JPS60168659A - Letter output device - Google Patents

Abstract

PURPOSE:To send out many letter species by a method in which when font patterns of KANJI etc., to be sent out are not found in the font pattern RAM, transference is demanded on the host side and rewriting is dynamically made. CONSTITUTION:In case where the font pattern of letters, e.g., KANJI, etc., to be sent out is not found in a font pattern memory 224, the transference of the font pattern is demanded on the host side 202 by a controller 222. When the font pattern is received, the font pattern, together with an age information to show the time of writing to RAM provided in the memory 224, is written. In the determination of location in the RAM in which the pattern is to be written, location in the RAM is orderly investigated by hashing, and when the location is vacant or age information by age counter stored correspondingly to the location is older than a given value, the font pattern is written in there and corresponding age information is rewritten.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a character output device particularly suitable for displaying and printing a large number of character types such as Chinese characters.

[Technical background of the invention and its problems]

When designing a character output device that outputs kanji and processes the characters, it is necessary to use some method to extract the corresponding font pattern from the kanji code.
The font patterns that can be used must be kept in mind. If the total number of Kanji character types that may be used is several thousand at most, a simple technique can be used in which all necessary font patterns are written in a ROM (read-only memory).

However, the total number of tXi character types is more than 50,000, and visually there are 6. (In order to output the desired font pattern, it is necessary to increase the number of dots that make up the font pattern as much as possible. Therefore, depending on the required specifications for the J device, all the necessary font patterns may be stored in the ROM. If you try to write to the ROM, the required ROM capacity will become too large, which is impractical. Furthermore, even if all fonts and patterns can be stored in the ROM, the creation of new kanji and the commonization of misspellings and slang characters are still continuing, so it is important that all fonts and patterns will be in the ROM in the future. is not necessarily guaranteed.

To solve this problem, conventionally the character output & placement shown in Figure 1 is 1.
It was proposed. In Figure 1, character output &a ff
Font turn memory in fj l 04
O is made up of a ROM 112 of 11 characters and a RAM (readable memory including J) 114 of relatively small size (for example, capacity for a+i oo character types). Then, when starting up this character output device 1 l O4, the host 10
2 sends to RAM 114 via interface 106 font patterns that are not stored in ROM 112 and will be used later. Thereafter, the character code indicating the kanji to be output from the host is loaded into the character code buffer 108 of this character output device. After that, the 1lI11 control 01f in the character output device 104 (not shown)
Accordingly, the font pattern corresponding to the character code in the character code buffer 108 is stored in the ROMI 2 or RAM 1 in the font pattern memory 110.
14, is written into the output memory 116, and is output to the CR1 display device 118, printer 120, or the like.

However, character height J-&C as shown in Figure 1
4 also has the following drawbacks. That is, in this character output device, all font patterns that will be needed from now on are stored in the RAM 114 at the time of startup, but the capacity of the RAM 114 cannot be made that large. Therefore, which font pattern to store in the RAM 114 for each application of the character output device must be managed by the user (or the creator of the system for that application). Tama Negative II! is large. In addition, depending on the necessary work, RA
A font pattern that goes around the size of M114 may be needed. This case could not be handled by the conventional character output device shown in FIG.

In addition, the host 102 has another character output device f l 44.
164 is connected, each character output device 9 has a different character set.
Character output device 104.144 connected to Jff from 2
．． This causes inconvenience in sending messages containing f5t'j: to the 164, or in exchanging data between character output devices.

[Inn9 of the Invention] It is an object of the present invention to eliminate the drawbacks of the above-mentioned conventional techniques, and to provide a character output &1α that can output an extremely large number of character types and is easy to handle for users.

[Keystone of invention]

In order to achieve this object, the character output device z of the present invention transfers the font pattern when the font pattern of a character such as a kanji to be output cannot be found in the font pattern memory. Request to host side. When a font pattern is received from the host, the font pattern is written into the RAM provided in the font pattern memory along with age information indicating the writing point. R to write font pattern
In determining the location in the AM, locations in the RAM are checked one after another by hashing. If the location is empty or the age information stored corresponding to that location is older than I91 ((fi), a font/turn is written there.・Rewrite the required age t/1 report to the latest one every time the pattern is used.

(Embodiments of the invention) Hereinafter, embodiments of the present invention will be described in detail based on the drawings.

FIG. 2 is a schematic block diagram of a character output device according to the present invention. In the figure, when a character code representing a Chinese character or the like to be output is given to a character output device 204 connected to a host 202, the character code is stored in a character code buffer 208 via an interface 206. The character code may also be input from the keyboard 226 of the character output device &. The control unit 222 reads the character code from the character code buffer 208, searches the ROM 212 and RAM 224 using the character code, and writes the corresponding phonograph turn to the output memory 216. This causes the CRT disk drive'!f&218, printer 220, etc. Output to is performed.

If the font pattern required for the selected character code cannot be found in either the ROM 212 or the RAM 224, the ili control + 311222 requests the transfer of the font pattern to the host 202 via the interface 206. . This transfer request is temporarily held in the interface 206 and taken in by polling performed from the host 202 to the character output devices 1204, 244, etc. Note that other means such as transmitting this transfer request to the host 202 using an interrupt may also be used.

When a font pattern is received from host 202, it is written to RAM 224 as well as output J memory 216. Note that the character code changed from the host etc. is once written to the character code buffer 208, and then stored in the ROM 212, R
Even if a font pattern corresponding to AM224 does not exist, the processing of that font pattern is postponed and a series of character codes are received from the host and the corresponding font pattern is written to the output memory without interruption. .

The RAM 224 has a character code section 3 as shown in FIG.
12 and an age information section 314, and a font RAM 304 having locations corresponding to people in each location where the numbers O to η-1 of the hash table 302 are filled out.

Furthermore, an age counter 306 is provided which is updated every time the output memory 216 is rewritten. Font RAM
When a font turn at a location in 304 is used, 6LJ of age counter 306 is written to the corresponding location in age bv report 314.

In order to simplify the explanation, in FIGS. 48 to 4c,
It is assumed that the output memory 21B displays 3 rows and 2 columns on the CRT data display device 21g, and the RAM 302 stores font patterns for three character types in locations θ to 2. Furthermore, the explanation will be made assuming that the ROM 212 is not provided.

However, the range in which R is reduced is displayed.

Now, with the display shown in Figure 4A being displayed, fj
As shown in the S4B diagram, the character code buffer is 16! ] is inserted by the host I-202. At this time, the controller 1-NOUIIfM1222 is stored in the RAM22.
New character code 1' f in hash table 302 in 4
Xl", but as you can see from Figure 4B, it cannot be found here.

Therefore, as shown in FIG. 4B, the control section 222
The location where the font pattern l' M J in the memory 216 should be stored is temporarily determined using a special pattern (it may be left blank). Thereafter, the font pattern 1 obtained from the host 202 is written to the output memory 216 and font RAM 304, and the age counter 306 is incremented. Font RA
Writing to M304 and accompanying hash table 30
Writing to 2 is performed as follows.

The following will explain the rewriting operation of the RAM 224 shown in FIG.

Character codes given from the host 202, keyboard 226, etc. are stored in the register 402.

If there is no font pattern corresponding to the given character code in the ROM 212 and RAM 224, the font pattern obtained by the transfer request to the host 202 is
The 'turn' is stored in register 410. Note that the initial value Mi of the age t+'J information Q in the register 406 is set to the current value of the age counter 306 by -i ships.

Write down these font patterns and character codes.
In order to determine the location of the font RAM 304 and hash table 302, the character code in the register 402 is first hashed in the hash section 404. Various specific methods of hashing are known, but the 4th
In the example shown in Figures 8 to 4C, where 7J
(varies depending on the capacity of AM 304 and hash table 302).

Hashing result W-t index Y as hashing table 3
Subtract 02 and get the age information q in that location. Further, if the location is empty, the operation of the hash section 404 is stopped by No. 178E. Then, the character code in register 402 and the value in age counter 306 are written to the location in hash table 302 indicated by index Y. At the same time, the font pattern in register 410 is written to the corresponding location in font RAM 304.

If the location read out by the inchfus Y is not empty, the comparison unit 408 compares the read age information q with the current count 613C of the age counter 306, and calculates the difference between them as determined by a predetermined value (fuM). If so, a write command W is given to the RAM 224, and the same write operation as in the case where the read location is empty is performed.

If the above difference is less than rll+M, hash portion 4
04 compares the number of rehashings performed so far with the upper limit number N set in advance, and if it is smaller than N,
Compare the engineering information Q in the register 406 with the read age information q. If q<Q (that is, if the read age information is older), the register 406
Age information 6 and inchphus P in 6 are respectively age tN
Replace with information q and influenza Y. Note that if it is Q10, register 4O6 is left as is. Thereafter, the hash section 404 performs rehashing, gives a new index Y, and reads the hash table 302 again (if not in FIG. 6, the process returns to (1) at the top of the flowchart). Rehashing may be performed, for example, by adding 1 to the current index Y (or the location to be read next may be stored in the hash table).6 While repeating the loop described above, +lj When the number of times of hatching reaches the limit N, this loop is removed, and the comparator 408 issues a blue filling command W, and also stores the inch P in the register 406 as the inch Y in the RAM 224.

As a result, the same four steps fj as in the case where the read location is empty are carried out.

Incidentally, the operation of the character height J apparatus of the present invention described above may be controlled using a processor. Further, although a large number of memories and counters are shown in the embodiments, a series of memory spaces may be appropriately divided and allocated to each of these, for example.

Furthermore, the characters to be output are not limited to Kanji characters, and can be applied to output of many character types and arbitrary character sets with complex font patterns.

Furthermore, various other methods can be arbitrarily used for hashing.

Furthermore, in the present invention, some character font patterns may be sent in advance from the host to a part of the RAM when the character height device is started up.

Moreover, it is not necessary to use ROM.

〔Effect of the invention〕

According to the present invention, the RAM in which the font pattern is stored is dynamically rewritten, so that the font pattern of the character output device can be used without increasing the capacity of the font pattern. It is possible to obtain flexible character output and placement that can output characters of Rewriting is done.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an example of a conventional character output device N, FIG. 2 is a schematic block diagram of a character output device according to the present invention, and FIG. 3 is a block diagram showing an example of a conventional character output device N. FIG. Memory configuration diagrams, FIGS. 48 to 4C are diagrams showing examples of rewriting the font/pattern memory in the present invention, and FIG. The diagram of the part related to f1, FIG. 6, is "Flowchart 1" showing the writing/rewriting operation. 202, host, 204.2441 character output device6. 206・Interface, 212: ROM, 216 Output memory, 218: CRT display device6. 220 Printer, 222 ilj control section, 224:R
AM, 226 keyboard, 302, hash table, 304: font RAM, 306, age counter, 312-character code section, 314: Age Information Department, applicant Yoko M. Kawa Tsuguo Figure 4C

Claims (1)

[Scope of Claims] A character output device connected to a second place device and outputting a font pattern of a given character code, comprising: a font pattern memory having a plurality of locations for storing font patterns; and,
++i an output memory for storing the font pattern read from the font pattern memory;
At least a portion of the location of
tf function, each location of the +1+I entry contains age information indicating the point in time when the 1IiJ font pattern in the location was most recently used; 4) 1) Search the location of the writable part in IIJ based on f1 which hashed the 11th character code, and find the location where the writing is to be performed based on the free status of location 11h and/or the age information in I+il. A character output device characterized by: