[Bindings] accept simple string in string<->xmlNode converter

Some use case ask for passing simple libxml content node (i.e just an
UTF-8 string) when a method argument or a field of the xmlNode* type.
This commit add a static method in bindings/utils.c named
lasso_string_fragment_to_xmlnode which does this transform by trying to
parse an XML document then by trying to parse a well balanced XML
fragment of only one node (if there is more than one node such as in the
string " xxx <tag/> yyy ", we free the node list and return NULL).

6 files changed, 69 insertions, 45 deletions

diff --git a/bindings/Makefile.am b/bindings/Makefile.am
index fde10816..42ccf450 100644
--- a/bindings/Makefile.am
+++ b/bindings/Makefile.am
@@ -21,5 +21,6 @@ EXTRA_DIST = bindings.py \
 	overrides.xml \
 	utils.py \
 	utility-scripts/error-analyzer.pl \
-	ghashtable.h
+	ghashtable.h \
+	utils.c
 
diff --git a/bindings/java/wrapper_top.c b/bindings/java/wrapper_top.c
index 5dcb5fda..e83a2737 100644
--- a/bindings/java/wrapper_top.c
+++ b/bindings/java/wrapper_top.c
@@ -7,6 +7,7 @@
 #include <string.h>
 #include "../ghashtable.h"
 #include "../../lasso/utils.h"
+#include "../utils.c"
 
 #define LASSO_ROOT "com/entrouvert/lasso/"
 #define check_exception (*env)->ExceptionCheck(env)
@@ -324,18 +325,10 @@ jstring_to_xml_node(JNIEnv *env, jstring jstr, xmlNode **xmlnode) {
     lasso_return_val_if_fail(jstring_to_local_string(env, jstr, &local_str), 0);
 
     if (local_str) {
-        doc = xmlReadDoc((unsigned char *)local_str, NULL, NULL, XML_PARSE_NONET);
-        if (!doc) {
-            exception(env, "could not read an xml document");
-            ret = 0;
-            goto out;
-        }
-        node = xmlDocGetRootElement(doc);
+        node = lasso_string_fragment_to_xmlnode(local_str, 0);
     }
-out:
-    lasso_assign_xml_node(*xmlnode, node)
-    if (doc)
-        lasso_release_doc(doc);
+    lasso_assign_new_xml_node(*xmlnode, node)
+    lasso_release_doc(doc);
     if (jstr && local_str)
         release_local_string(env, jstr, local_str);
     return ret;
diff --git a/bindings/perl/glist_handling.c b/bindings/perl/glist_handling.c
index d36c59b0..4cb09740 100644
--- a/bindings/perl/glist_handling.c
+++ b/bindings/perl/glist_handling.c
@@ -27,6 +27,7 @@
 #include <glib-object.h>
 #include <lasso/xml/xml.h>
 #include <lasso/utils.h>
+#include "../utils.c"
 
 /**
  * xmlnode_to_pv:
@@ -67,23 +68,15 @@ xmlnode_to_pv(xmlNode *node, gboolean do_free)
 
 static xmlNode *
 pv_to_xmlnode(SV *value) {
-	char *string;
-	xmlDoc *doc;
-	xmlNode *node = NULL;
+	int size;
 
 	if (! SvPOK(value))
 		return NULL;
-	string = SvPV_nolen(value);
+	string = SvPV(value, len);
 	if (! string)
 		return NULL;
 
-	doc = xmlReadDoc(BAD_CAST string, NULL, NULL, XML_PARSE_NONET);
-	if (! doc)
-		return NULL;
-	lasso_assign_xml_node(node, xmlDocGetRootElement(doc));
-	lasso_release_doc(doc);
-
-	return node;
+	return lasso_string_fragment_to_xmlnode(string, len);
 }
 
 /**
diff --git a/bindings/php5/wrapper_source_top.c b/bindings/php5/wrapper_source_top.c
index ebb8eb62..1b6bec7e 100644
--- a/bindings/php5/wrapper_source_top.c
+++ b/bindings/php5/wrapper_source_top.c
@@ -8,6 +8,7 @@
 #include "php_lasso.h"
 #include "../ghashtable.h"
 #include "../../lasso/utils.h"
+#include "../utils.c"
 
 /* utility functions */
 static void free_glist(GList **list, GFunc free_function);
@@ -126,17 +127,7 @@ get_string_from_xml_node(xmlNode *xmlnode)
 static xmlNode*
 get_xml_node_from_string(char *string)
 {
-	xmlDoc *doc;
-	xmlNode *node;
-
-	doc = xmlReadDoc((xmlChar*)string, NULL, NULL, XML_PARSE_NONET);
-	node = xmlDocGetRootElement(doc);
-	if (node != NULL) {
-		node = xmlCopyNode(node, 1);
-	}
-	lasso_release_doc(doc);
-
-	return node;
+	return lasso_string_fragment_to_xmlnode(string, 0);
 }
 
 static GList*
diff --git a/bindings/python/wrapper_top.c b/bindings/python/wrapper_top.c
index 76d9da95..7879bb13 100644
--- a/bindings/python/wrapper_top.c
+++ b/bindings/python/wrapper_top.c
@@ -5,6 +5,7 @@
 #include "../ghashtable.h"
 #include "../../lasso/debug.h"
 #include "../../lasso/utils.h"
+#include "../utils.c"
 
 #if PY_VERSION_HEX < 0x02050000 && !defined(PY_SSIZE_T_MIN)
 typedef int Py_ssize_t;
@@ -323,18 +324,10 @@ failure:
 
 static xmlNode*
 get_xml_node_from_pystring(PyObject *string) {
-	xmlDoc *doc;
-	xmlNode *node;
-
-	doc = xmlReadDoc((xmlChar*)PyString_AsString(string), NULL, NULL, XML_PARSE_NONET);
-	node = xmlDocGetRootElement(doc);
-	if (node != NULL) {
-		node = xmlCopyNode(node, 1);
-	}
-	lasso_release_doc(doc);
-
-	return node;
+	return lasso_string_fragment_to_xmlnode(PyString_AsString(string),
+			PyString_Size(string));
 }
+
 /** Return a tuple containing the string contained in a_list */
 static PyObject *
 get_list_of_strings(const GList *a_list) {
diff --git a/bindings/utils.c b/bindings/utils.c
new file mode 100644
index 00000000..d7dd4934
--- /dev/null
+++ b/bindings/utils.c
@@ -0,0 +1,53 @@
+/*
+ * In this file we put utility functions shared by all bindings.
+ *
+ * They usually are data structure manipulation or conversion functions.
+ */
+#include <libxml/tree.h>
+#include "../lasso/utils.h"
+
+/**
+ * lasso_string_fragment_to_xmlnode:
+ * @fragment: a fragment of an XML document
+ * @size: 
+ *
+ * Try to get one and only one node from a string, the node can be a simple string or a single node.
+ *
+ * Return value: a newly allocated xmlNode* or NULL if parsing fails.
+ */
+static xmlNode*
+lasso_string_fragment_to_xmlnode(const char *fragment, int size) {
+	xmlDoc *doc = NULL;
+	xmlNode *node = NULL;
+	xmlNode *list = NULL, *ref = NULL;
+	xmlParserErrors errors;
+
+	if (size == 0) {
+		size = strlen(fragment);
+	}
+
+	/* single node case, with preceding or following spaces */
+	doc = xmlReadMemory(fragment, size, NULL, NULL, XML_PARSE_NONET);
+	if (doc) {
+		node = xmlDocGetRootElement(doc);
+		if (node != NULL) {
+			node = xmlCopyNode(node, 1);
+			goto cleanup;
+		}
+		lasso_release_doc(doc);
+	}
+	/* simple string */
+	doc = xmlNewDoc(BAD_CAST "1.0");
+	ref = xmlNewNode(NULL, BAD_CAST "root");
+
+	xmlDocSetRootElement(doc, ref);
+	errors  = xmlParseInNodeContext(ref, fragment, size,
+		XML_PARSE_NONET, &list);
+	if (errors == XML_ERR_OK && list != NULL && list->next == NULL) {
+		node = xmlCopyNode(list, 1);
+	}
+cleanup:
+	lasso_release_doc(doc);
+	lasso_release_xml_node_list(list);
+	return node;
+}