JPS62102356A - Chinese character conversion system - Google Patents

Abstract

PURPOSE:To easily make corrections by a less procedure by checking which section of a character string a character specified with a cursor is in and reconverting the character string into KANJI (Chinese character) when a KANJI mixed character string which is converted at a time has a part different from words that a user requires. CONSTITUTION:A KANA (Japanese syllabary) document which is inputted as rendering data is analyzed grammatically and sectioned into characters or character strings regarded as words with a conversion key, and then converted into KANJI. Then when a character in the KANJI mixed character string after the conversion is indicated with the cursor and the conversion key is pressed, which section the character belongs to is discriminated and the conversion result before the corresponding section is determined and the part including the section and succeeding characters are restored to the starting KANA character string. Then, the KANA character string is analyzed grammatically again and converted into KANJI.

Description

[Detailed Description of the Invention] [Summary] In a conventional Japanese word processor that adopts a method of displaying a sentence of arbitrary length input as a reading in kana characters and converting it all at once into a string of characters mixed with kanji. When performing kanji conversion, if there is a part of the converted character string mixed with kanji that differs from the user's intention, even if it is a part of the character string that is to be deleted, use the "cancel key" etc. There was a problem in that the operability was poor because the entire inputted text had to be converted back to kana characters by pressing the button, and then divided and converted. In order to solve these conventional problems, the present invention, when there is a part of a character string that has been converted into kanji that differs from the intended part, uses a cursor to indicate the position of the character string and to instruct the conversion again. This paper discloses a control technology that can re-analyze the original kana character string corresponding to the position after the position and re-convert the kanji characters based on this analysis.

[Field of Industrial Application] The present invention relates to the control of a word processor that displays sentences of arbitrary length input as readings in kana characters and converts them all at once into a string of characters mixed with kanji. The present invention relates to a control that allows easy correction when a converted character string contains a part that differs from the user's intention.

[Prior art] Sentences of arbitrary length input as readings using keyboard input, voice input, etc. are displayed in kana characters on the screen of a display device, and these are collectively converted into characters mixed with kanji. There is a Japanese word processor that converts into columns. In such Japanese word processors, some of the sentences that have been converted into character strings mixed with kanji may contain homonyms that are different from the user's intention, or may have a meaning different from the user's intention due to incorrect grammatical analysis. It often happens that it becomes a sentence.

Conventionally, in cases like this, even if the erroneous part is only a small part of the entire text entered, by operating the "cancel key" etc., the entire text entered at that time can be reset to the original reading. The solution is to convert it back to the corresponding "kana character string" and convert it into several parts.

[Problems to be solved by the invention] As mentioned above, Japanese word processors that input sentences of arbitrary length at a time and then convert them into kanji all at once limit the input to single words. , it can perform conversions much faster than conventional Japanese word processors that perform kanji conversion each time, and can be expected to improve repeatability due to its ease of handling.
It has become popular recently.

Then, when the inputted Kana character strings of considerable length are converted all at once, as long as they are converted into the kanji-mixed character string that the user expected, the user can input the previous words in units of words. It is much faster and easier to operate than a system that converts kanji each time, but it does not always convert as intended each time you input it, and as mentioned above, it often A character string mixed with kanji that is different from the user's intention will appear.

At this time, even if the erroneous part is only a small part of the entire text input, by operating the ``cancel key'' etc., the entire text input at that time can be changed to the original reading. The solution is to return to the corresponding “kana character string′” and convert it into several parts, so in such cases, it is easier to input each word and convert it to kanji each time. It had the drawbacks of being slow and cumbersome to operate.

In view of these conventional shortcomings, the present invention has been developed to solve the problem of inputting sentences of arbitrary length and converting them all at once into a string of characters mixed with kanji, which may contain parts that are different from the user's intention. The purpose of this invention is to provide a method that can be easily corrected with a few steps even if there is a problem.

[Means for Solving the Problem] According to the present invention, this purpose is to display sentences of any length input as reading in kana characters and to collectively display them. In a Japanese word processor that uses a method of converting into a character string mixed with kanji, the above-mentioned kana character string is analyzed, separated into characters or character strings that are considered to be parts of speech, and kanji conversion is performed based on the separation, and after the kanji conversion. Each time the cursor is placed to specify a character in the string and a conversion is indicated,
Checking which of the delimiters of the character string the character specified by the cursor is in, and determining the result of the kanji conversion that has already been performed for the character string before the delimiter to which the character belongs, This is achieved by a kanji conversion method characterized by parsing the kana character string of the delimiter and subsequent character strings again, delimiting them into characters or character strings that are considered to be parts of speech, and performing kanji conversion based on the delimiters. .

[Example] FIG. 1 is a flowchart showing the control of one embodiment of the present invention.

In FIG. 1, a sentence of arbitrary length input as a reading from a keyboard or the like of a Japanese word processor is stored in a "reading area" on the memory as a JEF code. When the "conversion key" is pressed, it is grammatically analyzed and divided into characters or character strings that are considered words, and the number of characters in each division is stored. Then, for each segment, the contents are converted into kanji using a dictionary. As a result, the original kana character string becomes a character string mixed with kanji. At this time, the number of characters for each break (which differs from the number of characters in the above case due to the inclusion of kanji) is stored.

When the cursor points to any character in the kanji-mixed character string after conversion and the "conversion key" is pressed, the function identifies which delimiter the character pointed to by the cursor belongs to. , determines the conversion result before the corresponding break, and restores the break and the following as the original kana character string. At this time, the position of the character string delimiter can be easily identified because the number of characters for each delimiter is stored as described above.

For parts that have returned to kana character strings, the grammar is analyzed again, and based on this, the corresponding kana character strings are re-separated and converted into kanji.

With this kind of control, the user can easily convert the 6 kanji character strings by modifying and confirming them one after another, without having to cancel all the conversion results of a character string that has only been entered once. Needless to say, if the conversion is incorrect from the beginning, you can call up the next candidate by pressing the "conversion key" again, as in the past.

Figure 2 is a diagram illustrating an example of reconversion of an actual sentence, and the result of batch conversion after inputting ``Beads that fall in two and hang around the neck'' is ``Beads folded in two and hung around the neck.'' It shows the process of modifying this to ``a rosary that is folded in half and hung on the wrist.''

In Figure 2, (a) is the reading area (maximum 40 characters of JEF code), (b) is the area that stores the number of characters in each segment of the grammar analysis result (maximum 40 segments), and (c) is the clause (40 segments maximum).
Table (d) shows the conversion result storage area (up to 100 characters of JEF code), and the numbers 0 to 39 or O to 99 at the top of each figure are byte addresses, and the number of characters attached to the address. The numbers in the sections indicate the number of characters for each section.

Then, as shown in (,), when ``Yuzu that falls in two and hangs around the neck'' is input, this is converted to ``Beads that are folded in two and hung around the neck'' as shown in -(d). It shows the relationship between each delimiter (indicated by a dotted line) and its number of characters.

In the same figure (d), for example, when the cursor is placed at the position of "te" and the "conversion key'° is pressed, the number of characters for each clause shown in (c) Check using a table. In this case, (a')
As shown in the figure, it becomes ``juzu on the neck.'' This is grammatically analyzed, separated into ``tekubi/ni/kake/juzu'' and the number of characters in each separation is changed as shown in (b').

The result of reconversion is ``a rosary that is folded in half and hangs on the wrist'' as shown in (d'), and the number of characters for each break after this change is as shown in (C'). Stored in a character count table for each character. The arrows in (a') to (d') indicate the positions of reconversion.

- [Effects of the Invention] As explained above, according to the method of the present invention, a sentence of any length input as a reading using input from a keyboard or voice input can be displayed on the screen of a strawberry display device. In a Japanese word processor that displays kana characters and converts them all at once into a string of characters mixed with kanji, even if the conversion is not what the user intended, it is possible to easily correct it. Therefore, it is possible to easily create a Japanese word processor with excellent operability that takes full advantage of the advantages of the batch conversion method and can create efficient sentences, so the effect is great.

[Brief explanation of drawings]

FIG. 1 is a flowchart showing control in one embodiment of the present invention, and FIG. 2 is a diagram illustrating an example of re-conversion of an actual sentence. Agent: Patent attorney Sada Igeta. ζ−Sino Figure ゛)

Claims (1)

[Claims] A Japanese word processor that uses a system that displays sentences of arbitrary length input as readings in kana characters and converts them all at once into a string of characters mixed with kanji, analyzes the above kana character string and considers it as a part of speech. Separate each character or character string, perform kanji conversion based on the delimiter, and each time the cursor is placed to specify a character in the string after kanji conversion, the cursor Check which of the character string delimiters the specified character is in, determine the result of the kanji conversion that has already been performed for the character string before the delimiter to which the character belongs, and A kanji conversion method characterized in that the delimiter and the subsequent kana character string are analyzed again to delimit characters or character strings that are considered to be parts of speech, and kanji conversion is performed based on the delimiters.