US 6,635,088 B1Granted

Structured document and document type definition compression

AssigneeInternational Business Machines Corporation

InventorsJohn Raithel Hind, David Bruce Lection

Life of this patent

Estimating the term…

Loading patent…

Abstract

A method, system, and computer-readable code for reducing Extensible Markup Language (XML) and Document Type Definition (DTD) document file size through one or more novel compression techniques. In one aspect, a compression technique for tags is defined that reduces the size of tags within either or both of XML and DTD files. In another aspect, a compression technique for attributes within tags may be used to further reduce file size (again, in either XML or DTD files). In a further aspect, a compression technique for strings within an XML file is defined, whereby the decompression will be performed automatically by an XML parser. While these compression techniques are described with reference to XML, they may also be used advantageously with other notations derived from SGML (Standard Generalized Markup Language).

Claims9

Dependent claims

In a computing environment, computer readable code readable by a computer system in said environment, for reducing document file size by string compression, comprising:

a subprocess for reading an input file containing a structured document encoded in a derivative of Standard Generalized Markup Language (SGML) to locate each of a plurality of strings in said structured document, wherein said strings appear between tags in said structured document;
a subprocess for programmatically generating a unique entity name reference corresponding to each unique one of said located strings and substituting said corresponding programmatically-generated entity name reference for said located string, thereby creating a compressed version of said input file, provided that a first cost of substituting said entity name reference is less than a second cost of using said located string without substitution, and provided that said located string contains no embedded entity references;
and
a subprocess for programmatically creating an entity declaration and inserting said programmatically-created entity declaration into said compressed version of said input file for each of said unique entity name references, said entity declaration declaring said correspondence between said located string and said corresponding entity name reference, wherein entity declarations and entity name references are standard features of said derivative, and said inserted entity declarations and said substituted entity name references enable transparently recreating said structured document, by a standard parser for said derivative, using said compressed version of said input file.

1 more element

A system for reducing document file size by string compression in a computing environment, comprising:

means for reading an input file containing a structured document encoded in a derivative of Standard Generalized Markup Language (SGML) to locate each of a plurality of strings in said structured document, wherein said strings appear between tags in said structured document;
means for programmatically generating a unique entity name reference corresponding to each unique one of said located strings and substituting said corresponding programmatically-generated entity name reference for said located string, thereby creating a compressed version of said input file, provided that a first cost of substituting said entity name reference is less than a second cost of using said located string without substitution, and provided that said located string contains no embedded entity references;
and
means for programmatically creating an entity declaration and inserting said programmatically-created entity declaration into said compressed version of said input file for each of said unique entity name references, said entity declaration declaring said correspondence between said located string and said corresponding entity name reference, wherein entity declarations and entity name references are standard features of said derivative, and said inserted entity declarations and said substituted entity name references enable transparently recreating, by a standard parser for said derivative, said structured document using said compressed version of said input file.

1 more element

A method for reducing document file size by string compression in a computing environment, comprising the steps of:

reading an input file containing a structured document encoded in a derivative of Standard Generalized Markup Language (SGML) to locate each of a plurality of strings in said structured document, wherein said strings appear between tags in said structured document;
programmatically generating a unique entity name reference corresponding to each unique one of said located strings and substituting said corresponding programmatically-generated entity name reference for said located string, thereby creating a compressed version of said input file, provided that a first cost of substituting said entity name reference is less than a second cost of using said located string without substitution, and provided that said located string contains no embedded entity references;
and
programmatically creating an entity declaration and inserting said programmatically-created entity declaration into said compressed version of said input file for each of said unique entity name references, said entity declaration declaring said correspondence between said located string and said corresponding entity name reference, wherein entity declarations and entity name references are standard features of said derivative, and said inserted entity declarations and said substituted entity name references enable transparently recreating said structured document, by a standard parser for said derivative, using said compressed version of said input file.

1 more element

Dependencies are inferred from the claim text.