TW473673B - Method and apparatus for compressing scripting language content - Google Patents

Abstract

A method and apparatus for compressing scripting language content, such as hypertext markup language (HTML). Codewords are provided for HTML elements, such as tags and their attributes, to reduce the amount of data, e.g., in a Web page. The codewords are combined with translated (or coded) text to provide compressed HTML data. The codewords may have reserved bits to distinguish empty tags from container tags, and to indicate whether a container tag is a starting or closing tags, or to provide other information about the tag to aid in processing. The amount of data that must be communicated, e.g., to represent a Web page that is transmitted to a subscriber terminal, is thereby reduced. Additionally, the invention allows the use of a graphics engine or browser, e.g. in a subscriber terminal that processes/renders the compressed HTML data (e.g., codewords) directly without decompressing them. The technique is compatible with other compression techniques to provide even greater compression.

Description

473673 V. Description of the invention (1) Background of the invention The present invention relates to a method and a device for compressing the content of a descriptive language, such as HyperText Markup Language (HTML). M T M L is a system used to mark files to indicate how the files should be displayed and how different files should be linked together. HTML has been widely used to provide documents (such as web pages) in Int e r n e t. This file is organized into web space, where a web space contains a web page and links to other files in the local web space or in an external web space. Such links are known as superlinks. Files can contain moving images, images, text, graphics, and sound. HTML is a standard generalized markup language (SGML). It is defined by the International Standards Organization (iSQ) and reference number ISO 887 9: 1 986. HTML specifies the grammar and the syntax of markup tags. It is inserted into a data file and can define how the data should be represented by a computer program, known as a browser (such as plotting a daytime representation). The computer's browser and / or graphics engine processes the data to format the layout of the web page, so this page can be viewed by a user on a display or device. An SGML file contains three parts. The first section describes the characters, or codes, which are used in the language. The second part defines the type of document and identifies the "brother tag. The third part is a known example of the document and contains the actual text and tag labels. The three parts can be stored in different building plans. '' HTML browsers assume that files on different web pages contain a common character and

Page 6 47367

So only the text and markup tags can be changed for different web pages. HTML components include tags and chant-gold

Paper man 疋 疋 entity. Character entity is from ISO

The Latin-1 alphabet's pre-defined command-& MY tl, is not defined in ASCII, but is used to mark the HTML character "仏", "i", the beginning and end characters. For example The character entity "& 11:" means the character 疋 <(that is, "less than ,, symbol). HTML tags are included in the angle bracket + +,? ^ ^ #? To distinguish them from web page text: this; wear: appear alone (as a separate or blank tag), or can appear on the web page text The start and end of the stop (non-blank or container label). For example, &lt; P &gt; is -six & throw from ^ ^ ^ 工 '白 _ 表 戴 to indicate the beginning of a new paragraph Luo, and &lt; I &gt; and &lt; / I &gt; are Container tag to modify the contained text (eg &lt; I &gt; WelC〇me to my h〇me page. &Lt; / 1: &gt; represents an idiom, ieic〇me to my home page "should be typed in Italian) . ^〗>, Is the start tag, and π &lt; / I &gt; π is the end tag. Generally speaking, 'HTML tags provide text formatting, hypertext links to other web pages, links to sounds and pictures Like components. HTML tags also define input shelves for interactive web pages. In addition, some tags have one or more relevant attributes that can be set by this tag. For example, the tags &lt; A &gt; and &lt; / A &gt; An anchor code, which defines a paragraph of text as a hyperlink, or the target of another hyperlink. The attributes of this tag include HREF = uri, NAME = name, and TITLE = text. Therefore, "HTML stone horse" &lt; A HREF = ”http: /// www · uspto · g〇v” &gt; U · S · Patent and Trademark 0ffice &lt; / A &gt; '1f The text &quot; u.S. Patent and Trademark Office "appears in a special emphasis on

Page 7 473673 V. Description of the invention (3) 'In the browser (for example, a special color and / or adding an underline), it may indicate that the text is a hyperlink. When the user clicks on this text, the address of the web page goes to &quot; WWW · usp t 0 · gov. &Quot; Furthermore, the label has a secondary attribute or a secondary attribute. For example, the tag &lt; IMG &gt; Is a blank label indicating that a coaxial image is located on the same page. This attribute contains SRC = url, which specifies the URL of the slot where the image is to be contained, ALT = text, which specifies a URL that can be displayed in the image Text strings cannot be used, and ALIGN = [TOP | MIDDLE | BOTTOM], which recognizes how this image is aligned with adjacent text and other HTML components. Therefore, alig = is an attribute, TOP, MIDDLE and BOTTOM are secondary An example of HTML code is: &quot; &lt; IMG SRC: nfilename.GIF &quot; ALI'Jfilenajue :, ALIGN = middle &gt; 'f 〇 Moreover, HTML tags and attributes are generally referred to herein as "HTML" element π terms "Attributes" generally include hierarchies of different sub-attributes.

An HTML application can be used for users on the webpage: This user can save the user ^ ^ I support the user, which is stored by a HTML grid in a directory, and accessed by a servo * version . This kind of server cry is basically a web server, which meets the requirements of [Waiting] w &gt; Communication Supported by Webpage Play i§ Supported Communication Protocol ’is called Hypertext Conversion Communication Protocol T m ^ (Hypertext

Transfer Protocol, HTTP) 〇 In addition, HTML content can be stored in the header of a user communication network, such as: = / Satellite TV network. Because of the high-speed data rate of the Internet and the transparent network, there is an increasing trend toward the potential commercial benefits of providing two households = content combined with traditional TV program services. The "type of telephone, television and computer networks, For home computer networks

Page 8 473673 V. Explanatory Notes (4) The expectations can be increased. After the HTML content can be selected, it is directly provided by the headend, or the headend is only used as a high-speed link between the user and the remote web server. Servers conforming to other communication protocols, such as File Transfer Protocol (FTP), or GOPHER, can also access an HTTP browser through a proxy server. A proxy server is a form of gateway that allows a browser to use HTTP to communicate with a server that does not understand HTTP, but uses protocols such as FTP, Gopher, or other protocols. The proxy server accepts HTTP requests from the browser and translates them into a format suitable for the original server, such as an FTP request. Similarly, the proxy server translates FTP responses from the server into HTTP responses, so the browser can understand them. ·

Generally speaking, FTP files themselves are not translated. FTP is a high-level term. Lu ’is used to transfer files (such as Η T T P). This translation takes place at the level of the communication protocol. For example, a client browser can send an HTTP request, GET http: //www/.mvserver· com / somefile · txt HTTP / 1 1. On the proxy server, this will be translated into an FTP, GET, request and sent to the original server of FTP. The j? Τ p response from the original server is passed back to the proxy server (with the requested file attached), which is then translated (in the proxy server) to an HTTP response and contains this additional file. The case to be transmitted is not translated or modified. However, in some cases, the browser can, although it can decode certain graffiti or compressed formats. Therefore, the proxy shrine can translate (encode or compress) the attached file before it can be transmitted to the client. Oral

473673 V. Description of the invention (5) The proxy server can be a program running on the same machine, such as a browser, or an independent machine somewhere in the network, and serves a lot of browsing. For example, the user communication network The headend can provide a proxy server function. HTTP defines a set of rules that servers and browsers must follow when communicating with each other. Basically, when the user clicks on an image in an HTML page, the image is a hyperlink anchor, or the user enters a Uniform Resource Locator (URL) ). This URL contains a master station name, which is basically converted to an IP address through a domain name system (DNS) lookup table. Then, a connection is made to the master server, which uses the IP address (and possibly a connection number) returned by the DNS lookup table. The browser then sends a request to receive an object from the server, or to send data to the object on the server. The server sends a response to the browser, which contains a status code and response data. The connection between the browser and the server is then closed. &quot; 'URL is a unique address that actually identifies all files and resources on the Internet. However, due to the flexibility of HTML, the different tags, their attributes, and sub-attributes that they support need to be used to represent the amount of data on any given webpage. Therefore, the processing power required by a user terminal and browser cloud may not be enough to keep up with the data stream, thus causing delays in displaying data on the user's screen or other problems. Furthermore, because the bandwidth consumed by sending HTML data is increased, so

Page 10 473673

This reduces the bandwidth available to its users or increases the burden on channel capacity. The HTML data can be transmitted over the public switched telephone network (psTN), which can be via a local wireless network such as a cable telephone satellite television network: or through a combination of the foregoing. Specifically, § 'HTML's basic alphabet is Latin — 1 (iso 8859/1), which is an 8-bit letter whose characters can be used in most American and European languages. The 128-character standard ASCII (ISO 646) is a 7-bit combination of La tin-1. For simplicity and easier integration with other browsers, many web pages contain only one ASCI I character. For each character, it needs 8-bit or one-byte data (including a text, number, punctuation mark, or blank). For example, the HTML code: `` &lt; IMG SRC = nfilename. "GIF" ALT = &quot; filename ,, ALIGN = middle &gt; n has 52 characters, or 52 bytes of data. Therefore, 'There is a need to provide a system to compress and write language content, such as HTML or any similar language. This system must reduce the amount of bandwidth needed to transmit data to a browser or other (graphics) display engine. This system must be suitable for existing networks that transmit HTML data. This system must allow The browser on a terminal (such as a set-top box / decoder) in the user's TV network can directly process and display the compressed data without decompression. This system must reduce the The processing power required by the browser on a user terminal.

Page 11 473673 Description of the invention (7) j t j must be stable and fixed for a given web page for all HTML elements and attributes. This ί ^ t ί ^ is on the client / network side and server side of the network. And an M server ^ proxy server to form an interface between a client / browser and server, or other proxy server. ^ Ϊ Μ 一 uses a digital image communication protocol, such as MPEG_2, to deliver HTML-compliant web-compatible materials. This system must be compatible with a network that uses HTML to send Dixon], s: ^ # Ρ / ΤϋΛ + Μ transport control protocol / Internet protocol (TCP / IP) to transmit HTML data. This system must provide compression of the current version of the TM version and other similar markup versions of rm, along with its derivative versions, which are compatible with other bit-level scaling technologies. Yu = A system with these and other advantages. Summary of the Invention: The present invention relates to a device for compacting, such as HTML. Used for "methods and methods of writing language content; ™ or other written language elements, such as tagging and speaking to reduce the amount of data ', for example, on a web page, reserved bits are used to distinguish between blank tags and Container label, and means 俨 :: Basin: The device is worn as an opening or closing label, or can provide this tribute message to assist in processing. This technology is different from other compression technologies, and Asia can provide even higher Compression rate. The present invention can effectively reduce the amount of data to be transmitted, for example, to be transmitted to a user terminal on behalf of a web page. Furthermore, the present invention can

Page 12 473673 V. Description of the invention (8) The terminal starts the use of a drawing engine or browser, and directly processes and displays the compressed HTM L data without decompression, so the processing time and complexity can be greatly reduced. degree. A special method for processing written language data, including the steps of analyzing HTML data, and separating its text from its written language elements. The written language element contains tags and their attributes. Individual coded characters, such as double-byte coded characters, are used for each different label. The text is encoded, for example, A S C I I. The coded characters are then combined with the coded text in the appropriate sequence to provide compressed written language information. Encoded characters may contain reserved bits to indicate specific information, such as whether the related label is a blank label or a container label. For the container label, the encoding character can indicate whether the container label is an opening label or an ending label. The coded character can indicate whether a tag is a style tag tag or a structured tag tag. For structured markup tags, the coded character can indicate whether the structured markup tag is a block element, a list element, a table element, a format element, a hypertext link, an image between text, or a page mark tag. A different encoding character can be used for each different label attribute, including sub-labels. The encoding element can also indicate the attribute number associated with a tag. In a particularly advantageous implementation, the compressed written language is transmitted by a written language content server or headend to a user terminal in a communication network. For decompressing a compressed written language, such as a user terminal, the compressed written language data is analyzed to separate its encoding by its encoding characters.

Page 13

, Writing text language components and decoding Writing language data. Used for each corresponding different encoding to provide decoded text. The most combination of text to provide uncompressed compressed written text data can be transmitted as needed.

A related device is also revealed. Brief Description of the Drawings Figure 1 shows a user television circuit using H_TML compression according to the present invention. Figure 2 shows the HTML compression according to the present invention. FIG. 3 shows an HTML decompression according to the present invention. Detailed Description of the Invention The present invention relates to a method and device for writing language content, such as HTML. Figure 1 shows a user television network using HTML compression according to the present invention. Although the present invention can be implemented on different networks, it is particularly suitable for use in consumer television networks, and allows users (users) to access HTML data, such as on the Internet. Basically, users can access HTML content, such as web pages, and send them through a downstream channel on the Internet. For example

Page 14

W367J Invention Description (ίο) ----- Many technologies can be used to upload ML data through cable and satellite electric network. The user basically provides an upstream link and a telephone network to enter instructions, such as a URL address to request a display on a 11 'electricity. Some cable television networks have an upstream, 疋 ^ page for this purpose. The requester will receive the request in the headend or other central location and forward it to the content server specified by URL—. The content is sent back to the headend by the server, and then ready for delivery to the user. For example, HTML data can be included in digital MPEG-2 packets, which are in the same or different bands as the program service data (eg TV shows, music, etc.). 10 or 'HTML data can be loaded into the vertical blank space (VBI) of digital or analog TV signals. The invention is compatible with essentially any communication technology used to provide HTML feeds to end users. The content of the HTML is then returned on the user's terminal and displayed by a browser application or edge map processing engine for viewing on a video monitor, such as a television or computer monitor. The headend can act as a proxy server when forming an interface with the content server. For example, when the URL accepts a request from a user, the format is not compatible with the content server. In this example, the proxy server converts the URL request to the required format, and converts the content returned by the server into a format that the user's terminal can understand. FIG. 1 shows an exemplary embodiment. The network 100 includes a content server 1 1 0 ′, a head end 130, and a user terminal 150. Content server

Page 15 473673 V. Description of the invention (11) 1 1 0 represents the data in any number of networks; for example, the feeding device 'its storage ... enough = self-table-some terminal combination, can / fiber, or The broadcast signal of the satellite provider's, for example, the front-end in the horizon, the live-view network, 1300, as needed, and a modem, which allows the level end: the first operation of the second transport, such as the conventional telephone key In conclusion, ^ machine 150 communicates directly with the content server. Attentive :: From the head of the shop, laugh, do n’t look at it immediately, oxygen and stock information, and the content is also through a one-way or end group, including the example terminal 150. Strong 162 玎, Fen σ 5 It then transmits to Tongfang 162 to the terminal 150. The channel is multi-function WDM fine ds), or like a telephone link, or: Siam 164 allows headend 130 and sample content server]] This channel is basically-telephone Links or Ethernet to implement this pass: two! 1; 0-generally are far away from the head end 130, although it is possible that the head end can store content on -dead = media, such as digital video discs Or magnetic: mention; :: file 1 hard disk of feeder. &quot; Known network architectures can be used for this purpose. When the headend 130 provides local content, it provides a limited amount of content. This content can also be selected to correspond to the program service. In this example, a graphic may be overlaid on the TV program, and the user is notified of the available m [inner valley]. For example, in a television baseball broadcast, the user can point to a 473673 V, invention description (12), 罔 station to get the baseball score. In some examples, the entire local content may be broadcast continuously or periodically &apos; e.g., the same channel (or multiplexing mode) as the program service, or on a separate channel (or multiplexing mode). This can only happen on one-way networks, even if the user does not have an upstream link to the headend. A selection of the desired HTML content is then generated 150 on the user terminal. Known state access technologies can be used to provide access to HTML content at a billing rate. The invention is applicable in any of the ways described above. In the example of FIG. 1, it is assumed that the user has some upstream channels (丨 60 or 162) to enable the selected HTML content to be retrieved by the content server 110 and provided to the headend 130 Terminal 1 50. The content server 1 10, head end 1 30 and terminal 1 50 are respectively HTML compression functions 112, 132 and 152, and HTML compression functions 114, 134 and 154. However, not all functions are required. Generally speaking, providing compression of HTML data to a terminal is the most important subject. The HTML data is output by the terminal and, if any, is generally small. But this will also change, for example, the user sends HTML content to other users, or is authorized to send HTML content and modify the remote control server 110. The compression function 152 is used to compress the data transmitted from the terminal 15 to the headend 130 or the content server 110. Decompression function 5 4 is used to decompress the compressed data received from the headend 1 3 Q or the content server. Compression function 1 3 2 is used to compress and transmit the data from the headend 13 to the inside.

P.17 473673 —Five'B · ~ --- —-= monthly generator 110 or terminal 150. The decompression function 134 is used to decompress the compressed HTML data of the Uchiya server 11 or the terminal 150. The shrink function 2 is used to compress the HTML asset 130 or the terminal 150 transmitted from the content server 110. The decompression function 114 is used to decompress compressed HTML data from the headend 130 or the terminal 150. 8. The terminal 150 includes a user interface 58 for receiving user's finger. For example, the remote control is through a keyboard or infrared. For example, the user:-~ Staff instructions: Click on a graphic about url on 70, and start the download of the related inner valley to the terminal 150. -Browser 159 can be a full-featured browser application, such as a human brain or a small browser, which has only some basic functions, such as text display or limited drawing display capabilities. The browser 599 is combined with the drawing engine 156 to display text and images on the display 170 by the HTML content received at the terminal 150. &quot; T uses video ‘decoder 1 5 7 to display video, and uses compressed (or uncompressed) written language content for display 17 °. The display 170 may be a television screen or a video surveillance terminal like a pc. The processing power of the terminal 150 will affect the level of functions supported by the browser 159 and the graphics engine 156. Compression functions 1 12, 2, 1 32, and 1 52 can implement an HTML compression architecture shown in FIG. 2 and decompression functions 14, 134, and 154 can implement an HTML decompression architecture shown in FIG. 3. FIG. 2 shows an HTML compression method according to the present invention. Compression function: 200 pairs

473673 V. Description of the invention (14) The compression functions 11 2, 132 and 152 of FIG. 1 should be used. A buffer / parser 210 receives uncompressed HTML data. Please note that this HTML material can refer to the location where audio, video or graphic material can be found. The text is analyzed and provided to a conventional text encoding function 2 丨 5 to provide encoded text, such as AS C I I data. HTML elements' like tags, including their attributes, sub-attributes, sub-attributes, etc., are analyzed and provided to a compression function 2 2 0, which can have a lookup table 225 as needed, and can use known techniques get on. This lookup table 2 25 associates a coded character with each HTML element (tag and attribute). The length of the encoding character is selected according to the number of different tags and attributes to be encoded. A 16-bit coded character (rain bytes) is known to be suitable for processing existing tags while allowing future scalability. Furthermore, it is possible to reserve one or more 16 bits to designate the label as a blank label or a container label. For example, the most significant bit can be selected. For container labels, one or more other reserved bits can also indicate whether the label is a start or end label. Other information that can be pointed out is whether the label is a 〆 style mark woven or a structured mark label. In general, style tag tags (to specify thick line styles, fonts, quoted text, etc.) are used in structured tag tags (to specify table columns, table tags, anchors, etc.) , But the opposite is not recommended. For structured markup, the coded characters can specify whether the label is ~ block element, list element 'table element, format element, hypertext element 473673 V. Description of the invention (15) A coded character can also be specified with each The number of tag-related attributes. The number of bits reserved for this purpose must correspond to the maximum expected number of attributes. For example, three bits can be used to specify that up to eight attributes are associated with the tag here. In general, bits must remain in the coded characters to indicate the characteristics of the tags to help display HTML data. For example, the designation of the start and end container tags is useful because it tells a processor about the range of text to be modified.

For example, each character requires 8 bits of data (including a letter, number, punctuation, or blank), HTML code "&lt; IMG SRC = ', filename.GIF" ALT = &quot; fiiename " ALIGN = middle &gt; n has a total of 52 characters, or 52 bytes of data. By replacing one bit for each HTML element (such as &quot; IMGn, nSRC "," ALTn, and "ALIGN") The tuple encodes characters. There are 14 bytes that need to encode these elements, and it is reduced to 8 bytes, thus saving 6 bytes. In a given Η μL page, use this The number of bytes saved by the invention can save more for longer components, (for example: comparison "&lt; 81 ^ 0 (: 1 ^ 110 丁 £ &gt;), which can be reduced from 12 bytes to 2 Bytes, for "&lt; A &gt;", it is reduced from 3 bytes to 2 bytes), and the number of components in a web page. For each component, a coded character system Output by the compression function 2 2 0 and used in a combiner 2 3 0 to combine the coded characters in the appropriate sequence and provide according to the present invention Condensing the HTML data. This data includes character codes for the text, and from the compression function 220 for codewords HTML elements. Note that other known compression techniques, such Lempe 1 -Z i v algorithms and

Page 20 473673 V. Description of the invention (16)

Huffman encoding can be used to output compressed HTML data from the combiner 23, or for encoded text alone, or only for encoded characters. Furthermore, related video / audio data can also be compressed using the known techniques described above. FIG. 3 shows an HTML decompression according to the present invention. The decompression function 300 corresponds to the decompression function in FIG. 1, 4, 1, 3, 4, and 5 4. Here, the compressed HTML is received by a buffer / parser 31. The encoded text containing text data is used for a text decoding function 3 1 5 to reply to the text, and then used for a combiner 330, in which the HTML editor character is used for a decompression function 320. The lookup table 325 at the decompression function 320 is associated with each received HTML element. The related components are output to the combiner 330 to form uncompressed HTML data. Therefore, it can be understood that the present invention provides a method and device for compressing written content, such as HTML. Coded characters are used in HTML components, such as tags and their attributes, to reduce the amount of data, such as in a web page. Furthermore, the 'encoded character can have reserved bits to distinguish between a blank label and a container label' and indicate whether a container label is an opening or closing label, or provide other information about the label to assist in processing. This technology is compatible with other compression technologies5 and provides even more compression and shrinkage. The present invention can effectively compress the amount of data that must be transmitted, for example, to indicate a web page transmitted to a user terminal. In addition, the present invention allows the use of a graphics engine or browser, such as processing / displaying compressed ΗTLM data (such as coded characters) directly in a user terminal, without the need for decompression. This can greatly reduce processing time and complexity. In addition, each braille character has the same length, so it usually costs the same

Page 21 473673 V. Description of the invention (π) The processing time is similar, so the processing time becomes a determinable amount. The technology of the present invention can be implemented using any known hardware or software. Although the present invention has used different specific embodiments to add those skilled in the art, it can be understood that different adjustments and modifications can be made without departing from the spirit and scope of the scope of this patent. When discussing broadband or satellite TV broadband, they can learn about other networks, such as LANs, MANs, WANs, intranets, or a combination of the above. In other words, the present invention is applicable to compressed zip code, HTML or other similar languages (for example, σ, transcribed content, multimedia integrated language (SMIL)). Extensible Markup Language (XML) and / or resilience to illustrate that the invention application is modified. Internet, Internet, Include, or Sync

Claims (1)

473673 VI. Scope of Patent Application 1 · A method for processing written language data, which includes the following steps: (a) Analyze the written language data and separate its text from its written language element, which contains a label (B) provide individual coded characters for each different label; (c) code the text to generate coded text; and (d) combine the coded characters with the coded text to produce compressed written language data. 2. The method according to item 1 of the scope of patent application, wherein: at least one coded character indicates whether the relevant label is a blank label or a container label. 3. The method according to item 1 of the scope of patent application, wherein for the container label, at least one corresponding coded character indicates whether the container label is an opening label or an ending label. 4. The method according to item 1 of the scope of patent application, wherein at least one coded character indicates whether the corresponding label is a style label or a structure label. 5. The method according to item 1 of the scope of patent application, wherein for the structure mark label, at least one corresponding code character indicates whether the structure mark label is a block element 'list element' table element 5 format element 'hypertext bond , Built-in graphics, or web tag tags. 6. The method according to item 1 of the scope of patent application, wherein the written text element includes the attributes of the label, and further includes the following steps: · Provide a different encoding character for each different attribute. Page 23 473673 VI. Application for patent protection 7 · The method of item 1 of the patent application scope, wherein: the written language element contains the attributes of the tag; and the individual coded characters represent the number of attributes associated with each tag. 8. The method according to item 1 of the scope of patent application, further comprising the following steps: In a communication network, transmitting the compressed written language data from a written language content server to a user terminal. 9 · The method according to item 1 of the scope of patent application, further comprising the following steps: In a communication network, the compressed written language data is transmitted from one end to a user terminal. I 0 · If the method of the scope of patent application No. 1 further includes the following steps: In a communication network, transmit the compressed written language data to a user terminal; and · Process the compressed written language data without writing back Language elements to provide information suitable for display. II · The method according to item 1 of the scope of patent application, further comprising the following steps: (d) Analyze the compressed written language data and separate the written language elements from its text; (e) Decode the encoded text and provide decoding (F) providing separate written language elements for each corresponding different coded character contained in step (d); and (g) writing the language elements and decoding in step (f) Text combination to provide uncompressed written language information. 12. The method according to item 11 of the scope of patent application, wherein: the uncompressed written language data is a user terminal in a communication network Page 24 473673 VI. Scope of patent application Terminals use the steps (d)-(g) for processing. 1 3. The method of claim 1 in the scope of patent application, wherein: the written language data includes hypertext markup language (HTML) data. 14 · The method according to item 1 of the scope of patent application, further comprising the following steps: Temporarily storing compressed written language data in a proxy server for written language content frequently accessed by user terminals. 1 5. A device for processing written language data, comprising: a first analyzer for analyzing the written language data to separate its written language element from its text, the written language element including a label; first A device for providing a different coded character for each different label, a device for coding the text to provide the coded text; and a first combiner for combining the coded character with the coded text To provide compressed written text. 16 · The device according to item 15 of the scope of patent application, wherein: at least one coded character indicates whether the relevant label is a blank label or a container label. 17. The device according to item 15 of the scope of patent application, wherein for the container label, at least one corresponding coded character indicates whether the container label is an opening label or an ending label. 18 · The device according to item 15 of the scope of patent application, wherein: at least one coded character indicates whether the corresponding label is a style label or a structure label. 19. The device according to item 15 of the scope of patent application, wherein Page 25, 473673 6. At least one corresponding coded character indicates whether the structure tag label is a block component, a list component, a table component, a style component, a hypertext link, a built-in graphic, or Web tag labels. 2 〇 The device according to item 15 of the scope of patent application, wherein the written language element includes the attributes of the label, and further includes: a device for providing a different coded character for each different attribute. 2 1 · The device according to item 15 of the scope of patent application, wherein: the written language element includes attributes of tags; and individual coded characters represent the number of attributes associated with each tag. 2 2. The device according to item 15 of the patent application scope, wherein: In a communication network, the compressed language data is transmitted from a written language content server to a user terminal. -23. The device according to item 15 of the scope of patent application, wherein: In a communication network, the compressed written language data is transmitted from a head end to a user terminal. 2 4. The device according to item 15 of the scope of patent application, wherein the compressed language data is transmitted to a user terminal in a communication network, and further includes: at least one processor for processing the compressed written language data without The written language elements must be replied to provide information suitable for display. 2 5. The device according to item 15 of the scope of patent application, further comprising: a second analyzer for analyzing the compressed written language data to separate its coded characters from its coded text; the second device for Providing a different written language element for each corresponding different tag from the second analyzer; Page 26 473673 VI. Patent application device for decoding encoded text to provide decoded text; and a second combiner for combining the written language elements provided by the second device with the decoded text to provide Uncompressed written text. 26. The device according to item 25 of the scope of patent application, wherein: a user terminal on which a language data is written uncompressed on a communication network, the second analyzer, the second device, a device for decoding, and the first Two combiners for processing. 27. The device according to item 15 of the scope of patent application, wherein the compressed written language data includes hypertext markup language (HTML) data. 2 8. The device according to item 15 of the scope of patent application, further comprising: a device for temporarily storing and compressing written language data in a proxy server for the written language content frequently accessed by the user terminal. Page 27