Files
CXML/xml/xmls-compat.lisp
2008-02-24 19:03:48 +00:00

261 lines
8.8 KiB
Common Lisp

;;;; xml-compat.lisp -- XMLS-compatible data structures
;;;;
;;;; This file is part of the CXML parser, released under Lisp-LGPL.
;;;; See file COPYING for details.
;;;;
;;;; Developed 2004 for headcraft - http://headcraft.de/
;;;; Copyright: David Lichteblau
(defpackage cxml-xmls
(:use :cl :runes)
(:export #:make-node #:node-name #:node-ns #:node-attrs #:node-children
#:make-xmls-builder #:map-node #:make-xpath-navigator))
(in-package :cxml-xmls)
;;;; Knoten
(defun make-node (&key name ns attrs children)
`(,(if ns (cons name ns) name)
,attrs
,@children))
(defun node-name (node)
(let ((car (car node)))
(if (consp car)
(car car)
car)))
(defun (setf node-name) (newval node)
(let ((car (car node)))
(if (consp car)
(setf (car car) newval)
(setf (car node) newval))))
(defun node-ns (node)
(let ((car (car node)))
(if (consp car)
(cdr car)
nil)))
(defun (setf node-ns) (newval node)
(let ((car (car node)))
(if (consp car)
(setf (cdr car) newval)
(setf (car node) (cons car newval)))
newval))
(defun node-attrs (node)
(cadr node))
(defun (setf node-attrs) (newval node)
(setf (cadr node) newval))
(defun node-children (node)
(cddr node))
(defun (setf node-children) (newval node)
(setf (cddr node) newval))
;;;; SAX-Handler (Parser)
(defclass xmls-builder (sax:default-handler)
((element-stack :initform nil :accessor element-stack)
(root :initform nil :accessor root)
(include-default-values :initform t
:initarg :include-default-values
:accessor include-default-values)
(include-namespace-uri :initform t
:initarg :include-namespace-uri
:accessor include-namespace-uri)))
(defun make-xmls-builder (&key (include-default-values t)
(include-namespace-uri t))
"Make a XMLS style builder. When 'include-namespace-uri is true a modified
XMLS tree is generated that includes the element namespace URI rather than
the qualified name prefix and also includes the namespace URI for attributes."
(make-instance 'xmls-builder
:include-default-values include-default-values
:include-namespace-uri include-namespace-uri))
(defmethod sax:end-document ((handler xmls-builder))
(root handler))
(defmethod sax:start-element
((handler xmls-builder) namespace-uri local-name qname attributes)
(let* ((include-default-values (include-default-values handler))
(include-namespace-uri (include-namespace-uri handler))
(attributes
(loop
for attr in attributes
for attr-namespace-uri = (sax:attribute-namespace-uri attr)
for attr-local-name = (sax:attribute-local-name attr)
when (and (or (sax:attribute-specified-p attr)
include-default-values)
#+(or)
(or (not include-namespace-uri)
(not attr-namespace-uri)
attr-local-name))
collect
(list (cond (include-namespace-uri
(cond (attr-namespace-uri
(cons attr-local-name attr-namespace-uri))
(t
(sax:attribute-qname attr))))
(t
(sax:attribute-qname attr)))
(sax:attribute-value attr))))
(namespace (when include-namespace-uri namespace-uri))
(node (make-node :name local-name
:ns namespace
:attrs attributes))
(parent (car (element-stack handler))))
(if parent
(push node (node-children parent))
(setf (root handler) node))
(push node (element-stack handler))))
(defmethod sax:end-element
((handler xmls-builder) namespace-uri local-name qname)
(declare (ignore namespace-uri local-name qname))
(let ((node (pop (element-stack handler))))
(setf (node-children node) (reverse (node-children node)))))
(defmethod sax:characters ((handler xmls-builder) data)
(let* ((parent (car (element-stack handler)))
(prev (car (node-children parent))))
;; Be careful to accept both rods and strings here, so that xmls can be
;; used with strings even if cxml is configured to use octet string rods.
(if (typep prev '(or rod string))
;; um entities herum wird SAX:CHARACTERS mehrfach aufgerufen fuer
;; den gleichen Textknoten. Hier muessen wir den bestehenden Knoten
;; erweitern, sonst ist das Dokument nicht normalisiert.
;; (XXX Oder sollte man besser den Parser entsprechend aendern?)
(setf (car (node-children parent))
(concatenate `(vector ,(array-element-type prev))
prev
data))
(push data (node-children parent)))))
;;;; SAX-Treiber (fuer Serialisierung)
(defun map-node
(handler node
&key (include-xmlns-attributes sax:*include-xmlns-attributes*)
(include-namespace-uri t))
(if include-namespace-uri
(map-node/lnames (cxml:make-namespace-normalizer handler)
node
include-xmlns-attributes)
(map-node/qnames handler node include-xmlns-attributes)))
(defun map-node/lnames (handler node include-xmlns-attributes)
(sax:start-document handler)
(labels ((walk (node)
(unless (node-ns node)
(error "serializing with :INCLUDE-NAMESPACE-URI, but node ~
was created without namespace URI"))
(let* ((attlist
(compute-attributes/lnames node include-xmlns-attributes))
(uri (node-ns node))
(lname (node-name node))
(qname lname) ;let the normalizer fix it
)
(sax:start-element handler uri lname qname attlist)
(dolist (child (node-children node))
(typecase child
(list (walk child))
((or string rod)
(sax:characters handler (string-rod child)))))
(sax:end-element handler uri lname qname))))
(walk node))
(sax:end-document handler))
(defun map-node/qnames (handler node include-xmlns-attributes)
(sax:start-document handler)
(labels ((walk (node)
(when (node-ns node)
(error "serializing without :INCLUDE-NAMESPACE-URI, but node ~
was created with a namespace URI"))
(let* ((attlist
(compute-attributes/qnames node include-xmlns-attributes))
(qname (string-rod (node-name node)))
(lname (nth-value 1 (cxml::split-qname qname))))
(sax:start-element handler nil lname qname attlist)
(dolist (child (node-children node))
(typecase child
(list (walk child))
((or string rod)
(sax:characters handler (string-rod child)))))
(sax:end-element handler nil lname qname))))
(walk node))
(sax:end-document handler))
(defun compute-attributes/lnames (node xmlnsp)
(remove nil
(mapcar (lambda (a)
(destructuring-bind (name value) a
(unless (listp name)
(setf name (cons name nil)))
(destructuring-bind (lname &rest uri) name
(cond
((not (equal uri "http://www.w3.org/2000/xmlns/"))
(sax:make-attribute
;; let the normalizer fix the qname
:qname (if uri
(string-rod (concatenate 'string
"dummy:"
lname))
(string-rod lname))
:local-name (string-rod lname)
:namespace-uri uri
:value (string-rod value)
:specified-p t))
(xmlnsp
(sax:make-attribute
:qname (string-rod
(if lname
(concatenate 'string "xmlns:" lname)
"xmlns"))
:local-name (string-rod lname)
:namespace-uri uri
:value (string-rod value)
:specified-p t))))))
(node-attrs node))))
(defun compute-attributes/qnames (node xmlnsp)
(remove nil
(mapcar (lambda (a)
(destructuring-bind (name value) a
(when (listp name)
(error "serializing without :INCLUDE-NAMESPACE-URI, ~
but attribute was created with a namespace ~
URI"))
(if (or xmlnsp
(not (cxml::xmlns-attr-p (string-rod name))))
(sax:make-attribute :qname (string-rod name)
:value (string-rod value)
:specified-p t)
nil)))
(node-attrs node))))
;;;; XPath
(defun make-xpath-navigator ()
(make-instance 'xpath-navigator))
(defclass xpath-navigator ()
((parents :initform (make-hash-table))
(prefixes :initform (make-hash-table))
(children :initform (make-hash-table))
(attributes :initform (make-hash-table))
(namespaces :initform (make-hash-table))))
(defmethod initialize-instance :after ((instance xpath-navigator) &key)
(with-slots (prefixes) instance
(setf (gethash "http://www.w3.org/XML/1998/namespace" prefixes)
"xml")))