecutive addresses, to 0, and then writes the consecutive addresses to the dictionary memory 20. counter 28-1
The content of is the lowest two bits of the address of the dictionary memory 20 (
Specify LSB). Therefore, address register 1
Contents of 8-1 ω1. - Due to ω1o, the upper address of the dictionary memory 20 is specified and the content of the counter 28-1 j=0
The address (ω+o+0) specified by the lower address of the dictionary memory 20 by
00 to read out ω21 and set it in the address register 18-1. Next, the contents ω2 of address register 18-1. The next memory 200 and expansion memory 300 of the dictionary memory 20 are accessed by the address (ω2100) with the upper address being the upper address and the contents of the counter 28-1 being the lower address, and the first
The concatenated address ω2□ of 21 for the th candidate character and 22 for the second candidate character is read out. 2+ is set in the register 18-2 for the first candidate character read, and the concatenated address ω2□ of 22 is set in the register 18-2 for the second candidate character.
Set to 3. Then, the comparator 22-2 compares the input character set in the register 22-1 with 2□ and the first candidate character set in the register 18-2 with 21 to determine whether they match or do not match. . Since the two do not match, it is determined that they do not match, and 2□ is read out as the next candidate character, but at this time the counter 28-
The address (ω21+1) of the next memory 200 is generated by incrementing the value of 1 by 1 to change only the lower address of the dictionary memory 20, and when the next memory 200 is accessed, 22 is read into the register 18-2 as the next candidate character. put out. At this time, the contents ω2□ of the address register 181 specifying the upper address remain unchanged. This operation is repeated in the same way, but the counter 28-
1 to generate consecutive addresses blindly,
Since the dictionary memory 20 is to be enlarged, this embodiment is designed to generate four consecutive addresses. For example, if the character code is 8 bits, an address exceeding 9 bits has no meaning. Therefore, once in four searches, the concatenated address ω11 obtained by accessing the first memory 100 is used instead of the continuous address of the next memory 200. That is, when the counter 28-1 generates the consecutive lower address "00°01.10.IIJ" four times while the upper address is fixed, it simultaneously switches to the next consecutive address "00".
Register 18 on the fourth access to register 18-3
-3 is stored in the first memory 100 is set in the address register 18-1. For example, the upper address ω of the next memory 200 in FIG.
9. For example, when the lower address is incremented by the counter 28-1, ω3110 (=ω3.) ω31+1 (=ω3□) ω31+2 (=ω33) ω31+3 (=ω34) are generated as consecutive addresses, and the fifth address is is the link address ω3. to the next consecutive address stored in the next memory 200. are successively generated, and the generation of consecutive addresses is repeated from the beginning as the upper address. When the input character and the candidate character match in the comparator 22-2 through such a dictionary search, the NOR circuit 2 simultaneously
4-1, the contents of register 18-3 (next memory 20
0's concatenated address) is all 0's, and the dictionary search is repeated until all 0's are found. If register 1
If 8-3 is all 0, it is detected that there are no more candidate characters to search for. In this case, the MPU 14 and the pipeline control circuit 26 terminate the search process of the dictionary search circuit 16, and output the address of the last matching candidate character in the dictionary search so far as the code word code (ω). In the case of Figure 4, input character rK4+j moves to next memory 2.
Since the contents of 00 are all 0, the dictionary search is ended at this stage, and the address (ω41+0) of the last matched candidate character rK4+J is output as the code word C0de(ω). Next, the MPU 14 inputs the last remaining input character "K4□"
The address (ω4°+1) is registered in the extended memory 300, and the address (ω41+0) of the next memory 200 is registered in the extended memory 300.
) After registering the connected address ω4□, a new dictionary search is performed using the input character rK42J as the initial character string i. In this way, in the present invention, candidate characters and concatenated addresses can be retrieved by continuously generating addresses, so that only the row address can be changed while the column address is fixed as shown in FIG. 6 as the dictionary memory 20. A high-speed page mode using continuous addresses can be used, and candidate characters and their concatenated addresses can be read out at high speed, so dictionary searches can be performed at high speed.

【Effect of the invention】

As explained above, according to the present invention, a linked list using the external hash method is constructed of consecutive addresses in a dictionary search for LZW encoding, so that once one address is determined, it is possible to start by predicting the address, and the dictionary memory For example, by realizing a high-speed page mode when using a DRAM, the performance of the memory element can be fully utilized and dictionary searches can be performed at high speed.

[Brief explanation of the drawing]

FIG. 1 is a diagram explaining the principle of the present invention; FIG. 2 is a configuration diagram of an embodiment of the invention; FIG. 3 is a diagram illustrating the configuration of an embodiment showing details of the dictionary search circuit of the invention; FIG. An explanatory diagram of the LZW code search procedure and registration procedure; Figure 5 is a tree structure diagram showing the dictionary registration contents of the present invention; Figure 6 is the DR when using the high-speed page mode of the present invention.
Timing chart of AM read mode; Figure 7 is a conventional LZW encoding process flow diagram; Figure 8 is a conventional LZW decoding process flow diagram; Figure 9 is an explanatory diagram of LZW encoding; Figure 10 is a dictionary configuration example Fig. 11 is an explanatory diagram of LZW encoding; Fig. 12 is an explanatory diagram of the list structure of the external hash method;
The figure is a flowchart of conventional LZW encoding processing using the external hash method; Figure 14 is an explanatory diagram of the LZW code search procedure and registration procedure of Figure 13; Figure 15 shows the dictionary registration contents of Figure 14. Figure 16 is a timing chart of the DRAM read mode in which the high-speed page mode cannot be used. In the figure, 10: Original data 12: DMA control circuit 14: MPU 16: Dictionary search means (dictionary search circuit) 18: Multiple character reading circuit 18-1: Address register 18-2.18-3: Register 20: Dictionary Memory 22: Coincidence check circuit 22-1: Register 22-2: Comparator 24: Connection detection circuit 24-1: NOR circuit 26: Pipeline control circuit 28: Continuous address circuit 28-1: Counter 30: Input/output circuit 100・First memory 200 Next memory 300; Expansion memory

Claims (3)

[Claims] (1) Divide the encoded data into different subsequences, add a different reference number to each subsequence, and register it in a dictionary, and match the input data with the maximum length among the subsequences in the dictionary. A data compression device that specifies and encodes a subsequence using a reference number, comprising a next memory (200) having a first memory (100) and an extended memory (300) according to a list structure of an external hash method, a dictionary memory (20) in which concatenated addresses of external hash addresses based on the input data are partially composed of consecutive addresses of the next memory (200);
) for searching candidate data in the extended memory (300) that matches input data; . (2) In the dictionary search method of the data compression device according to claim 1, the dictionary search means (16) performs a match check between the input data and candidate data, the presence or absence of candidate data, and the reading of the next candidate data in parallel. A dictionary search method for a data compression device, characterized in that it is equipped with a pipeline control means (26) for performing the following operations. (3) A dictionary search method for a data compression device according to claim 1, wherein a high-speed page mode is used as an access mode for the dictionary memory (20).