JPS6272061A - Dictionary control system in document preparing processing - Google Patents

Abstract

PURPOSE:To execute dictionary access of a converting module at a high speed by storing the dictionary data in the secondary memory medium to a writable read memory part, and further, storing the dictionary data stored in the read memory part to a packet memory part in the prescribed number of data unit including the word to be retrieved. CONSTITUTION:A read memory part is set to memory capacity (for example, 1k byte) in which the memory retrieving time is minimized. On the other hand, the area of a packet memory is set so that dictionary data from the read memory part can store plural packets for the prescribed number of the data (for example, 256 byte as one packet). The read memory part and the packet memory part are controlled by the read memory control part and the packet memory control part respectively, and by these memory part and memory control part, the dictionary control is executed. In KANA/KANJI converting operation, when the input action is executed, a dictionary control part 13 retrieves whether or not the retrieved data exist in a packet memory part 15.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a dictionary management system in document creation processing such as a word processor, and particularly to a dictionary management system that allows high-speed and efficient access.

Conventional technology Kana-Kanji conversion technology has been established as an essential man-machine interface for office automation equipment such as Japanese word processors, and a free-writing Kana-Kanji conversion method that requires almost no conversion key operations is being put into practical use. be. In this type of kana-kanji conversion using the free literary ability, the number of times the dictionary is accessed due to keyboard input (that is, the number of times the dictionary is searched) is extremely large, thousands of times the number of accesses to a normal dictionary. .

Therefore, in order to achieve such a huge number of accesses at high speed, conventionally the main memory was configured with RAM, and dictionary information was loaded onto this RAM from a secondary storage medium such as a floppy disk. The created large-capacity dictionary is resident in the main memory to achieve high-speed kana-kanji conversion with the above-mentioned large number of accesses.

Problems to be Solved by the Invention However, in such a conventional dictionary management system, a large capacity storage medium of 200 to 4 QoK bytes is required in order to expand the dictionary data on the RAM that constitutes the main memory. There is a problem in that it is necessary and costs a lot of money. In addition, in devices such as IWPs (integrated work terminals), the main memory often occupies a considerable area for processing images and graphics, so it is necessary to reduce the dictionary area in order to perform these processing tasks. Although this is desired, there is a problem in that the conventional example, which requires the use of the above-mentioned RAM, cannot meet this demand.

The present invention was made in view of such conventional problems, and its purpose is to
Another object of the present invention is to provide a dictionary management system for document creation processing/stem with limited main memory.

Means for Solving the Problems In order to achieve the above object, the present invention provides a read memory unit that can rewrite dictionary data in a secondary storage medium in predetermined capacity units defined by sounds entered with a keyboard or the like. Store it in
Furthermore, the dictionary data stored in the read memory section is stored in the packet memory section in units of a predetermined number of data including the term to be searched, and the kana-kanji conversion section searches the packet memory section based on the input from the input section, and If the data to be searched does not exist in the packet memory section, the read memory section is searched and the dictionary data is transferred to the packet memory section, and if the data to be searched does not exist in the read memory section, the specified data is retrieved from the secondary storage medium. The gist of this system is that when the dictionary data for the content is transferred to the read memory section, it is also transferred to the Nihacket memory section, and the packet memory is updated in units of the predetermined number of data using the first-in, first-out procedure. .

The active read memory section is set to a storage capacity (eg, 1 byte) that allows memory search time to be minimized. on the other hand,
The area of the packet memory is set so that a plurality of packets of dictionary data from the read memory section can be stored for each predetermined number of data (for example, 256 bytes per bucket). The read memory section and the packet memory section are managed by a read memory management section and a packet memory management section, respectively, and dictionary management is performed by these memory sections and the memory management section. In the kana-kanji conversion operation, when an input operation is performed, the searched data is sent from the kana-kanji conversion section to the dictionary management section. In the dictionary management section, first, the packet memory management section searches whether the data to be searched exists in the packet memory section, and if the data exists, this data is read out and sent to the kana-kanji conversion section.

If the searched data does not exist, the read memory management section searches whether the searched data exists in the read memory section, and if it does exist, one bucket of dictionary data including this searched data is transferred and written to the packet memory section. At the same time, the data to be searched is read out. When the searched data does not exist in the read memory section, a search is performed on the secondary storage medium, and dictionary data, for example, one byte worth of dictionary data including the searched data, is written to the read memory section. One bucket of dictionary data including the data to be searched is transferred and written from the memory section to the packet memory section, and at the same time, the data to be searched is read out. Writing of dictionary data to the packet memory section is carried out in units of 1 batch, according to the first-in, first-out method.
K gives the Kana-Kanji conversion system a learning function.

Embodiment FIGS. 1 to 4 are diagrams showing an embodiment of a system for executing a dictionary management method in document creation processing according to the present invention. The dictionary tube salt system according to this embodiment includes an input section 11 for inputting data for document creation, a kana-kanji conversion section 12 for performing kana-kanji conversion based on the input data, and a kana-kanji conversion section 12 for converting data into kana-kanji based on commands from the kana-kanji conversion section 12. A dictionary management unit 13 that performs searches, dictionary data necessary for dictionary management,
It consists of a secondary storage medium that stores other data, and an output section 19 that sends out the results of kana-kanji conversion as output data.

The input section 11 (2) is composed of a data input mechanism such as a keyboard or a document reading device, and is configured to input data necessary for creating a target.

The dictionary management unit 13 includes a rewritable packet memory unit 16 that divides and stores all dictionary data in packet units each consisting of a predetermined number of bytes, and an input unit that stores all dictionary data f′ stored in a secondary storage medium. A rewritable read memory unit 17 that stores data in predetermined capacity units defined by data, a packet memory management unit 14 that manages the packet memory unit 16, and a read memory management unit 16 that manages the read memory unit 17. Become. In this embodiment, the bucket memory section 15fi has a storage area divided into a plurality of packets (16 packets in FIG. 1) as 266-byte t1 packets. 16 is searched, and data is written into the packet memory section 16 in accordance with the first-in, first-out procedure in units of packets. The read memory section 17 is 1
It is composed of a rewritable storage medium having a capacity of 1 byte, and among the dictionary data stored therein, 266 bytes of data including search data are transferred and written as one packet to the packet memory 15. As shown in FIG. 4, the packet memory management unit 14 includes a queue management circuit 31, a packet memory write circuit 32 that performs all writing of dictionary data to the packet memory unit 16, and and a packet memory read circuit 33 for reading data to be searched. The deep read memory management section 16 includes a read memory management circuit 34 , a read memory read circuit 35 that reads dictionary data from the read memory section 17 , and a read memory write circuit that writes dictionary data to the read memory section 17 . It consists of 36. These circuits and each circuit and the memory sections 16 and 17 are connected by control signal lines, and data is written into and read from the respective memory sections 15 and 17 in accordance with each command signal.

Furthermore, a recording medium such as a floppy disk is used as the secondary storage medium 18, and dictionary data is stored in this secondary storage medium in a format as shown in FIG. 2, for example. According to this embodiment, each independent word ("kanji") is the minimum unit of dictionary data.

A number indicating words such as "conversion" and "terminal" is assigned to each record 23 (represented by RO, R1, R2, etc. in FIG. Dictionary data is divided into pages by a predetermined number of records 23 (corresponding to one packet) or less (represented by PO, P1, P2, etc. in Figure 2 a), and the dictionary data is organized and stored in page units. are doing. On the other hand, an index section 20 divided into a plurality of segments 22 is provided in the main memory of the kana-kanji conversion section 12 of the document creation processing system, and each segment 22 of this index section 2o and each page in the secondary storage medium 18 are There is a one-to-one correspondence between them. In order to take such measures, one segment 22 of the index section 20 is composed of a heading section 222L and a pointer section 22b, as shown in FIG. In addition, the record 23 assigned to each term in the secondary storage medium 18 is as shown in FIG.
Part-of-speech information part 23 representing the part of speech of the term stored in 3? L
, a reading information section 23b representing the reading of the term, and a kanji notation section 230 representing the kanji notation of the term.

In such a configuration, when a document creation process is performed, the data input from the input unit 11 is decoded by the kana-kanji converter 12, and the kana-kanji converter 12 converts the data into kanji based on the kana sounds. 13 issues a data search command (signal line 1) In the dictionary management section 13, the command signal from the kana-kanji conversion section 12 is first received by the queue management circuit 31, and the data to be searched corresponding to the command is stored in the bucket memory section 1.
The bucket memory read circuit 33 determines whether the bucket memory exists in
Search using . If the searched data is found in this search, this data is transferred from the packet memory section 16 to the kana-kanji conversion section 12 (data bus 6), where it is converted into kana-kanji and outputted from the output section 19. On the other hand, when the data to be searched does not exist in the packet memory unit 15, this fact is sent from the queue management circuit 31 to the read memory management circuit 3.
4 (signal line 2).

The IJ-to-memory management circuit 34, which has received the signal from the queue management circuit 31, uses the all-read memory reading circuit 36 to search whether or not the above-mentioned data to be searched exists in the read memory section. In this search, if there is the data to be searched, the read memory reading circuit 36 and the packet memory loading circuit 32 operate simultaneously, and 266 bytes (1 packet) of dictionary data including the data to be searched are read from the read memory section 1. is transferred to and written into the packet memory unit 15 (data bus 5). In the packet memory section 15, the dictionary data transferred from the read memory section 17 is written in a first-in, first-out procedure, so that the oldest data with the least number of accesses is replaced by the newer data. Then, the searched data is read out from this newly written dictionary data and transferred to the kana-kanji conversion section 12, where kana-kanji conversion is performed. On the other hand, if the search target data does not exist in the search for the read memory section 17, the search for the secondary storage medium is performed (signal line 3).
.

The storage location of the data to be searched is determined based on the information in the index section 20 provided in the main memory of the ephemeral Kanji conversion section 12 for the secondary storage medium. Then, one byte worth of dictionary data including the data to be searched is transferred to the read memory section 17 and written (data bus 4) o In the read memory section 15, all of the dictionary data previously stored is newly transferred. It will be replaced by the incoming dictionary data. Next, in the same manner as described above, data is transferred and written from the read memory section 17 to the packet memory section 16, and the searched data is transferred from this packet memory section 16 to the kana-kanji conversion section 12, where the kana-kanji conversion is performed. Q. As described in detail, according to the present invention, access to a secondary storage medium such as a floppy disk is performed when independent word data, which is data to be searched, is stored in either the packet memory section or the read memory section. Since the packet memory unit is managed only when the dictionary does not exist in the dictionary and the packet memory section is managed according to the frequency of use, dictionary access by the conversion module is speeded up. Furthermore, since this speeding up of dictionary access can be achieved without incorporating a large amount of memory into the document creation system, it is possible to introduce Kana-Kanji conversion into the document creation system, which has excellent operability while reducing costs. You can get the effect of getting.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a document creation processing system that implements the dictionary management method of the present invention, FIG. 2 is an index correspondence diagram between the kana-kanji conversion unit and the secondary storage medium in the document creation processing system, and FIG. 3 4 is a storage state diagram of independent word data in the secondary storage medium, and FIG. 4 is a diagram of the configuration and signal flow inside the dictionary management section. 11... Input unit, 12... Kana-Kanji conversion unit, 13... Dictionary management unit, 14... Packet memory management unit, 15... Packet memory section, 16...Read memory management section, 17...
...Read memory section, 18...Secondary storage medium,
19...Output section, 20...Index section, 21...Packet. Name of agent: Patent attorney Toshio Nakao and 1 other person 1st
Figure 2? Figure 3? 3

Claims (1)

[Claims] The dictionary comprises an input section, a kana-kanji conversion section, a secondary storage medium storing dictionary data to be converted into kana-kanji, and a dictionary management section provided between the kana-kanji conversion section and the secondary storage medium. The management section includes a read memory section that temporarily stores the dictionary data in the secondary storage medium in a predetermined capacity unit, and a part of the dictionary data stored in the read memory section that stores the data to be searched in another capacity unit. a packet memory section to
It is composed of means for managing the writing and reading of data in the read memory section and means for managing the writing and reading of data in the packet memory section, and searches the packet memory section based on input from the kana-kanji conversion section, and searches the packet memory section based on the input from the kana-kanji conversion section. If the data to be searched does not exist in the read memory section, the dictionary data in the above capacity unit containing the data to be searched is searched, and if the data to be searched does not exist in the read memory section, the above data is transferred from the secondary storage medium. A dictionary management method for document creation processing, characterized in that dictionary data in units of capacity is transferred to a read memory section and transferred and written to a packet memory section, thereby updating the packet memory section.