CN106372065B - A method and system for developing a multilingual website - Google Patents

Abstract

The invention relates to the technical field of natural language processing, in particular to a method and system for developing a multilingual website. The method for developing a multilingual website includes: step a: developing a static webpage of the multilingual website; step b: calling a machine translation interface to perform multilingual translation processing on the Chinese data dynamically added in the multilingual website; step c: reading Get translation data, load and render the dynamic web page of the multilingual website according to the translation data. The present invention adopts machine translation and manual intervention to correct processing methods, greatly reduces translation errors, and makes the display effect of web pages more accurate; by selecting the Unicode encoding format of utf-8, the garbled characters generated during web page rendering are avoided; The caching mechanism solves the problem of resource consumption and loading delay caused by the need to re-call the machine translation interface each time during the real-time translation loading process, while reducing manual intervention.

Description

A method and system for developing a multilingual website

technical field

The invention relates to the technical field of natural language processing, in particular to a method and system for developing a multilingual website.

Background technique

With the rapid development of Internet commercialization, a large number of e-commerce websites have emerged, and the market competition has become increasingly fierce. In recent years, China's e-commerce has developed rapidly, and its application in various fields has been continuously expanded and deepened. The transaction volume has hit new highs, which has driven the vigorous development of related industries.

We know that Uyghur is an ancient language with a long history. Currently, there are many books and documents written in Uyghur, and there are many historical materials. It stores thousands of Uyghur language and life information, and its historical significance and cultural value are precious. Therefore, folk information processing technology is closely related to the future development prospects of Uyghur language. At present, with the gradual improvement of the culture and knowledge level of the Uyghur people, the number of people who can create Uyghur language web pages is also increasing. Many people or groups have built various types of Uyghur websites to disseminate various information. Like ordinary Chinese websites, these websites provide news browsing, information download and other functions. It is not the same, which leads to the situation that Uyghur web pages have been in a state of incompatibility with each other. Most of the information of Uyghur web pages cannot be shared. At the same time, the conversion between different codes has to consume a lot of work time and scientific research materials. .

Xinjiang Uygur Autonomous Region is a multi-ethnic and multi-lingual place. E-commerce shopping has become a popular trend. The success of Taobao confirms that this trend will continue to be popular. However, most shopping platforms in Xinjiang are ordinary websites in Chinese. Most Uyghur compatriots who are not familiar with Chinese are difficult to use, so a standardized Uyghur-Chinese bilingual shopping platform is urgently needed. Then, to develop a standardized Uyghur version of the e-commerce platform is not a simple Uyghur-language version of a static web page, but a perfect shopping mechanism, which requires real-time dynamic management, dynamic additions, deletions, changes, and searches, which cannot be satisfied by manual translation. The demand for this kind of dynamic change of massive data, therefore, we need the dynamic change of the machine translation assistance platform.

Machine translation is the process of using computers to convert one natural language into another natural language. Since the development of machine translation, a variety of machine translation systems based on different principles have emerged. In general, machine translation systems can be roughly divided into four categories: rule-based machine translation, instance-based machine translation, statistical-based machine translation, and hybrid machine translation. Different machine translation systems have their own strengths. For example, rule-based machine translation systems are good at translating sentences that conform to rules, and the translation quality is high; statistics-based machine translation systems are versatile and automatically learn language knowledge from corpora.

Relevant references on Uyghur-Chinese machine translation include:

[1] Lan Boxiong, Zheng Xiaona, Xu Xin. Supply Chain Management in the E-Commerce Era [J]. China Management Science, 2000, 03: 2-8.

[2] Tang Yan. Analysis of China's E-commerce Online Shopping Platform Industry Organization [D]. Wuhan University of Technology, 2012.

[3] Chen Yun, Zhang Penghua, Ren Lihua. A Review of Machine Translation Research [J]. Value Engineering, 2013, 01: 174-176.

[4] Zhu Hai. Research on machine translation system fusion based on obfuscated network [D]. University of Science and Technology of China, 2010.

[5]Nagao M.A.Framework of a mechanical translation between Japanese and English by analogy principle[M].North Holland Publications,1984.

[6] Mairehaba Aili. Research on some key issues of Uyghur-Chinese machine translation based on examples [D]. Xinjiang University, 2014.

[7] Arif Kurban, Ablimiti Abdureyimu, Turgen Ibrayin. Design of an Electronic Dictionary for Uyghur-Chinese Machine Translation [J]. Computer Engineering and Applications, 2006 , 20:76-78.

[8] Rexiti of Kaharjan Abi. Research on Chinese-Uighur-Uighur-Chinese Bidirectional Machine Translation System Based on Cases [D]. Shanghai Jiaotong University, 2012.

[9] Gulisong Nasserdin, Maimaiti Saifudin. Research and Design of Electronic Dictionary for Uyghur-Chinese Machine Translation System [J]. Journal of Xinjiang Normal University (Natural Science Edition), 1997, 01:32 -36.

In order to solve the problem of Uyghur machine translation, Chinese Patent Application No. 201310740830.3 discloses a Uyghur translation engine method using a self-service payment terminal for electricity bills. The patent selects display types such as Chinese and Uyghur from the self-service payment terminal; if Chinese is selected, no need Perform machine translation; if Uyghur is selected, the translation engine will be started to translate the information in the database and displayed on the terminal interface, thereby greatly reducing the cost and time of manual translation between Chinese and Uyghur. The disadvantage of this patent lies in the fact that real-time machine translation is performed when selecting Uyghur language. Although the cost and time of manual translation are greatly reduced, there is still a lack of caching mechanism or Uyghur language database storage in advance to reduce the delay when web pages are loaded.

Another Chinese patent application No. 201310197369.1 discloses an enterprise comprehensive information management system. The patent submits a request for information management to the internationalization synchronization module through the client, and the request includes the choice of language and application mode; the international synchronization module receives the request and divides it. Language management, and then transmit it to the information unified management module; the information unified management module judges and manages the different information in the requests managed by languages; transmits the different information after classification management to the history record module; the history record module receives the classification Different information after management and transfer client. This patent solves the problem of synchronous updating of pages in different language environments. Users can fully grasp the details of the internal personnel, salary, files, tasks and property of the enterprise; all operation steps of users are synchronously saved in the history record module, and can be stored at any time at any time. It can be restored and viewed without hassle. However, the shortcomings of this patent are: the international synchronization module is managed by modules and languages. When a large amount of data is updated by the client, each module needs to be updated synchronously in real time. On the one hand, there is no preprocessing process, and there is a refresh delay in data return; on the other hand, There may be errors in data updates, and there is no human involvement in the correction process.

To sum up, the translation modes of the existing Uyghur-Chinese bilingual machine translation technology are relatively simple, and dynamic real-time machine translation is generally used, without a caching mechanism or data preprocessing process. In B/C mode, webpage rendering may have garbled characters and Lazy loading issue.

SUMMARY OF THE INVENTION

The present invention provides a method and system for developing a multilingual website, aiming to solve one of the above-mentioned technical problems in the prior art at least to a certain extent.

In order to solve the above problems, the present invention provides the following technical solutions: a method for developing a multilingual website, comprising:

Step a: Develop static web pages for multilingual websites;

Step b: calling the machine translation interface to perform multilingual translation processing on the Chinese data dynamically added in the multilingual website;

Step c: read the translation data, load and render the dynamic web page of the multilingual website according to the translation data.

The technical solution adopted in the embodiment of the present invention further includes: in the step a, the multilingual website includes at least Chinese, Uyghur or/and Kazakh; the static webpage for developing the multilingual website is specifically: using Unicode characters Set the UTF-8 encoding format for static web development of multilingual websites.

The technical solution adopted in the embodiment of the present invention further includes: in the step b, the multilingual translation processing of the Chinese data dynamically added in the multilingual website specifically includes:

Step b1: encapsulate the translation interface, take out the Chinese data dynamically added in the website database in batches, store the Chinese data in the document, read the Chinese data in the document line by line, and call the machine translation interface for automatic translation for each line read ;

Step b2: perform manual correction processing on the stored translation data;

Step b3: Store the translation data processed by manual correction in the website database in a corresponding format.

The technical solution adopted in the embodiment of the present invention further includes: in the step c, the loading and rendering of the dynamic web page of the multilingual website according to the translation data specifically includes: when storing the translation data, each Uyghur or Kazakh language The character encoding is converted into a four-digit hexadecimal string, and when the webpage is rendered, the Uyghur or Kazakh language read from the website database is encoded again.

The technical solution adopted in the embodiment of the present invention further includes: the step c further includes: performing caching processing on the loaded web page; the web page caching processing includes file caching and memory caching.

Another technical solution adopted in the embodiment of the present invention is: a multilingual website development system, comprising:

Static web page development module: used to develop static web pages for multilingual websites;

Machine translation module: used to call the machine translation interface to perform multilingual translation processing on the Chinese data dynamically added in the multilingual website;

Web page rendering module: used to read translation data, load and render the dynamic web page of the multilingual website according to the translation data.

The technical solution adopted in the embodiment of the present invention further includes: the multilingual website includes at least Chinese, Uyghur or/and Kazakh; the static webpage development module of the static webpage development module specifically develops the static webpage of the multilingual website: UTF-8 using the Unicode character set -8 encoding format for static web development of multilingual websites.

The technical solution adopted in the embodiment of the present invention further includes a website database module, which is used for storing Chinese data dynamically added in the multilingual website; the machine translation module further includes:

Translation unit: used to encapsulate the translation interface, take out the Chinese data dynamically added in the website database module in batches, store the Chinese data in the document, read the Chinese data in the document line by line, and call the machine for each line read. Translation interface for automatic translation;

Error correction unit: used to perform manual correction processing on the stored translation data;

Storage unit: used for storing the translation data processed by manual correction in the website database module in a corresponding format.

The technical solution adopted in the embodiment of the present invention further includes: the webpage rendering module loads and renders the dynamic webpage of the multilingual website according to the translation data, which specifically includes: when storing the translation data, encoding and converting each character in Uyghur or Kazakh It is converted into a four-digit hexadecimal string, and when the webpage is rendered, the Uyghur or Kazakh language read from the website database module is encoded and converted again.

The technical solution adopted in the embodiment of the present invention further includes a data cache module, which is used for performing cache processing on the loaded web page; the web page cache processing includes file cache and memory cache.

Compared with the prior art, the beneficial effects of the embodiments of the present invention are: the multilingual website development method and system of the embodiments of the present invention adopt a combination of template development of static web pages and dynamic data invoking a machine translation interface, which greatly reduces manual translation. the cost and time; the use of machine translation and manual intervention to correct the processing method greatly reduces translation errors and makes the web page display effect more accurate; by selecting the utf-8 Unicode encoding format, the garbled code generated during web page rendering is avoided; through dynamic The loading cache mechanism solves the problem of resource consumption and loading delay caused by the need to re-call the machine translation interface each time during the real-time translation loading process, while reducing manual intervention.

Description of drawings

Fig. 1 is the flow chart of the multilingual website development method of the embodiment of the present invention;

Fig. 2 is the overall frame diagram of the multilingual website of the embodiment of the present invention;

Fig. 3 is the training flow chart of statistical machine translation;

Fig. 4 is the multilingual artificial assisted translation flow chart of the embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a multilingual website development system according to an embodiment of the present invention.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

Please refer to FIG. 1 and FIG. 2 , FIG. 1 is a flowchart of a method for developing a multilingual website according to an embodiment of the present invention, and FIG. 2 is an overall frame diagram of a multilingual website according to an embodiment of the present invention. The multilingual website development method of the embodiment of the present invention includes the following steps:

Step 10: Develop a multilingual website template theme, and perform static web page development of a multilingual website through the UTF-8 encoding format of the Unicode character set;

In step 10, the multilingual website in the embodiment of the present invention includes at least Chinese, Uyghur, Kazakh, and the like. In the process of developing Uyghur and Kazakh web pages, unified coding is a key technology. Uyghur and Kazakh belong to the Altai language family, and their texts borrow Arabic and some Persian letters. Uyghur has a total of 32 letters. Kazakh There are 33 characters in the language. Uyghur and Kazakh characters are handwritten characters. Each letter has 4 forms of expression according to its position in the word: individual form, initial form, mid-word form, and suffix form. The position in determines the form of appearance. Therefore, Uyghur and Kazakh characters have some particularities when inputting and editing them, which are embodied as follows: (1) The writing direction is from right to left, and the line direction is from top to bottom. The writing direction is opposite, which makes the processing technology more complicated when Uyghur and Kazakh are mixed with Chinese and English; (2) Kazakh has 33 letters, including 9 vowels and 24 consonants. Kazakh vowels are more harmonious, and consonant assimilation is more common. Uyghur consists of 32 letters, and there are more than 120 character forms, each letter has 4 different writing forms, the initial form with the tail connected to the next letter, the middle form with the head and tail connected with adjacent letters, the head with The last form of the last letter connected to the last letter and the independent form of the first and last letters are not connected to the adjacent letters, and which form to use is determined according to the position of the letter; (3) Uyghur punctuation marks, such as commas, question marks, etc. Chinese and English symbols are in opposite directions.

There are many kinds of identical character sets in the field of computer application, and when users of different languages browse webpages in different languages, garbled characters often appear because of different character sets used. Simplified (GB2312) character set is used in general Chinese website system, but for Uyghur and Kazakh website system, the character set of Chinese website does not support their language. Therefore, for websites that want to provide Uyghur, Kazakh and Chinese multilingual versions, a character set that supports Chinese, Uyghur and Kazakh should be selected. The Unicode character set has a unified and unique encoding represented by two bytes (also 4 bytes) for each character of each language, which satisfies cross-language and cross-platform character decoding and conversion processing, but due to The Unicode character set is not compatible with the IS088:59-1 character set, and occupies a large space (for English letters, Unicode also needs two bytes to represent), thus resulting in the UTF character set. There are two types of UTF character sets: UTF-8 and UTF-16. Among them, the encoding specifications of UTF-16 and Unicode are consistent, while UTF-8 is different. It defines an "interval rule". This rule It can maintain maximum compatibility with IS088:59-1 encoding, and can also be used to represent characters in all languages, so UTF-8 is the most ideal choice when designing and developing multi-language versions of websites in Wei, Kazakhstan, and Han. Through the UTF-8 encoding format, the static webpage of the multilingual website displays the normal webpage without garbled characters. The client selects the Uyghur version mode, and the system will load the Uyghur template theme.

Step 20: select the language version mode through the client, and display the static web page of the corresponding template theme according to the language version mode;

Step 30: encapsulate the machine translation interface into a webAPI, call the machine translation interface, and perform multilingual translation processing on the Chinese data dynamically added in the website database;

In step 30, the multilingual website template development of the embodiment of the present invention includes the development of static web pages and the calling process of invoking the machine translation interface during dynamic data loading. The development of multi-language e-commerce platforms such as Weiha requires the development of static web pages, batch The dynamic addition of content also needs to be translated into multilingual versions. The embodiment of the present invention encapsulates a machine translation interface, and invokes the machine translation interface to efficiently solve multilingual translation when adding in batches. After coding preprocessing for static web pages, store the dynamically added Chinese data in the website database, and add the Uyghur language field or Kazakh language field corresponding to the Chinese data in the website database to store the translated Uyghur language or Kazakh language , for dynamic web page rendering and loading.

At present, machine translation methods include rule law and statistical law;

1. Rules method: The text is analyzed according to the language rules, and then translated with the help of computer programs. Most commercial machine translation systems use the rules approach. The operation of the rule-based machine translation system is realized through three consecutive stages: analysis, conversion, and generation; it is divided into three levels according to the complexity of the three stages:

①Direct translation: word-to-word translation;

②Translation translation: The translation process should refer to and take into account the lexical, syntactic and semantic information of the original text. Because the range of information sources is too broad, there are too many grammatical rules and there are contradictions and conflicts between them, the conversion and translation are complicated and error-prone;

③ Interlingua translation.

2. Statistical SMT: As shown in Figure 3, it is the training flow chart of statistical machine translation. Through statistical analysis of a large number of parallel corpora, a statistical translation model (vocabulary, comparison or language pattern) is constructed, and then this model is used for translation. Generally, the entry with the highest probability of occurrence in the statistics is selected as the translation. The probability algorithm is based on the Yeas' theorem. Suppose that an English sentence A is to be translated into Chinese, and all Chinese sentences B are possible or impossible potential translations of A. Pr(A) is the probability of occurrence of an expression similar to A, and Pr(B|A) is the probability of occurrence of A translated into B. Finding the maximum of the two parameters narrows the search of sentences and their corresponding translations to find the most appropriate translation. SMT is divided into two types according to the level of text analysis: word-based SMT and phrase-based SMT. The latter one is currently commonly used, and this is what Google uses. The translated text is automatically divided into fixed-length word sequences, and then each word sequence is statistically analyzed in the corpus to find the translation with the highest corresponding probability.

Specifically, please refer to FIG. 4 , which is a flowchart of a multilingual human-assisted translation according to an embodiment of the present invention. The multilingual translation processing of the dynamically added Chinese data in the website database includes the following steps:

Step 31: Take out the Chinese data (such as commodity data) dynamically added in the website database in batches, store the Chinese data in the document, and write a program to read the Chinese data in the document line by line, and call the machine translation interface for automatic translation for each line read. It is Uyghur or Kazakh data, and the translated Uyghur or Kazakh data is stored in the result document in unicode encoding format;

In step 31, the translation interface is firstly encapsulated, each character string is input, the Chinese data in the website database is taken out and stored in a document, the translation interface is called in turn by line reading, the returned result list is stored in the result document, and then the Insert into the corresponding fields of the website database. This process is an automated and regularly executed process. In the process of all data transmission and invoking the translation interface, the unification of the character encoding format is emphasized.

Step 32: Manually correct the automatically translated Uyghur or Kazakh data;

In step 32, due to the limited number of Viha thesaurus, it is impossible to achieve 100% accurate translation of all Chinese phrases. In order to improve the accuracy of translation, the present invention corrects a small amount of errors in the translation process through manual participation, It greatly reduces translation errors and makes the display effect of web pages more accurate.

Step 33: Read and store the manually corrected Uyghur or Kazakh data in the corresponding format into the corresponding fields of the website database, and automatically process the entire operation process periodically to complete all static and dynamic two-way rendering processes.

Step 40: read the translation data from the website database, load and render the dynamic web page of the corresponding template theme according to the translation data, and perform cache processing on the loaded web page;

编码转换后：“062A 0648 064A”)，在网页渲染时，对从网站数据库中读出的维吾尔语或哈萨克语再做一次编码转换，这样就不存在乱码问题了。In step 40, due to the limitation of the computer in the Uyghur language, when the Uyghur language website renders the web page, there is a Unicode encoding and the corresponding operating system, and the encoding format supported by the browser and the website database converts the input and output problems, For almost all website database drivers, the ISO-8859-1 encoding format is used by default when transferring data between the program and the website database. Therefore, when the website platform stores Uyghur data in the website database, the website database driver will convert the Unicode encoding format into ISO-8859-1 format for storage. When the webpage is rendered, the Uyghur language data read from the website database becomes garbled. In order to solve the problem of garbled characters caused by the incompatibility of Uyghur, Kazakh and Chinese in reading, writing and storage methods, the embodiment of the present invention proposes a code conversion method for Uyghur and Kazakh. , each character encoding of Kazakh is converted into a four-digit hexadecimal string (such as:

After encoding conversion: "062A 0648 064A"), when the webpage is rendered, the Uyghur or Kazakh language read from the website database will be encoded again, so that there is no garbled problem.

Cache processing usually includes: (1) Data cache: refers to the website database query PHP cache mechanism, each time a page is accessed, it will first detect whether the corresponding cache data exists, if not, connect to the database, get the data, and put The query results are serialized and saved to a file, and the same query results can be obtained directly from the cache table or file in the future. (2) Page caching: Every time you visit a page, it will first detect whether the corresponding cached page file exists. If it does not exist, it will connect to the database, get the data, display the page and generate the cached page file at the same time, so that the next time you visit The page file comes into play (template engines and some common PHP caching mechanism classes on the Internet usually have this function). (3) Time-triggered cache: Check whether the file exists and the timestamp is less than the set expiration time. If the timestamp of the file modification is greater than the current timestamp minus the expiration timestamp, then use the cache, otherwise update the cache. (4) Content-triggered caching: When data is inserted or updated, the PHP caching mechanism is forced to be updated. (5) Static cache: Static cache refers to static, directly generating text files such as HTML or XML, and regenerating it once when there is an update, which is suitable for pages that do not change very much. (6) Memory cache: Memcached is a high-performance, distributed memory object PHP cache mechanism system, which is used to reduce database load and improve access speed in dynamic applications. (7) php cache (8) MYSQL cache (9) web cache based on reverse proxy (10) DNS polling. The cache processing in this embodiment of the present invention mainly includes file cache and memory cache. The main function of the cache is to reduce the pressure on the database and the PHP calculator, reduce the delay caused by calling the database data of the machine translation website during web page rendering, and solve the problem of resource consumption caused by the need to re-call the machine translation interface each time during the real-time translation loading process. , while reducing manual intervention. The queried data is directly stored in the cache, and there is no need to repeatedly query the website database, and the pressure on mysql will be reduced; and the operation of php is mainly reflected in, for example, caching the results obtained by a complex recursive operation, without wasting CPU every time. complex operations. During the caching process, the Uyghur encoding process still adopts the above-mentioned Unicode encoding standard.

The embodiments of the present invention are not limited to solving the coding compatibility problem between Uyghur Han and Ha Han. Similar coding processing methods are also used between similar Ke Han, and the whole process of calling the machine translation interface to process the website database data and manual participation The same applies to other multilingual jobs.

Please refer to FIG. 5 , which is a structural diagram of a multilingual website development system according to an embodiment of the present invention. The multilingual website development structure of the embodiment of the present invention includes a static webpage development module, a static webpage display module, a website database module, a machine translation module, a webpage rendering module and a data cache module;

The static webpage development module is used to develop the template theme of the multilingual website, and the static webpage development of the multilingual website is carried out through the UTF-8 encoding format of the Unicode character set; wherein, the multilingual website in the embodiment of the present invention includes at least Chinese, Uyghur, Kazakh etc. In the process of developing Uyghur and Kazakh web pages, unified coding is a key technology. Uyghur and Kazakh belong to the Altai language family, and their texts borrow Arabic and some Persian letters. Uyghur has a total of 32 letters. Kazakh There are 33 characters in the language. Uyghur and Kazakh characters are handwritten characters. Each letter has 4 forms of expression according to its position in the word: individual form, initial form, mid-word form, and suffix form. The position in determines the form of appearance. Therefore, Uyghur and Kazakh characters have some particularities when inputting and editing them, which are embodied as follows: (1) The writing direction is from right to left, and the line direction is from top to bottom. The writing direction is opposite, which makes the processing technology more complicated when Uyghur and Kazakh are mixed with Chinese and English; (2) Kazakh has 33 letters, including 9 vowels and 24 consonants. Kazakh vowels are more harmonious, and consonant assimilation is more common. Uyghur consists of 32 letters, and there are more than 120 character forms, each letter has 4 different writing forms, the initial form with the tail connected to the next letter, the middle form with the head and tail connected with adjacent letters, the head with The last form of the last letter connected to the last letter and the independent form of the first and last letters are not connected to the adjacent letters, and which form to use is determined according to the position of the letter; (3) Uyghur punctuation marks, such as commas, question marks, etc. Chinese and English symbols are in opposite directions.

There are many kinds of identical character sets in the field of computer application, and when users of different languages browse webpages in different languages, garbled characters often appear because of different character sets used. Simplified (GB2312) character set is used in general Chinese website system, but for Uyghur and Kazakh website system, the character set of Chinese website does not support their language. Therefore, for websites that want to provide Uyghur, Kazakh and Chinese multilingual versions, a character set that supports Chinese, Uyghur and Kazakh should be selected. The Unicode character set has a unified and unique encoding represented by two bytes (also 4 bytes) for each character of each language, which satisfies cross-language and cross-platform character decoding and conversion processing, but due to The Unicode character set is not compatible with the IS088:59-1 character set, and occupies a large space (for English letters, Unicode also needs two bytes to represent), thus resulting in the UTF character set. There are two types of UTF character sets: UTF-8 and UTF-16. Among them, the encoding specifications of UTF-16 and Unicode are consistent, while UTF-8 is different. It defines an "interval rule". This rule It can maintain maximum compatibility with IS088:59-1 encoding, and can also be used to represent characters in all languages, so UTF-8 is the most ideal choice when designing and developing multi-language versions of websites in Wei, Kazakhstan, and Han. Through the UTF-8 encoding format, the static webpage of the multilingual website displays the normal webpage without garbled characters. The client selects the Uyghur version mode, and the system will load the Uyghur template theme.

The static webpage display module is used to display the static webpage of the corresponding template theme according to the language version mode selected by the client;

The website database module is used to store the Chinese data dynamically added in the multilingual website;

The machine translation module is used to encapsulate the machine translation interface into a webAPI, call the machine translation interface, and perform multilingual translation processing on the Chinese data dynamically added in the website database module; wherein, the multilingual website template development in the embodiment of the present invention includes the development of static web pages Unlike the calling process of calling the machine translation interface when loading dynamic data, the development of multi-language e-commerce platforms such as Weiha requires the development of static web pages, and the dynamic addition of batch content also needs to be translated into multilingual versions. The embodiment of the present invention encapsulates the machine translation interface. , call the machine translation interface to efficiently solve the multilingual translation when adding batches. After coding preprocessing for static web pages, store the dynamically added Chinese data in the database module, and add the Uyghur or Kazakh fields corresponding to the Chinese data in the database module to store the translated Uyghur or Kazakh , for dynamic web page rendering and loading.

Specifically, the machine translation module includes a translation unit, an error correction unit and a storage unit;

The translation unit is used to take out the Chinese data in the database module in batches, store the Chinese data in the document, and write a program to read the Chinese data in the document line by line, and call the machine translation interface for each line read to automatically translate it into Uyghur or Kazakh data, and store the translated Uyghur or Kazakh data in the result document in unicode encoding format; among them, the translation interface is firstly encapsulated, a string is input, and the Chinese data in the database module is taken out and stored in a document. The translation interface is called in sequence by line reading, the returned result list is stored in the result document, and then inserted into the corresponding fields of the database module in turn. This process is an automated and regular process. In the process of all data transmission and translation interface invocation, the character encoding format is emphasized. unity.

The error correction unit is used to manually correct the automatically translated Uyghur or Kazakh data; among them, due to the limited number of Viha thesaurus, it is not possible to achieve 100% accurate translation of all Chinese phrases. In order to improve translation The invention corrects a small amount of errors in the translation process through manual participation, greatly reduces the translation errors, and makes the display effect of the webpage more accurate.

The storage unit is used to read and store the manually corrected Uyghur or Kazakh data in the corresponding format into the corresponding fields of the database module, automatically process the entire operation process periodically, and complete all static and dynamic two-way rendering processes.

编码转换后：“062A 0648 064A”)，在网页渲染时，对从数据库模块中读出的维吾尔语或哈萨克语再做一次编码转换，这样就不存在乱码问题了。The webpage rendering module is used to read the translation data from the website database module, load and render the dynamic webpage of the corresponding template theme according to the translation data; among them, due to the limitation of the computer in the Uyghur language, when the Uyghur language website renders the webpage, there is a A Unicode encoding and the corresponding operating system, the encoding format supported by the browser and the database to convert the input and output problems, for almost all database drivers, the default is to use ISO-8859-1 when transferring data between the program and the database. encoding format. Therefore, when the website platform stores Uyghur data in the database module, the database driver will convert the Unicode encoding format into ISO-8859-1 format for storage. When the webpage is rendered, the Uyghur data read from the database module becomes garbled. In order to solve the problem of garbled characters caused by the incompatibility of Uyghur, Kazakh and Chinese in reading, writing and storage methods, the embodiment of the present invention proposes a code conversion method for Uyghur and Kazakh. , each character encoding of Kazakh is converted into a four-digit hexadecimal string (such as:

After encoding conversion: "062A 0648 064A"), when the webpage is rendered, perform encoding conversion on the Uyghur or Kazakh language read from the database module, so that there is no problem of garbled characters.

The data cache module is used to cache the loaded web page; the cache process usually includes: (1) Data cache: refers to the database query PHP cache mechanism, each time a page is accessed, it will first check whether the corresponding cache data exists, if not If it exists, connect to the database, get the data, and serialize the query result and save it to the file. In the future, the same query result will be directly obtained from the cache table or file. (2) Page caching: Every time you visit a page, it will first detect whether the corresponding cached page file exists. If it does not exist, it will connect to the database, get the data, display the page and generate the cached page file at the same time, so that the next time you visit The page file comes into play (template engines and some common PHP caching mechanism classes on the Internet usually have this function). (3) Time-triggered cache: Check whether the file exists and the timestamp is less than the set expiration time. If the timestamp of the file modification is greater than the current timestamp minus the expiration timestamp, then use the cache, otherwise update the cache. (4) Content-triggered caching: When data is inserted or updated, the PHP caching mechanism is forced to be updated. (5) Static cache: Static cache refers to static, directly generating text files such as HTML or XML, and regenerating it once when there is an update, which is suitable for pages that do not change very much. (6) Memory cache: Memcached is a high-performance, distributed memory object PHP cache mechanism system, which is used to reduce database load and improve access speed in dynamic applications. (7) php cache (8) MYSQL cache (9) web cache based on reverse proxy (10) DNS polling. The cache processing in this embodiment of the present invention mainly includes file cache and memory cache. The main function of the cache is to reduce the pressure on the database and the PHP calculator, reduce the delay caused by calling the machine translation database data when the web page is rendered, and solve the problem of resource consumption caused by the need to re-call the machine translation interface each time during the real-time translation loading process. It also reduces manual intervention. The queried data is directly stored in the cache, and there is no need to repeatedly query the database, and the pressure on mysql will be reduced; and the operation of php is mainly reflected in, for example, caching the result obtained by a complex recursive operation, without wasting the CPU every time for complex operation. During the caching process, the Uyghur encoding process still adopts the above-mentioned Unicode encoding standard.

The multilingual website development method and system of the embodiment of the present invention adopts the combination of template development of static webpage and dynamic data calling machine translation interface, which greatly reduces the cost and time of manual translation; Reduce translation errors and make web page display more accurate; by choosing utf-8 Unicode encoding format to avoid garbled characters during web page rendering; Resource consumption and loading delay problems caused by the need to re-call the machine translation interface, while reducing manual intervention.

The above description of the disclosed embodiments enables any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (5)

1. A multi-language website development method is characterized by comprising the following steps:

step a: developing static web pages of the multi-language website;

step b: calling a machine translation interface to perform multilingual translation processing on the dynamically added Chinese data in the multilingual website;

step c: reading translation data, and loading and rendering the dynamic webpage of the multi-language website according to the translation data;

in the step b, the performing multilingual translation on the dynamically added chinese data in the multilingual website specifically includes:

step b 1: packaging translation interfaces, taking out dynamically added Chinese data in a website database in batches, storing the Chinese data in a document, reading the Chinese data in the document according to lines, and calling a machine translation interface to perform automatic translation when reading one line each time;

step b 2: carrying out manual correction processing on the stored translation data;

step b 3: storing the translation data subjected to manual correction processing into the website database according to a corresponding format;

in the step c, the loading and rendering the dynamic webpage of the multi-language website according to the translation data specifically comprises: when the translation data is stored, each character code of Uygur language or Kazakh is converted into a 16-system character string with four bits, and when the webpage is rendered, the Uygur language or Kazakh read from the website database is subjected to code conversion again;

the step c further comprises: caching the loaded webpage; the webpage caching process comprises file caching and memory caching.

2. The method for developing multilingual website of claim 1, wherein, in the step a, the multilingual website comprises at least chinese, uygur or/and kazak; the static web page for developing the multilingual website specifically comprises the following steps: and carrying out static webpage development on the multilingual website through the UTF-8 coding format of the Unicode character set.

3. A multilingual website development system, comprising:

a static webpage development module: static web pages for developing multi-lingual sites;

a machine translation module: the system is used for calling a machine translation interface to perform multilingual translation processing on the dynamically added Chinese data in the multilingual website;

the webpage rendering module: the system is used for reading translation data, loading and rendering the dynamic webpage of the multi-language website according to the translation data;

the machine translation module includes:

a translation unit: the system comprises a website database module, a translation interface, a machine translation interface and a database module, wherein the website database module is used for storing Chinese data dynamically added into the website database module in batches, storing the Chinese data into a document, reading the Chinese data in the document according to lines, and calling the machine translation interface for automatic translation when each line is read;

an error correction unit: the translation data storage unit is used for storing translation data;

a storage unit: the translation data used for manual correction processing is stored in the website database module according to a corresponding format;

the step of loading and rendering the dynamic webpage of the multi-language website by the webpage rendering module according to the translation data specifically comprises the following steps: when the translation data is stored, each character code of Uygur language or Kazakh is converted into a 16-system character string with four bits, and when the webpage is rendered, the Uygur language or Kazakh read from the website database module is subjected to code conversion again;

the system also comprises a data caching module, wherein the data caching module is used for caching the loaded webpage; the webpage caching process comprises file caching and memory caching.

4. The multi-lingual site development system of claim 3, wherein the multi-lingual site comprises at least Chinese, Uygur, and/or Kazakh; the static webpage development module is used for developing the static webpage of the multi-language website and specifically comprises the following steps: and carrying out static webpage development on the multilingual website through the UTF-8 coding format of the Unicode character set.

5. The multi-lingual website development system of claim 4, further comprising a website database module for storing dynamically added Chinese data in the multi-lingual website.

Images