sfeed_web: attribute parsing improvements, improve man page

Fix attribute parsing and now decode entities. The following now works (from
helsinkitimes.fi):

	<base href="https://www.helsinkitimes.fi/" />
	<link href="/?format=feed&amp;type=rss" rel="alternate" type="application/rss+xml" title="RSS 2.0" />
	<link href="/?format=feed&amp;type=atom" rel="alternate" type="application/atom+xml" title="Atom 1.0" />

Properly associate attributes with the actual tag, this now parses properly
(from ascii.jp).

	<link rel="apple-touch-icon-precomposed" href="/img/apple-touch-icon.png" />
	<link rel="alternate" type="application/rss+xml"  />

2 files changed, 72 insertions, 35 deletions

diff --git a/sfeed_web.1 b/sfeed_web.1
index 1f7d0cc..5b93d9b 100644
--- a/sfeed_web.1
+++ b/sfeed_web.1
@@ -1,4 +1,4 @@
-.Dd March 15, 2020
+.Dd October 22, 2020
 .Dt SFEED_WEB 1
 .Os
 .Sh NAME
@@ -21,13 +21,12 @@ url<TAB>content\-type<newline>
 .Bl -tag -width Ds
 .It url
 Found relative or absolute url.
-If the url is relative and the
+.Pp
+For relative urls if a <base href="..." /> tag is found it will be used,
+otherwise if the
 .Ar baseurl
-option is
-specified then the url is made absolute.
-If the url is relative and no
-.Ar baseurl
-option is specified then it is printed as is.
+option is specified then that is used, if neither are set then the relative url
+is printed.
 .It content\-type
 Usually application/atom+xml or application/rss+xml.
 .El
diff --git a/sfeed_web.c b/sfeed_web.c
index 308d22c..c9cea05 100644
--- a/sfeed_web.c
+++ b/sfeed_web.c
@@ -10,65 +10,101 @@
 #define STRP(s) s,sizeof(s)-1
 
 static XMLParser parser;
-static int isbase, islink, isfeedlink;
-static char abslink[4096], feedlink[4096], basehref[4096], feedtype[256];
+static int isbasetag, islinktag, ishrefattr, istypeattr;
+static char linkhref[4096], linktype[256], basehref[4096];
+static char abslink[4096];
 
 static void
-printfeedtype(const char *s, FILE *fp)
+printvalue(const char *s)
 {
 	for (; *s; s++)
-		if (!isspace((unsigned char)*s))
-			fputc(*s, fp);
+		if (!iscntrl((unsigned char)*s))
+			putchar(*s);
 }
 
 static void
 xmltagstart(XMLParser *p, const char *t, size_t tl)
 {
-	isbase = islink = isfeedlink = 0;
-	feedlink[0] = '\0';
+	isbasetag = islinktag = 0;
 
 	if (!strcasecmp(t, "base"))
-		isbase = 1;
+		isbasetag = 1;
 	else if (!strcasecmp(t, "link"))
-		islink = 1;
+		islinktag = 1;
 }
 
 static void
 xmltagstartparsed(XMLParser *p, const char *t, size_t tl, int isshort)
 {
-	if (!isfeedlink)
+	if (!islinktag)
 		return;
 
-	if (absuri(abslink, sizeof(abslink), feedlink, basehref) != -1)
-		fputs(abslink, stdout);
+	if (strncasecmp(linktype, STRP("application/atom")) &&
+	    strncasecmp(linktype, STRP("application/xml")) &&
+	    strncasecmp(linktype, STRP("application/rss")))
+		return;
+
+	if (absuri(abslink, sizeof(abslink), linkhref, basehref) != -1)
+		printvalue(abslink);
 	else
-		fputs(feedlink, stdout);
+		printvalue(linkhref);
 	putchar('\t');
-	printfeedtype(feedtype, stdout);
+	printvalue(linktype);
 	putchar('\n');
 }
 
 static void
+xmlattrstart(XMLParser *p, const char *t, size_t tl, const char *a, size_t al)
+{
+	ishrefattr = istypeattr = 0;
+
+	if (!isbasetag && !islinktag)
+		return;
+
+	if (!strcasecmp(a, "href")) {
+		ishrefattr = 1;
+		if (isbasetag)
+			basehref[0] = '\0';
+		else if (islinktag)
+			linkhref[0] = '\0';
+	} else if (!strcasecmp(a, "type") && islinktag) {
+		istypeattr = 1;
+		linktype[0] = '\0';
+	}
+}
+
+static void
 xmlattr(XMLParser *p, const char *t, size_t tl, const char *n, size_t nl,
 	const char *v, size_t vl)
 {
-	if (isbase) {
-		if (!strcasecmp(n, "href"))
-			strlcpy(basehref, v, sizeof(basehref));
-	} else if (islink) {
-		if (!strcasecmp(n, "type")) {
-			if (!strncasecmp(v, STRP("application/atom")) ||
-			    !strncasecmp(v, STRP("application/xml"))  ||
-			    !strncasecmp(v, STRP("application/rss"))) {
-				isfeedlink = 1;
-				strlcpy(feedtype, v, sizeof(feedtype));
-			}
-		} else if (!strcasecmp(n, "href")) {
-			strlcpy(feedlink, v, sizeof(feedlink));
-		}
+	if (isbasetag && ishrefattr) {
+		strlcat(basehref, v, sizeof(basehref));
+	} else if (islinktag) {
+		if (ishrefattr)
+			strlcat(linkhref, v, sizeof(linkhref));
+		else if (istypeattr)
+			strlcat(linktype, v, sizeof(linktype));
 	}
 }
 
+static void
+xmlattrentity(XMLParser *p, const char *t, size_t tl, const char *a, size_t al,
+              const char *v, size_t vl)
+{
+	char buf[16];
+	int len;
+
+	if (!ishrefattr && !istypeattr)
+		return;
+
+	/* try to translate entity, else just pass as data to
+	 * xmlattr handler. */
+	if ((len = xml_entitytostr(v, buf, sizeof(buf))) > 0)
+		xmlattr(p, t, tl, a, al, buf, (size_t)len);
+	else
+		xmlattr(p, t, tl, a, al, v, vl);
+}
+
 int
 main(int argc, char *argv[])
 {
@@ -79,6 +115,8 @@ main(int argc, char *argv[])
 		strlcpy(basehref, argv[1], sizeof(basehref));
 
 	parser.xmlattr = xmlattr;
+	parser.xmlattrentity = xmlattrentity;
+	parser.xmlattrstart = xmlattrstart;
 	parser.xmltagstart = xmltagstart;
 	parser.xmltagstartparsed = xmltagstartparsed;