 |
Claims  |
|
|
The invention in which an exclusive right is claimed is defined by the
following:
1. A method for synchronizing a hypertext markup language (HTML) file
cached at a client location with a corresponding changed HTML file stored
at a server location to enable the changed HTML file stored at the server
location to be used at the client location without transmitting all of the
changed HTML file to the client location, comprising the steps of:
(a) extracting a macro name file and a macro definition file from the HTML
file cached at the client location, said macro definition file including a
definition that defines a content of at least a portion of the HTML file,
and said macro name file including a macro name associated with each said
definition;
(b) unless the macro name file is already available at the server location,
transmitting the macro name file from the client location to the server
location;
(c) generating a macro compressed file from the changed HTML file using the
macro name file at the server location, said macro compressed file
indicating changes made to the changed HTML file relative to the HTML file
cached at the client location and including the macro name for each
portion of the changed HTML file that is identical to a corresponding
portion of the HTML file cached at the client location;
(d) transmitting the macro compressed file to the client location from the
server location; and
(e) expanding the macro compressed file at the client location to recover
the changes and to identify any portion of the changed HTML file that is
identical to the HTML file cached at the client location, said changes
being applied to the HTML file cached at the client location to make a new
cached HTML file that is substantially identical to the changed HTML file
at the server location.
2. The method of claim 1, further comprising the step of using the new
cached HTML file to render a display at the client location.
3. The method of claim 1, wherein the HTML file cached at the client
location comprises a plurality of lists, each list representing one
portion of said file and having a checksum associated with it that is the
macro name for the list, said macro definition file comprising a macro
definition for each list that concisely defines a content for each list.
4. The method of claim 3, wherein the macro definition file comprises an
idiom dictionary that defines the lists comprising the HTML file cached at
the client location.
5. The method of claim 3, further comprising the step of providing a client
agent at the client location to extract the macro definition file and the
macro name file, said client agent responding to a request entered by a
user at the client location to access a site from which the HTML file
cached at the client location was obtained, by appending the macro name
file for said HTML file after an address for the site, and expanding the
macro compressed file that is received from the server location.
6. The method of claim 5, further comprising the step of providing a server
agent at the server location to fetch the changed HTML file from the
server location and produce the macro compressed file, said macro
compressed file comprising the macro name for any list in the changed HTML
file that is identical to the list in the HTML file cached at the client
location and the list defining any portion of the changed HTML file that
differs from that of the HTML file cached at the client location.
7. The method of claim 6, wherein the client agent expands the macro
compressed file by replacing macro names included in the macro compressed
file with a corresponding content determined from the macro definition
file.
8. The method of claim 7, further comprising the step of conveying the new
HTML file to a web browser for use in the display at the client location.
9. A system for updating an HTML file disposed at a first location so that
it is substantially identical to a changed HTML file disposed at a second
location, without transmitting all of the changed HTML file to the first
location from the second location, comprising:
(a) a client computer disposed at the first location, said client computer
including a display for displaying data, and a memory in which the HTML
file is cached and for storing machine instructions that effect a first
set of functions;
(b) a communication link coupling the first and the second locations in
communication with each other;
(c) a server computer disposed at the second location, said server computer
including a memory in which the changed HTML file is stored and for
storing machine instructions that implement a second set of functions;
(d) said first set of functions including:
(i) causing the client computer to extract a macro name file and a macro
definition file from the HTML file cached in the memory of the client
computer, said macro definition file including a definition that defines a
content of at least a portion of the HTML file, and said macro name file
including a macro name associated with each said definition;
(ii) unless macro name file is already available at the server computer,
transmitting the macro name file from the client computer to the server
computer; and
(iii) expanding a macro compressed file on the client computer, said macro
compressed file being received from the server computer, indicating
changes made to the changed HTML file relative to the HTML file cached by
the client computer, and including the macro name for each portion of the
changed HTML file that is identical to a corresponding portion of the HTML
file cached by the client computer, said changes being applied to the HTML
file cached in the memory of the client computer to make a new cached HTML
file that is substantially identical to the changed HTML file stored in
the memory of the server computer, said new cached HTML file being stored
in the memory of the client computer; and
(e) said second set of functions including:
(i) causing the server computer to generate the macro compressed file from
the changed HTML file using the macro name file, said macro compressed
file indicating the changes in the changed HTML file relative the HTML
file cached in the memory of the client computer; and
(ii) causing the server computer to transmit the macro compressed file to
the client computer.
10. The system of claim 9, wherein the new cached HTML file is used to
produce a displayed image on the display of the client computer at the
first location.
11. The system of claim 9, wherein the HTML file cached on the client
computer comprises a plurality of lists, each list having a checksum
associated with it that is the macro name for the list, said macro name
file comprising the macro names for the lists, said macro definition file
comprising a macro definition for each list that concisely defines a
content for each list.
12. The system of claim 11, wherein the macro definition file comprises an
idiom dictionary that defines the lists comprising the HTML file cached in
the memory of the client computer.
13. The system of claim 11, wherein the first set of functions effected by
the client computer further include producing a client agent on the client
computer to extract the macro definition file and the macro name file,
said client agent including the macro name file with an address for the
second location that is transmitted over the communications link, and
expanding the macro compressed file received from the server computer.
14. The system of claim 13, wherein the second set of functions effected by
the server computer further include providing a server agent on the server
computer to fetch the changed HTML file from the server computer and
produce the macro compressed file, said macro compressed file comprising
the macro name for any list in the changed HTML file that is identical to
the list in the HTML file cached by the client computer, and the
compressed file including any portion of the changed HTML file that
differs from that of the HTML file cached on the client computer.
15. The system of claim 14, wherein the client agent expands the macro
compressed file by replacing macro names included in the macro compressed
file with a corresponding content determined from the macro definition
file.
16. The system of claim 15, wherein the first set of functions include
passing the new HTML file to a web browser for use in displaying an image
on the display of the client computer.
17. An article of manufacture for use with a computer, comprising:
(a) a computer readable memory medium; and
(b) a plurality of machine instructions stored on said computer readable
memory medium, said machine instructions controlling a computer to
implement a plurality of functions, said functions including:
(i) extracting a macro name file and a macro definition file from a cached
HTML file;
(ii) as required, transmitting the macro name file to a different location,
said different location storing a changed HTML file; and
(iii) expanding a macro compressed file received from the different
location to recover changes in the changed HTML file, said changes being
applied to the cached HTML file to make a new cached HTML file that is
substantially identical to the changed HTML file stored at the different
location.
18. The article of manufacture of claim 17, wherein the machine
instructions cause the new cached HTML file to be used for displaying an
image.
19. An article of manufacture for use with a computer, comprising:
(a) a computer readable memory medium; and
(b) a plurality of machine instructions stored on said computer readable
memory medium, said machine instructions controlling a computer to
implement a plurality of functions, said functions including:
(i) generating a macro compressed file for a changed HTML file using a
macro name file derived from a corresponding HTML file that is cached at a
different location, said macro compressed file indicating changes in the
changed HTML file relative the HTML file cached at the different location;
and
(ii) transmitting the macro compressed file to the different location to
enable the cached HTML file to be updated with the changes.
20. The article of manufacture of claim 19, wherein the machine
instructions produce the macro compressed file by identifying changes in
the changed HTML file relative to the macro name file. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
FIELD OF THE INVENTION
The present invention generally relates to the transfer of information over
a network, and more specifically, to the transfer of hierarchically
structured files, such as hypertext markup language (HTML) files, over a
network to a recipient computer on which the files are cached.
BACKGROUND OF THE INVENTION
Ready access to a multitude of World Wide Web (WWW) sites over the Internet
has resulted in an astonishing growth in the amount of traffic and is
likely to cause increasing delays in the transfer of data between servers
and the client computers accessing them. This problem is particularly
troubling to those using relatively slow modems to access the web. High
speed access is often not readily available when the Web is being accessed
from a personal data assistant or other small, portable computing device.
Furthermore, even though higher speed modems are available for use by
conventional computers, many slower modems remain in use. Clearly, a
technique for increasing the efficiency with which HTML documents are
transferred over the Internet (and over other networks) by slower speed
devices would improve the effective data transfer rate, thereby benefiting
all users of the network.
To better appreciate how data transfer efficiency might be improved on the
Internet, it is helpful to understand the nature of the Web page documents
being transferred. The HTML language is used for writing hypertext
documents. These documents are more formally referred to as Standard
Generalized Markup Language (SGML) documents and they conform to a
particular Document Type Definition (DTD). An HTML document includes a
hierarchical set of markup elements. Most elements have a start tag,
followed by content, followed by an end tag. The content is a combination
of text and nested markup elements. Tags indicate how the document is
structured and how to display the document, as well as destinations and
labels for hypertext links. There are tags for markup elements such as
titles and headers, text attributes such as bold and italic, lists,
paragraph boundaries, links to other documents or other parts of the same
document, and in-line graphic images. The HTML language also enables an
HTML document to include images that are stored as separate files. When
the user views the HTML document, any included image is displayed as part
of the document, at the point where the image element occurred in the
document.
Although HTML files may include graphics, text in such files is more likely
to change in a dynamic fashion. One solution to this problem is to store
or "cache" HTML files that were previously transferred from the server
computers, so that the next time the user connects to a site recently
visited, the HTML file defining a web page for that Uniform Resource
Locator (URL) site can be loaded into the client computer display from the
cache stored on the client hard drive instead of being transferred over
the Internet. The user perceives a much faster loading of a page from a
cache, and less data need to be transferred over the Internet. However,
under conventional procedures, if any change in the text content of an
HTML file has occurred since the user last connected to a URL site, the
cached file on the client hard drive will normally be discarded, and the
entire revised HTML will be transferred from the server to the client
computer. The user must wait for the transfer to complete before all of
the content of the HTML file is fully visible. The server and the browser
software used by the client typically cooperate to detect whether it will
be necessary to transfer a new HTML file from the server to the client
computer or whether a cached HTML file stored on the client computer can
instead be used. Currently, there is no provision for splitting up an HTML
file so that only the portion that has changed since the file was last
cached is transferred from the server to the client.
To minimize the amount of data that must be transferred over the Internet
in order to display an HTML file, it would be desirable to provide a
mechanism for dividing an HTML file into a plurality of units, so that
only those units that have changed must be transferred to synchronize the
cached HTML file data on the client computer with that stored on the
server. Furthermore, in those cases where it is necessary to avoid the
need to modify existing web browser or web site server software, this
mechanism should occur in a manner that is transparent to the server and
browser software. By avoiding the need to modify server and web browser
software so as to accomplish this objective, it should be possible to
effect the desired result independent of the server and web browser
software programs used to access the HTML files. However, as new browser
and web site server software is developed, it would be preferable to
integrate this approach into the respective software so that it operates
more efficiently. By providing a method for transferring primarily the
changed portions of an HTML file, a substantial improvement in the data
transfer rate for web sites that have previously been accessed can be
achieved when the sites are again accessed. The result will improve the
apparent speed at which the HTML file for each such site is displayed on
the client computer and greatly reduce the amount of data transferred over
the net. Further, it will be apparent that this technique can be used when
transferring an HTML file over almost any type of network, including local
area and wide area networks and on intranets, between any two points that
are connected, such as between two sites on a wide area network.
SUMMARY OF THE INVENTION
The present system for rapidly transferring files includes four operative
components--a web browser, a client agent, a server agent, and a web
server. The communication between the web browser and the client agent is
fast. Similarly, the communication between the web server and server agent
is also fast. However, the link between the client agent and server agent
is relatively slow, being limited by the speed of a modem or data link
coupling the client agent to the server agent. The present invention
compensates for the slow data transfer rate over this link. Each component
can be run in a different computer or can be executed with another
component on the same computer. For example, the web browser and client
agent are very likely to be executed on the same client computer. The
server agent can be executed in the proxy server or combined with the web
server for execution on the server computer.
In accord with the present invention, a method is defined for transferring
changes to synchronize an HTML file cached at a client location with a
corresponding changed HTML file stored at a server location. The method
enables the changed HTML file stored at the server location to be used at
the client location without transmitting all of the changed HTML file to
the client location and includes the step of extracting a macro name file
and a macro definition file from the HTML file cached at the client
location. The macro definition file includes a definition that defines a
content of at least a portion of the HTML file, and the macro name file
includes a macro name associated with each such definition. In the next
step of the method, the macro name file is transmitted from the client
location to the server location. However, if the macro name file is
already available at the server location, it need not be transmitted. A
macro compressed file is generated at the server location from the changed
HTML file, using the macro name file. This macro compressed file indicates
changes made to the changed HTML file relative to the HTML file cached at
the client location and includes the macro name for each portion of the
changed HTML file that is identical to a corresponding portion of the HTML
file cached at the client location. The macro compressed file is
transmitted to the client location from the server location. The macro
compressed file is expanded at the client location to recover the changes
and to identify any portion of the changed HTML file that is identical to
the HTML file cached at the client location. These changes are applied to
the HTML file cached at the client location to make a new cached HTML file
that is substantially identical to the changed HTML file stored at the
server location.
Preferably, the method further comprises the step of using the new cached
HTML file to produce a display at the client location. Like a list
processing (LISP) language program, the HTML file cached at the client
location comprises a plurality of lists. Each list represents one portion
of the cached file and has a checksum associated with it that comprises
the macro name for the list. The macro definition file includes a macro
definition for each list that concisely defines a content for each list.
Further, the macro definition file comprises an idiom dictionary that
defines the lists for the HTML file cached at the client location. In
addition, the method includes the step of providing a client agent at the
client location to extract the macro definition file and the macro name
file. This client agent responds to a request entered by a user at the
client location to access a URL for a site corresponding to that from
which the HTML file cached at the client location was obtained, by
appending the macro name file for the HTML file after a URL address for
the site. Also, the client agent expands the macro compressed file that is
received from the server location. The client agent expands the macro
compressed file by replacing macro names included in the macro compressed
file with a corresponding content determined from the macro definition
file. The new HTML file is preferably passed to a web browser for use in
the display at the client location.
Another aspect of the present invention is directed to a system for
updating an HTML file disposed at a first location so that it is
substantially identical to a changed HTML file disposed at a second
location, without transmitting all of the changed HTML file to the first
location from the second location. The system implements functions
generally consistent with the steps of the method discussed above.
Still another aspect of the present invention is directed to articles of
manufacture that include a memory medium adapted to be read by a computer.
The memory medium stores machine instructions, which when executed by the
server computer and the client computer, cause those computers to
implement functions that are generally consistent with the steps of the
method discussed above.
BRIEF DESCRIPTION OF THE DRAWING FIGURES
The foregoing aspects and many of the attendant advantages of this
invention will become more readily appreciated as the same becomes better
understood by reference to the following detailed description, when taken
in conjunction with the accompanying drawings, wherein:
FIG. 1 is a schematic block diagram showing a server computer and several
client computers communicating over a network, such as the Internet;
FIG. 2 is a schematic block diagram of the internal components of a digital
computer on which the present invention is implemented when communicating
over the Internet and is applicable to either a server or a client
computer;
FIG. 3 is a flow chart showing the overall steps of the present method for
rapidly updating a cached HTML file stored on a client computer using a
macro compressed file that is transmitted from the server to the client;
and
FIG. 4 is a flow chart illustrating the steps used to produce the macro
name file and the macro definition file from the cached HTML file.
DESCRIPTION OF THE PREFERRED EMBODIMENT
FIG. 1 schematically illustrates a network 12 that includes three client
computers 14a, 14b, and 14c that are communicating through the network to
a server computer 16. Although many more client computers may be coupled
to server computer 16 through the network, for simplification of the
drawing and this discussion, only three client computers are illustrated,
and two of these are shown in phantom view to make clear that the present
invention arises in connection with communication between one client
computer and the server computer. Network 12 may comprise the Internet, a
local area network, a wide area network, or virtually any other type of
network in which client and servers communicate using HTML documents. The
present invention is particularly applicable to the Internet and to
intranet networks, since it facilitates communication of HTML documents
from the server to the client computer in order to update an HTML document
that was previously transferred from the server and is now stored on the
client computer.
As noted in the Background of the Invention, an object of the present
invention is to minimize the total amount of data communicated from server
16 to any of the client computers requesting an HTML document from the
server by making greater use of the data for the HTML document that is
cached on the client computer than is currently done. Each client computer
is generally conventional in its design, and may comprise a desktop
computer, a workstation, a laptop or portable computer, or any other
suitable digital computer having storage capacity for caching HTML
documents and an interface for coupling to the server over a network.
With reference to FIG. 1, each client computer includes a processor chassis
18 in which may be disposed a hard drive 22 and/or a floppy drive 24.
Generally, hard drive 22 will be used for nonvolatile storage of
application programs and data, including the caching of HTML document data
in connection with the present invention. It is contemplated that the
client computer may be a network workstation that does not include a hard
drive or may be a portable device such as a PDA that does not include
either a hard drive or a floppy. Each client computer will likely include
a monitor or display screen 20 on which HTML documents can be displayed
and presented to the user of the client computer.
Similarly, server computer 16 includes a hard drive 22 and may include a
floppy drive 24. Typically, the nonvolatile storage capacity of hard drive
22 on server computer 16 is substantially greater than that of the hard
drives on the client computers that are linked to the server. However, it
is also contemplated that a device such as a router, which performs the
function of a server computer, can be used and such a device may not
include a hard drive; because it accesses data stored on a different
device. Server computer 16 may optionally include a monitor or display
screen 20, but typically does not.
The link between each of the client computers and the server computer will
be through any of the links typically used for connecting digital
computers in a network, such as telephone or data lines. In any case, the
details employed in linking the client computers to the server computer in
network 12 are of little consequence in connection with the present
invention.
As shown in FIG. 2, the digital computer used for either the client
computer or the server computer includes a central processing unit (CPU)
26 that is bidirectionally coupled to hard drive 22 to store or access
previously stored application programs and/or data. CPU 26 is also coupled
to display 20. Although not shown, an appropriate video interface and
appropriate video driver software are used to enable the CPU to render
graphics and other aspects of an HTML document on display 20. A keyboard
and mouse 28 are connected to CPU 26 to provide user input to applications
running on the computer. Either the keyboard or the mouse can be employed
by the user to provide an input to CPU 26 that initiates connection of the
client computer to a specific site at which server computer 16 is
disposed, in order to access an HTML document stored on hard drive 22 of
server computer 16. The document is referenced by the URL of the site of
the server computer and by the document HTML file name.
Application programs such as an Internet browser are stored on hard drive
22 in the form of machine language instructions. When the user wants to
connect to the Internet, e.g., using a dial-up service provider or by some
other means (not shown), the user will typically run the Internet browser
program, such as Microsoft Corporation's Internet Explorer.TM.. The
machine instructions defining the Internet browser program are loaded from
hard drive 22 into a memory 30 for access by CPU 26. Memory 30 includes
both volatile random access memory (RAM) and read only memory (ROM). The
ROM portion of memory 30 is used during boot-up of the computer to
establish the basic input/output capabilities of the system, and the RAM
portion stores the machine instructions and other data of a temporary
nature that are used by CPU 26 in effecting the functions implemented when
the application is run on the computer. Similarly, the present invention
comprises a set of machine instructions that define the functions
implemented by CPU 26. The machine instructions implemented by the client
computer for the present invention initiate a client agent to effect the
rapid updating of an HTML file that is cached on hard drive 22 in the
client computer when a user selectively accesses a site from which the
HTML file was previously downloaded. Similarly, at the server computer,
machine instructions are loaded into memory that enable the server
computer to function as a web server and in connection with the present
invention, to provide a server agent by running the instructions.
Each site on the Internet is identified by a unique URL that enables it to
be found so that the client computer is coupled to the server computer at
that site to access one or more HTML documents stored on the hard drive of
the server computer. Similar techniques are used in connection with
intranets to access a specific HTML document, which might be stored on one
of several different intranet server computers. Furthermore, it should be
noted that the present invention can be used on two computers that are
linked together by any communication medium, which may not require that
the source of the HTML document be identified.
On client computer 14a, the client agent implements the functions described
below that enable HTML files cached on the client computer to be updated
with data from server computer 16. Similarly, on server computer 16, the
server agent implements the functions described below that are required on
the server side of the system.
In FIG. 3, the steps required for updating an HTML file cached on the
client computer and stored on hard drive 22 are illustrated in a flow
chart. The flow chart begins with a start block 40 and continues in a
block 42 in which the user requests connection to a URL site. Generally,
this request is made by entering the URL identifier for the site in a web
browser, or selecting the site from among a number of sites that are
stored as bookmarks for access by the web browser. Examples of browsers
that include this capability are Microsoft Corporation's Internet
Explorer.TM., and Netscape Corporation's Navigator.TM. web browsers. In
response to the request entered in the browser in block 42, a decision
block 44 determines if an HTML file for that URL site has been cached or
stored on hard drive 22. If the site was recently visited by the user and
the HTML file was then transferred from the server computer at the site to
the client computer, the HTML file will likely still be cached on the
client computer hard drive. If not, a block 46 indicates that the web
browser transmits the request for the HTML file to the web server at the
URL site. The request transmitted by the browser is sent over the Internet
(or other type of network) to the appropriate server, which responds as
indicated in a block 48 by sending the entire requested HTML file to the
web browser running on the client computer. Since the HTML file was not
cached on the client computer when requested by the user, the web browser
stores the HTML file for later use in a cache folder on the hard drive and
proceeds to a block 52 in which the web browser executes the HTML file to
render the document on the display of the client computer.
Those familiar with HTML documents will appreciate that such a web document
may include one or more hypertext links to other web sites, animations
running under ActiveX.TM. or Java.TM., forms, sound clips, and various
other components. New features are being developed for inclusion in HTML
documents at a rapid pace. As more features are added, the complexity and
the amount of data comprising an HTML file increases proportionally.
Accordingly, the present invention is particularly useful in that it
enables a substantially smaller amount of data to be transmitted to update
a cached HTML file when changes occur in the HTML file on the server
computer that are not included in the HTML file cached on the client
computer.
The next time that the user requests a connection to the same URL site to
access the HTML file stored there, decision block 44 determines that the
HTML file for that site has been cached on hard drive 22 of the client
computer. In response to this determination, the client agent extracts a
macro name file and a macro definition file from the cached HTML file, as
noted in a block 56. This step is a key aspect of the present invention
and further explanation is required to clearly describe it.
To understand how the present invention enables an HTML file cached on the
client computer to be rapidly updated, it is first necessary to understand
that HTML files are structured files, having an explicit beginning and end
for most of the constructs or components comprising the file. As noted
above, the specification defining HTML files is not fixed, but continues
to expand as new features and functionality are added. Because of their
structure, HTML files can be modeled in a way similar to LISP programs,
which also have explicit beginning and ends for each of the constructs
comprising them. Each construct in an HTML file comprises a list that is
enclosed by an open parenthesis and a close parenthesis. These lists can
be nested, but not interleaved. The technique for accelerating the
updating of HTML files applies a technique that is analogous to the
methods used for LISP file synchronization, but the technique is applied
in an entirely different manner and for a very different purpose.
Details of the steps referenced in block 56 from FIG. 3 are illustrated in
the flow chart shown in FIG. 4. Turning now to FIG. 4, a block 70
indicates that a cached HTML file (on the client computer) is opened. The
client computer then determines a macro name for each construct or list in
the cached HTML file. While it might be possible to use other schemes to
identify each construct, it is important that the method used to determine
the macro name be consistent and readily accomplished. For this reason,
the preferred embodiment of the present invention determines a macro name
that is equal to a checksum for each construct or list in the cached HTML
file. In connection with an HTML document, each construct may comprise a
different section of text, a graphic image, a Visual Basic script or Java
script component, a form, or virtually any other discrete portion of the
HTML document. The present invention is applicable to the Visual Basic
script and Java script components, which may be rather large, because
these components have a hierarchical structure, i.e., nested begin-end
structures. Since each construct or list is defined and represented in the
document by alphanumeric characters, a checksum of those characters for
each list can be determined for use as the macro name. A conventional
approach well known to those of ordinary skill in the art is used to
determine the checksum. It is extremely unlikely that any two different
constructs in a given HTML document would have identical checksums, so
that each macro name is generally unique and corresponds to a different
construct in the document.
In a block 74, the macro names for the cached HTML document are stored in a
macro name file in the same sequence that the lists or constructs to which
they correspond appear within the HTML document. These macro names
comprise an idiom set for the document. In a macro definition, each
sub-construct is represented by its macro name. A parent construct may
include a plurality of child constructs, and each has its own macro name
in the macro definition file. It should be apparent that the macro name
file identifies each construct or list in the HTML file and that the
checksums comprising the macro names in the macro name file are
substantially smaller in size than the constructs or lists that the macro
names represent and correspond to. Accordingly, the macro name file for a
document is much smaller in size than the data required to store the
original HTML document.
Corresponding to each macro name in the macro name file are macro
definitions. A macro definition is simply the content of that portion of
the body of the cached HTML file represented by a macro name. Thus, there
is a direct correspondence between each macro name and the macro
definition or portion of the HTML file represented by the macro name. In a
block 76, the macro definitions are extracted from the body of HTML file
for each different construct or list. Finally, in a block 78, a macro
definition file comprising the idiom dictionary is produced using the
macro definitions that were extracted. In a block 80, all of the files are
closed.
Returning back to FIG. 3, in a block 58, the client agent appends the macro
name file to the URL for the site at which the cached HTML file is stored
and transmits it to the web server at that site, over the network. The
existence of the client agent and its actions in producing the macro name
file and macro definition file are completely transparent to the web
browser, which need not be modified in any way to implement the present
invention. The client agent simply responds to the request made by the
user to access a web document using the web browser by insuring that the
macro name file for any HTML file being requested is appended to the URL
that is transmitted over the network from the client computer to the
server computer. However, as noted above, it may be desirable to integrate
the client agent into the program that implements the web browser. Also,
in certain applications of the present invention in which two end points
are coupled together in a communication link, it may not be necessary to
provide a URL for the storage site of the HTML file. It is further
contemplated that the server computer may have already produced the macro
name file for the HTML file cached by the client computer, e.g., when the
file was originally transmitted to the client computer. If so, it will not
be necessary for the client computer to send the macro name file to the
server computer.
In a block 60, the server agent, which is executed on the server computer,
fetches the HTML file that was requested by the URL from the hard drive on
the web server computer where it is stored and compresses the HTML file
using the macro name file supplied by the client agent. The step
implemented in block 60 produces a compressed file from the HTML file that
was obtained from the server. The compressed file is produced by replacing
each construct in the HTML file with the macro name that represents it, if
the macro name is also found in the macro name file received from the
client computer. If a construct in the HTML file that is obtained from the
server hard drive is not represented by a macro name in the macro name
file received from the client computer, then that construct or list is
inserted into the compressed file. If a parent construct does not match in
the changed HTML file, then as many child constructs for that portion of
the HTML file will be used as possible. The compressed file thus uses the
macro names provided by the client computer to represent the constructs or
lists that are already in the cached HTML file on the client computer. The
order in which the macro names are inserted into the compressed file may
be different than their order in the HTML file cached on the client
computer. Only those portions of the HTML files stored on the server
computer that have changed relative to the HTML file cached by the client
computer must be transmitted in full back to the client computer.
If the HTML file stored on the server computer is identical to that cached
by the client computer, the compressed file produced by the server agent
will include only the macro names received from the client agent. More
typically, the compressed file will comprise the macro names for the
portions of the HTML files stored on the server computer that have not
changed interspersed with the full constructs or lists for those sections
of the HTML file that have changed. It will be apparent that the size of
the compressed file will become larger as a function of the differences
between the HTML file stored on the server computer and that cached by the
client computer, i.e., the more changes that have occurred, the larger
will be the compressed file. If no change has occurred in the HTML file,
the compressed file could include simply the macro name representing the
text for the entire HTML file as one const | | |
