html: Stop implying <p> start tags
Only <html>, <head> or <body> should be implied. Opening extra <p> tags
has always been a libxml2 quirk.
diff --git a/HTMLparser.c b/HTMLparser.c
index eac9964..1af775e 100644
--- a/HTMLparser.c
+++ b/HTMLparser.c
@@ -1165,19 +1165,6 @@
};
/*
- * The list of HTML elements which are supposed not to have
- * CDATA content and where a p element will be implied
- *
- * TODO: extend that list by reading the HTML SGML DTD on
- * implied paragraph
- */
-static const char *const htmlNoContentElements[] = {
- "html",
- "head",
- NULL
-};
-
-/*
* The list of HTML attributes which are of content %Script;
* NOTE: when adding ones, check htmlIsScriptAttribute() since
* it assumes the name starts with 'on'
@@ -1568,48 +1555,22 @@
}
/**
- * htmlCheckParagraph
+ * htmlStartCharData
* @ctxt: an HTML parser context
*
- * Check whether a p element need to be implied before inserting
- * characters in the current element.
- *
- * Returns 1 if a paragraph has been inserted, 0 if not and -1
- * in case of error.
+ * Prepare for non-whitespace character data.
*/
-static int
-htmlCheckParagraph(htmlParserCtxtPtr ctxt) {
- const xmlChar *tag;
- int i;
-
- if (ctxt == NULL)
- return(-1);
- if (ctxt->options & HTML_PARSE_HTML5)
- return(0);
-
- tag = ctxt->name;
- if (tag == NULL) {
- htmlAutoClose(ctxt, BAD_CAST"p");
- htmlCheckImplied(ctxt, BAD_CAST"p");
- htmlnamePush(ctxt, BAD_CAST"p");
- if ((ctxt->sax != NULL) && (ctxt->sax->startElement != NULL))
- ctxt->sax->startElement(ctxt->userData, BAD_CAST"p", NULL);
- return(1);
- }
+static void
+htmlStartCharData(htmlParserCtxtPtr ctxt) {
+ if (ctxt->options & (HTML_PARSE_NOIMPLIED | HTML_PARSE_HTML5))
+ return;
if (!htmlOmittedDefaultValue)
- return(0);
- for (i = 0; htmlNoContentElements[i] != NULL; i++) {
- if (xmlStrEqual(tag, BAD_CAST htmlNoContentElements[i])) {
- htmlAutoClose(ctxt, BAD_CAST"p");
- htmlCheckImplied(ctxt, BAD_CAST"p");
- htmlnamePush(ctxt, BAD_CAST"p");
- if ((ctxt->sax != NULL) && (ctxt->sax->startElement != NULL))
- ctxt->sax->startElement(ctxt->userData, BAD_CAST"p", NULL);
- return(1);
- }
- }
- return(0);
+ return;
+
+ if (xmlStrEqual(ctxt->name, BAD_CAST "head"))
+ htmlAutoClose(ctxt, BAD_CAST "p");
+ htmlCheckImplied(ctxt, BAD_CAST "p");
}
/**
@@ -2972,7 +2933,7 @@
/*
* Add leading whitespace to html or head elements before
- * calling htmlCheckParagraph.
+ * calling htmlStartCharData.
*/
for (i = 0; i < size; i++)
if (!IS_WS_HTML(buf[i]))
@@ -2994,7 +2955,7 @@
if (size <= 0)
return;
- htmlCheckParagraph(ctxt);
+ htmlStartCharData(ctxt);
}
if ((mode == 0) &&
@@ -4084,7 +4045,7 @@
SKIP(2);
if (ctxt->input->cur >= ctxt->input->end) {
- htmlCheckParagraph(ctxt);
+ htmlStartCharData(ctxt);
if ((ctxt->sax != NULL) && (!ctxt->disableSAX) &&
(ctxt->sax->characters != NULL))
ctxt->sax->characters(ctxt->userData,
@@ -4243,7 +4204,7 @@
} else if (IS_ASCII_LETTER(NXT(1))) {
htmlParseElementInternal(ctxt);
} else {
- htmlCheckParagraph(ctxt);
+ htmlStartCharData(ctxt);
if ((ctxt->sax != NULL) && (!ctxt->disableSAX) &&
(ctxt->sax->characters != NULL))
ctxt->sax->characters(ctxt->userData, BAD_CAST "<", 1);
@@ -5187,7 +5148,7 @@
ctxt->checkIndex = 0;
} else {
ctxt->instate = XML_PARSER_CONTENT;
- htmlCheckParagraph(ctxt);
+ htmlStartCharData(ctxt);
if ((ctxt->sax != NULL) && (!ctxt->disableSAX) &&
(ctxt->sax->characters != NULL))
ctxt->sax->characters(ctxt->userData,
diff --git a/result/HTML/758518-entity.html b/result/HTML/758518-entity.html
index 0f66150..7dde7c6 100644
--- a/result/HTML/758518-entity.html
+++ b/result/HTML/758518-entity.html
@@ -1,2 +1,2 @@
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">
-<html><body><p>&jÙ</p></body></html>
+<html><body>&jÙ</body></html>
diff --git a/result/HTML/758518-entity.html.sax b/result/HTML/758518-entity.html.sax
index 25aa72a..4c512a0 100644
--- a/result/HTML/758518-entity.html.sax
+++ b/result/HTML/758518-entity.html.sax
@@ -2,10 +2,8 @@
SAX.startDocument()
SAX.startElement(html)
SAX.startElement(body)
-SAX.startElement(p)
SAX.characters(&j, 2)
SAX.characters(Ù, 2)
-SAX.endElement(p)
SAX.endElement(body)
SAX.endElement(html)
SAX.endDocument()
diff --git a/result/HTML/758605.html b/result/HTML/758605.html
index ce5cbe5..77f70a0 100644
--- a/result/HTML/758605.html
+++ b/result/HTML/758605.html
@@ -1,3 +1,3 @@
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">
-<html><body><p>&:ê
-</p></body></html>
+<html><body>&:ê
+</body></html>
diff --git a/result/HTML/758605.html.sax b/result/HTML/758605.html.sax
index c6dc85a..ba116b7 100644
--- a/result/HTML/758605.html.sax
+++ b/result/HTML/758605.html.sax
@@ -2,11 +2,9 @@
SAX.startDocument()
SAX.startElement(html)
SAX.startElement(body)
-SAX.startElement(p)
SAX.characters(&:, 2)
SAX.characters(ê
, 3)
-SAX.endElement(p)
SAX.endElement(body)
SAX.endElement(html)
SAX.endDocument()
diff --git a/result/HTML/implied1.html b/result/HTML/implied1.html
index 6c86065..a937900 100644
--- a/result/HTML/implied1.html
+++ b/result/HTML/implied1.html
@@ -1,6 +1,6 @@
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">
<html>
<head> </head>
-<body><p>x
-</p></body>
+<body>x
+</body>
</html>
diff --git a/result/HTML/implied1.html.sax b/result/HTML/implied1.html.sax
index b3d7165..531bdf6 100644
--- a/result/HTML/implied1.html.sax
+++ b/result/HTML/implied1.html.sax
@@ -5,10 +5,8 @@
SAX.characters( , 3)
SAX.endElement(head)
SAX.startElement(body)
-SAX.startElement(p)
SAX.characters(x
, 2)
-SAX.endElement(p)
SAX.endElement(body)
SAX.endElement(html)
SAX.endDocument()
diff --git a/testparser.c b/testparser.c
index 1d0f9e7..4004ac8 100644
--- a/testparser.c
+++ b/testparser.c
@@ -474,7 +474,7 @@
err = 1;
}
- node = xmlDocGetRootElement(doc)->children->children->children;
+ node = xmlDocGetRootElement(doc)->children->children;
if (!xmlStrEqual(node->content, BAD_CAST "-\xC3\xA4-")) {
fprintf(stderr, "testHtmlPushWithEncoding failed\n");
err = 1;
