guix-play/gnu/packages/patches/libxslt-generated-ids.patch
Mathieu Lirzin c74c7c1317 gnu: libxslt: Make generated documentation reproducible.
* gnu/packages/patches/libxslt-generated-ids.patch: New file.
* gnu/packages/patches/libxslt-remove-date-timestamps.patch: Likewise.
* gnu/packages/xml.scm (libxslt)[source]: Use them.
* gnu/local.mk (dist_patch_DATA): Add them.
2016-05-02 17:54:01 +02:00

174 lines
4.8 KiB
Diff

This makes generated IDs deterministic.
Written by Daniel Veillard.
This should be fixed in next release (2.29).
See https://bugzilla.gnome.org/show_bug.cgi?id=751621.
diff --git a/libxslt/functions.c b/libxslt/functions.c
index 6448bde..5b00a6d 100644
--- a/libxslt/functions.c
+++ b/libxslt/functions.c
@@ -651,6 +651,63 @@ xsltFormatNumberFunction(xmlXPathParserContextPtr ctxt, int nargs)
}
/**
+ * xsltCleanupIds:
+ * @ctxt: the transformation context
+ * @root: the root of the resulting document
+ *
+ * This clean up ids which may have been saved in Element contents
+ * by xsltGenerateIdFunction() to provide stable IDs on elements.
+ *
+ * Returns the number of items cleaned or -1 in case of error
+ */
+int
+xsltCleanupIds(xsltTransformContextPtr ctxt, xmlNodePtr root) {
+ xmlNodePtr cur;
+ int count = 0;
+
+ if ((ctxt == NULL) || (root == NULL))
+ return(-1);
+ if (root->type != XML_ELEMENT_NODE)
+ return(-1);
+
+ cur = root;
+ while (cur != NULL) {
+ if (cur->type == XML_ELEMENT_NODE) {
+ if (cur->content != NULL) {
+ cur->content = NULL;
+ count++;
+ }
+ if (cur->children != NULL) {
+ cur = cur->children;
+ continue;
+ }
+ }
+ if (cur->next != NULL) {
+ cur = cur->next;
+ continue;
+ }
+ do {
+ cur = cur->parent;
+ if (cur == NULL)
+ break;
+ if (cur == (xmlNodePtr) root) {
+ cur = NULL;
+ break;
+ }
+ if (cur->next != NULL) {
+ cur = cur->next;
+ break;
+ }
+ } while (cur != NULL);
+ }
+
+fprintf(stderr, "Attributed %d IDs for element, cleaned up %d\n",
+ ctxt->nextid, count);
+
+ return(count);
+}
+
+/**
* xsltGenerateIdFunction:
* @ctxt: the XPath Parser context
* @nargs: the number of arguments
@@ -701,7 +758,39 @@ xsltGenerateIdFunction(xmlXPathParserContextPtr ctxt, int nargs){
if (obj)
xmlXPathFreeObject(obj);
- val = (long)((char *)cur - (char *)&base_address);
+ /*
+ * Try to provide stable ID for generated document:
+ * - usually ID are computed to be placed on elements via attributes
+ * so using the element as the node for the ID
+ * - the cur->content should be a correct placeholder for this, we use
+ * it to hold element node numbers in xmlXPathOrderDocElems to
+ * speed up XPath too
+ * - xsltCleanupIds() clean them up before handing the XSLT output
+ * to the API client.
+ * - other nodes types use the node address method but that should
+ * not end up in resulting document ID
+ * - we can enable this by default without risk of performance issues
+ * only the one pass xsltCleanupIds() is added
+ */
+ if (cur->type == XML_ELEMENT_NODE) {
+ if (cur->content == NULL) {
+ xsltTransformContextPtr tctxt;
+
+ tctxt = xsltXPathGetTransformContext(ctxt);
+ if (tctxt == NULL) {
+ val = (long)((char *)cur - (char *)&base_address);
+ } else {
+ tctxt->nextid++;
+ val = tctxt->nextid;
+ cur->content = (void *) (val);
+ }
+ } else {
+ val = (long) cur->content;
+ }
+ } else {
+ val = (long)((char *)cur - (char *)&base_address);
+ }
+
if (val >= 0) {
sprintf((char *)str, "idp%ld", val);
} else {
diff --git a/libxslt/functions.h b/libxslt/functions.h
index e0e0bf9..4a1e163 100644
--- a/libxslt/functions.h
+++ b/libxslt/functions.h
@@ -64,6 +64,13 @@ XSLTPUBFUN void XSLTCALL
int nargs);
/*
+ * Cleanup for ID generation
+ */
+XSLTPUBFUN int XSLTCALL
+ xsltCleanupIds (xsltTransformContextPtr ctxt,
+ xmlNodePtr root);
+
+/*
* And the registration
*/
diff --git a/libxslt/transform.c b/libxslt/transform.c
index 24f9eb2..2bdf6bf 100644
--- a/libxslt/transform.c
+++ b/libxslt/transform.c
@@ -700,6 +700,7 @@ xsltNewTransformContext(xsltStylesheetPtr style, xmlDocPtr doc) {
cur->traceCode = (unsigned long*) &xsltDefaultTrace;
cur->xinclude = xsltGetXIncludeDefault();
cur->keyInitLevel = 0;
+ cur->nextid = 0;
return(cur);
@@ -6092,6 +6093,13 @@ xsltApplyStylesheetInternal(xsltStylesheetPtr style, xmlDocPtr doc,
if (root != NULL) {
const xmlChar *doctype = NULL;
+ /*
+ * cleanup ids which may have been saved in Elements content ptrs
+ */
+ if (ctxt->nextid != 0) {
+ xsltCleanupIds(ctxt, root);
+ }
+
if ((root->ns != NULL) && (root->ns->prefix != NULL))
doctype = xmlDictQLookup(ctxt->dict, root->ns->prefix, root->name);
if (doctype == NULL)
diff --git a/libxslt/xsltInternals.h b/libxslt/xsltInternals.h
index 95e8fe6..8eedae4 100644
--- a/libxslt/xsltInternals.h
+++ b/libxslt/xsltInternals.h
@@ -1786,6 +1786,8 @@ struct _xsltTransformContext {
int funcLevel; /* Needed to catch recursive functions issues */
int maxTemplateDepth;
int maxTemplateVars;
+
+ unsigned long nextid;/* for generating stable ids */
};
/**