parser: Always throw entity errors if external DTD is loaded

When parsing with XML_PARSE_DTDLOAD, missing entities are always an error. Also consolidate behavior when validating. See b717abdd.
2024-10-26 12:25:09 +03:00 · 2024-05-02 16:23:04 +02:00 · 2024-05-02 16:23:04 +02:00 · fdc5ff3657
commit fdc5ff3657
parent 39e5b35bd0
9 changed files with 38 additions and 57 deletions
--- a/parser.c
+++ b/parser.c
@ -7602,11 +7602,22 @@ xmlLookupGeneralEntity(xmlParserCtxtPtr ctxt, const xmlChar *name, int inAttr) {
     * standalone='yes'.
     */
    if (ent == NULL) {
-	if ((ctxt->standalone == 1) ||
+	if (((!ctxt->validate) && (ctxt->loadsubset)) ||
+            (ctxt->standalone == 1) ||
 	    ((ctxt->hasExternalSubset == 0) &&
 	     (ctxt->hasPErefs == 0))) {
 	    xmlFatalErrMsgStr(ctxt, XML_ERR_UNDECLARED_ENTITY,
 		     "Entity '%s' not defined\n", name);
+        } else if (ctxt->validate) {
+            /*
+             * [ VC: Entity Declared ]
+             * In a document with an external subset or external
+             * parameter entities with "standalone='no'", ...
+             * ... The declaration of a parameter entity must
+             * precede any reference to it...
+             */
+            xmlValidityError(ctxt, XML_ERR_UNDECLARED_ENTITY,
+                             "Entity '%s' not defined\n", name, NULL);
 	} else {
 	    xmlWarningMsg(ctxt, XML_WAR_UNDECLARED_ENTITY,
 		          "Entity '%s' not defined\n", name, NULL);
@ -7823,37 +7834,19 @@ xmlParsePEReference(xmlParserCtxtPtr ctxt)
 	(ctxt->sax->getParameterEntity != NULL))
 	entity = ctxt->sax->getParameterEntity(ctxt->userData, name);
    if (entity == NULL) {
-	/*
-	 * [ WFC: Entity Declared ]
-	 * In a document without any DTD, a document with only an
-	 * internal DTD subset which contains no parameter entity
-	 * references, or a document with "standalone='yes'", ...
-	 * ... The declaration of a parameter entity must precede
-	 * any reference to it...
-	 */
-	if ((ctxt->standalone == 1) ||
-	    ((ctxt->hasExternalSubset == 0) &&
-	     (ctxt->hasPErefs == 0))) {
+	if (((!ctxt->validate) && (ctxt->loadsubset)) ||
+            (ctxt->standalone == 1)) {
 	    xmlFatalErrMsgStr(ctxt, XML_ERR_UNDECLARED_ENTITY,
 			      "PEReference: %%%s; not found\n",
 			      name);
-	} else {
-	    /*
-	     * [ VC: Entity Declared ]
-	     * In a document with an external subset or external
-	     * parameter entities with "standalone='no'", ...
-	     * ... The declaration of a parameter entity must
-	     * precede any reference to it...
-	     */
-            if ((ctxt->validate) && (ctxt->vctxt.error != NULL)) {
-                xmlValidityError(ctxt, XML_ERR_UNDECLARED_ENTITY,
-                                 "PEReference: %%%s; not found\n",
-                                 name, NULL);
-            } else
-                xmlWarningMsg(ctxt, XML_WAR_UNDECLARED_ENTITY,
-                              "PEReference: %%%s; not found\n",
+	} else if (ctxt->validate) {
+            xmlValidityError(ctxt, XML_ERR_UNDECLARED_ENTITY,
+                             "PEReference: %%%s; not found\n",
+                             name, NULL);
+        } else {
+            xmlWarningMsg(ctxt, XML_WAR_UNDECLARED_ENTITY,
+                          "PEReference: %%%s; not found\n",
                              name, NULL);
-            ctxt->valid = 0;
 	}
    } else {
 	/*
@ -8098,30 +8091,18 @@ xmlParseStringPEReference(xmlParserCtxtPtr ctxt, const xmlChar **str) {
 	(ctxt->sax->getParameterEntity != NULL))
 	entity = ctxt->sax->getParameterEntity(ctxt->userData, name);
    if (entity == NULL) {
-	/*
-	 * [ WFC: Entity Declared ]
-	 * In a document without any DTD, a document with only an
-	 * internal DTD subset which contains no parameter entity
-	 * references, or a document with "standalone='yes'", ...
-	 * ... The declaration of a parameter entity must precede
-	 * any reference to it...
-	 */
-	if ((ctxt->standalone == 1) ||
-	    ((ctxt->hasExternalSubset == 0) && (ctxt->hasPErefs == 0))) {
+	if (((!ctxt->validate) && (ctxt->loadsubset)) ||
+            (ctxt->standalone == 1)) {
 	    xmlFatalErrMsgStr(ctxt, XML_ERR_UNDECLARED_ENTITY,
 		 "PEReference: %%%s; not found\n", name);
-	} else {
-	    /*
-	     * [ VC: Entity Declared ]
-	     * In a document with an external subset or external
-	     * parameter entities with "standalone='no'", ...
-	     * ... The declaration of a parameter entity must
-	     * precede any reference to it...
-	     */
+	} else if (ctxt->validate) {
+            xmlValidityError(ctxt, XML_ERR_UNDECLARED_ENTITY,
+                             "PEReference: %%%s; not found\n",
+                             name, NULL);
+        } else {
 	    xmlWarningMsg(ctxt, XML_WAR_UNDECLARED_ENTITY,
 			  "PEReference: %%%s; not found\n",
 			  name, NULL);
-	    ctxt->valid = 0;
 	}
    } else {
 	/*
--- a/python/tests/input_callback.py
+++ b/python/tests/input_callback.py
@ -101,7 +101,7 @@ run_test(desc="Loading entity with custom callback",
        exp_status="loaded", exp_err=[
            (  3, 'failed to load "http://example.com/dtds/sample.dtd": Attempt to load network entity\n'),
            ( -1, "Attempt to load network entity: http://example.com/dtds/sample.dtd"),
-            (  3, "Entity 'sample.entity' not defined\n")
+            (  4, "Entity 'sample.entity' not defined\n")
            ])

 # Register a catalog (also accessible via pystr://) and retry
@ -114,7 +114,7 @@ run_test(desc="Loading entity and unregistering callback",
        test_callback=lambda: libxml2.popInputCallbacks(),
        exp_status="loaded", exp_err=[
            ( 3, "failed to load \"py://strings/dtds/sample.dtd\": No such file or directory\n"),
-            ( 3, "Entity 'sample.entity' not defined\n")
+            ( 4, "Entity 'sample.entity' not defined\n")
            ])

 # Try to load the document again
--- a/result/ent7.sax
+++ b/result/ent7.sax
@ -1,4 +1,3 @@
-xmlParseDocument returned error 26
 SAX.setDocumentLocator()
 SAX.startDocument()
 SAX.internalSubset(item, , )
--- a/result/ent7.sax2
+++ b/result/ent7.sax2
@ -1,4 +1,3 @@
-xmlParseDocument returned error 26
 SAX.setDocumentLocator()
 SAX.startDocument()
 SAX.internalSubset(item, , )
--- a/result/noent/ent7.sax2
+++ b/result/noent/ent7.sax2
@ -1,4 +1,3 @@
-xmlParseDocument returned error 26
 SAX.setDocumentLocator()
 SAX.startDocument()
 SAX.internalSubset(item, , )
--- a/result/noent/xml2.sax2
+++ b/result/noent/xml2.sax2
@ -1,4 +1,3 @@
-xmlParseDocument returned error 26
 SAX.setDocumentLocator()
 SAX.startDocument()
 SAX.internalSubset(test, , )
--- a/result/xml2.sax
+++ b/result/xml2.sax
@ -1,4 +1,3 @@
-xmlParseDocument returned error 26
 SAX.setDocumentLocator()
 SAX.startDocument()
 SAX.internalSubset(test, , )
--- a/result/xml2.sax2
+++ b/result/xml2.sax2
@ -1,4 +1,3 @@
-xmlParseDocument returned error 26
 SAX.setDocumentLocator()
 SAX.startDocument()
 SAX.internalSubset(test, , )
--- a/runtest.c
+++ b/runtest.c
@ -11,6 +11,8 @@
 * daniel@veillard.com
 */

+#define XML_DEPRECATED
+
 #include "libxml.h"
 #include <stdio.h>
 #ifdef HAVE_UNISTD_H
@ -1601,9 +1603,11 @@ oldParseTest(const char *filename, const char *result,
     * base of the test, parse with the old API
     */
 #ifdef LIBXML_SAX1_ENABLED
+    xmlGetWarningsDefaultValue = 0;
    doc = xmlParseFile(filename);
+    xmlGetWarningsDefaultValue = 1;
 #else
-    doc = xmlReadFile(filename, NULL, 0);
+    doc = xmlReadFile(filename, NULL, XML_PARSE_NOWARNING);
 #endif
    if (doc == NULL)
        return(1);
@ -1622,9 +1626,11 @@ oldParseTest(const char *filename, const char *result,
     * Parse the saved result to make sure the round trip is okay
     */
 #ifdef LIBXML_SAX1_ENABLED
+    xmlGetWarningsDefaultValue = 0;
    doc = xmlParseFile(temp);
+    xmlGetWarningsDefaultValue = 1;
 #else
-    doc = xmlReadFile(temp, NULL, 0);
+    doc = xmlReadFile(temp, NULL, XML_PARSE_NOWARNING);
 #endif
    if (doc == NULL)
        return(1);