strings:Text2Xml

Converts a flat sequence of text characters into an XML document that represents punctuation marks and whitespace by special elements and attributes. This is a starting point for parsing the text for some purpose.

Syntax

Namespace declaration

You need to declare this namespace only if you call the function directly from the USCSXSL component but outside Blend. In a Blend context (ublendit.exe, uscsxsl.blend()) the namespace is already declared in the Blend transformation, which is in xsl\Util\Batch.1.0.xsl in your USoft installation folder.

xmlns:strings="USoft:Strings"

Function call

strings:Text2Xml( text, 'boundary-aliases' )

boundary-aliases  ::=  page-alias line-alias element-alias

The required boundary-aliases is a quoted string of 3 character sequences separated by 2 spaces. For details, see the Example and the "Boundary aliases" section below.

Example

<root xmlns:pc="Processing.Command" pc:hideme="true">
  <pc:defs>
    URE..TEXT. Application
      MYAPP
  </pc:defs>
  <pc:evaluate select="strings:Text2Xml(//pc:defs, 'P l elem')"/>
</root>

This example returns the following XML document. For comments, see the sections below the XML document.

<P>
  <l>
    <elem ws="">URE</elem>
    <pm ws="">.</pm>
    <pm ws="">.</pm>
    <elem ws="">TEXT</elem>
    <pm ws="s">.</pm>
    <elem ws="ntt">Application</elem>
    <eol/>
  </l>
  <l>
    <elem ws="nt">MYAPP</elem>
    <eol/>
  </l>
  <l>
    <eof/>
  </l>
</P>

Boundary aliases

With the boundary-aliases setting you can influence what the root element, each of the line elements, and each of the text elements is called in the result document.

The value of boundary-aliases is a string of 3 character sequences separated by 2 spaces, in the example:

P l elem

The first or page-alias sequence is the name of the root element of the return document.

The second or line-alias is the element name used for each section in the document that represents a text line.

The third or element-alias is the element name used for each text element other than punctuation marks.

Punctuation marks

Punctuation marks include full stops ( . ), commas ( , ), colons ( : ), and semicolons ( ; ). Each punctuation mark is parsed as an individual symbols. In contrast to non-punctuation text, shown in elements by the name of the page-alias boundary alias, a punctuation mark is shown in a <pm> element, for example:

<pm ws="">.</pm>

Whitespace

Each text element has a ws attribute representing all the whitespace characters (if any exist)  following that text in the original. The value of a ws attribute is made up of one or more of the following placeholders for whitespace characters:

For example, a text element followed by 1 newline and 2 tabs has the following ws attribute:

ws="ntt"

A text element that has no whitespace following it has an empty ws attribute:

ws=""

Notice that whitespace preceding the FIRST text element of the input is lost in the return document, as the example in this help topic shows.

Special elements

The following special elements are infixed in the result document in the places stated: