catalogLoad: Manipulate XML catalog contents

Description

These functions allow the R user to programmatically control the XML catalog table used in the XML parsing tools in the C-level libxml2 library and hence in R packages that use these, e.g. the XML and Sxslt packages. Catalogs are consulted whenever an external document needs to be loaded. XML catalogs allow one to influence how such a document is loaded by mapping document identifiers to alternative locations, for example to refer to locally available versions. They support mapping URI prefixes to local file directories/files, resolving both SYSTEM and PUBLIC identifiers used in DOCTYPE declarations at the top of an XML/HTML document, and delegating resolution to other catalog files. Catalogs are written using an XML format. Catalogs allow resources used in XInclude nodes and XSL templates to refer to generic network URLs and have these be mapped to local files and so avoid potentially slow network retrieval. Catalog files are written in XML We might have a catalog file that contains the XML In the XDynDocs package, we refer to OmegahatXSL files and DocBook XSL files have a catalog file of the form

The functions provided here allow the R programmer to empty the current contents of the global catalog table and so start from scratch ( catalogClearTable ), load the contents of a catalog file into the global catalog table ( catalogLoad ), and to add individual entries programmatically without the need for a catalog table.

In addition to controlling the catalogs via these functions, we can use catalogResolve to use the catalog to resolve the name of a resource and map it to a local resource.

catalogDump allows us to retrieve an XML document representing the current contents of the in-memory catalog . More information can be found at http://xmlsoft.org/catalog.html and http://www.sagehill.net/docbookxsl/Catalogs.html among many resources and the specification for the catalog format at http://www.oasis-open.org/committees/entity/spec-2001-08-06.html.

Usage

catalogLoad(fileNames)
catalogClearTable()
catalogAdd(orig, replace, type = "rewriteURI")
catalogDump(fileName = tempfile(), asText = TRUE)

Arguments

orig

a character vector of identifiers, e.g. URIs, that are to be mapped to a different name via the catalog. This can be a named character vector where the names are the original URIs and the values are the corresponding rewritten values.

replace

a character vector of the rewritten or resolved values for the identifiers given in orig. Often this omitted and the original-rewrite pairs are given as a named vector via orig.

type

a character vector with the same length as orig (or recycled to have the same length) which specifies the type of the resources in the elements of orig. Valid values are rewriteURI, rewriteSystem, system, public.

fileNames

a character vector giving the names of the catalog files to load.

fileName

the name of the file in which to place the contents of the current catalog

asText

a logical value which indicates whether to write the catalog as a character string if filename is not specified.

Value

These functions are used for their side effects on the global catalog table maintained in C by libxml2. Their return values are logical values/vectors indicating whether the particular operation were successful or not.

References

This provides an R-like interface to a small subset of the catalog API made available in libxml2.

Examples

Run this code

# Add a rewrite rule
# 
# 	
catalogAdd(c("http://www.omegahat.net/XML" = system.file("XML", package
= "XML")))
catalogAdd("http://www.omegahat.net/XML", system.file("XML", package =
"XML"))
catalogAdd("http://www.r-project.org/doc/",
           paste(R.home(), "doc", "", sep = .Platform$file.sep))
	
#
#          This shows how we can load a catalog and then resolve a
#          systemidentifier that it maps.
# 	
catalogLoad(system.file("exampleData", "catalog.xml", package = "XML"))
catalogResolve("docbook4.4.dtd", "system")
catalogResolve("-//OASIS//DTD DocBook XML V4.4//EN", "public")