seedxml

A reckoner seed: xml's parse and write, a node as a tagged list — machines/seeds/seedxml.shoddy

Summary

seedxml bridges xml into the reckoner, Shoddy’s stack calculator. Rule R4.16 says there is no TREE cell at the keyboard. So an Xml node converts to a tagged LIST, one shape per case: { "elem" tag attrs kids }, { "comment" text } and so on. The one exception is XText, which travels as a plain STRING. An element’s attributes are a dict — a list of name-and-value pairs — the same shape seeddict already established. XcOfXml/XcToXml, the conversion functions, are Included by seedhtml rather than duplicated, since html.shoddy answers the same Xml type.

Why It's Useful

The guard this seed adds is not the parse — it is everything around it. xml.shoddy’s reader already answers a Result rather than aborting on bad syntax. And XMLLOAD and XMLSAVE go through TRYREADFILE and TRYWRITEFILE rather than the aborting whole-file builtins. So a path that is not there, a directory handed where a file was meant, or an unwritable destination is a refusal or an honest False.

The larger thing it buys is that a parsed tree needs no words of its own. An element is a tagged LIST and its attributes are the same LIST-of-PAIR dict seeddict established. So the dictionary a session already has walks it:

FIRST and NTH take the tag or the children.
DGET reads an attribute.
MAP and FILTER run over the child list.

A seed that invented XMLATTR would have added a second spelling of a word that was already there.

User's Guide

Parse, and write back out. The round trip is exact for anything xml itself will accept:

> "<note><to>batley</to></note>" XMLPARSE XMLTEXT
x: "<note><to>batley</to></note>"
> CLEAR "<note><to>batley</to></note>" XMLPARSE 2 XMLPRETTY
x: "<note>
  <to>batley</to>
</note>
"

Attributes need single quotes when they are typed. The reckoner’s tokenizer — the part that cuts a line into words — has no escape inside a string literal. So there is no way to put a " inside one, and the usual XML spelling of an attribute is quoted. XML accepts either quote. The apostrophe form is therefore the one that can be typed, and it is written back out in the conventional form:

> "<a b='1'>hi</a>" XMLPARSE XMLTEXT
x: "<a b="1">hi</a>"

For anything longer than a line, XMLLOAD is the way in and the quoting question does not arise.

The two refusals come from different places and say so. The first is xml’s own reader naming where the syntax went wrong. The second is the guarded read:

> "<a><b></a>" XMLPARSE
?: XMLPARSE: </a> CLOSES <b> AT 7
> "nope.xml" XMLLOAD
?: XMLLOAD: cannot read the file

Word Reference

Word	Description
XMLPARSE ( str -- doc )	The string, parsed as XML — an element is { "elem" tag attrs kids }, attrs a dict.
XMLTEXT ( doc -- str )	The value, written out as XML text.
XMLPRETTY ( doc n -- str )	The value, written out as XML text indented n spaces per level.
XMLLOAD ( path -- doc )	The file, read and parsed as XML — a refusal names why it could not be, whether that is the file or the syntax.
XMLSAVE ( path doc -- ok )	Write the value to path as XML text; True on success.

Who Uses It

	User	How
	halifax	The calculator's XML words: `XMLPARSE`, `XMLTEXT`, `XMLPRETTY`, `XMLLOAD` and `XMLSAVE`.
	seedhtml	Includes this seed to reuse `XcOfXml`/`XcToXml` for the same `Xml` type html.shoddy answers.
	sparky	Sparky folds it too, so a model calling `eval` reaches the same words halifax puts at a prompt.

The Machines It Uses

	Machine	Why
	cuttle	The `Cell` type every bridged word reads its arguments from and answers into.
	reckoner	`RckReg`, `RckSeeding` and the argument readers every registered word is built from.
	seq	List plumbing under the tree converters.
	xml	The domain this seed bridges: `XmlRead`, `XmlText`, `XmlPretty`.