tscrape

commit 5df58d27f557292778cdc5dee306f18db8c980f7
parent f8629e681a16fc3af086355a44c942df57291b4b
Author: Hiltjo Posthuma <hiltjo@codemadness.org>
Date:   Sat,  1 Feb 2020 15:02:27 +0100

sync XML improvements

Diffstat:
M
tscrape.c
 | 
8
++++----
M
xml.c
 | 
24
++++++++----------------
M
xml.h
 | 
2
++

3 files changed, 14 insertions(+), 20 deletions(-)
diff --git a/tscrape.c b/tscrape.c
@@ -107,10 +107,10 @@ isclassmatch(const char *classes, const char *clss, size_t len)
 }
 
 /* convert XML and some HTML entities */
-static ssize_t
+static int
 html_entitytostr(const char *s, char *buf, size_t bufsiz)
 {
-	ssize_t len;
+	int len;
 
 	if ((len = xml_entitytostr(s, buf, bufsiz)) > 0)
 		return len;
@@ -244,7 +244,7 @@ xmlattrentity(XMLParser *x, const char *t, size_t tl, const char *a, size_t al,
               const char *v, size_t vl)
 {
 	char buf[16];
-	ssize_t len;
+	int len;
 
 	if (!state)
 		return;
@@ -267,7 +267,7 @@ static void
 xmldataentity(XMLParser *x, const char *d, size_t dl)
 {
 	char buf[16];
-	ssize_t len;
+	int len;
 
 	if (!(state & Text))
 		return;
diff --git a/xml.c b/xml.c
@@ -1,8 +1,5 @@
-#include <sys/types.h>
-
 #include <ctype.h>
 #include <errno.h>
-#include <limits.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
@@ -255,11 +252,6 @@ namedentitytostr(const char *e, char *buf, size_t bufsiz)
 		{ "gt;",   '>'  },
 		{ "apos;", '\'' },
 		{ "quot;", '"'  },
-		{ "AMP;",  '&'  },
-		{ "LT;",   '<'  },
-		{ "GT;",   '>'  },
-		{ "APOS;", '\'' },
-		{ "QUOT;", '"'  }
 	};
 	size_t i;
 
@@ -274,7 +266,7 @@ namedentitytostr(const char *e, char *buf, size_t bufsiz)
 			return 1;
 		}
 	}
-	return 0;
+	return -1;
 }
 
 static int
@@ -291,12 +283,12 @@ numericentitytostr(const char *e, char *buf, size_t bufsiz)
 	errno = 0;
 	/* hex (16) or decimal (10) */
 	if (*e == 'x')
-		l = strtoul(e + 1, &end, 16);
+		l = strtol(++e, &end, 16);
 	else
-		l = strtoul(e, &end, 10);
-	/* invalid value or not a well-formed entity or too high codepoint */
-	if (errno || *end != ';' || l > 0x10FFFF)
-		return 0;
+		l = strtol(e, &end, 10);
+	/* invalid value or not a well-formed entity or invalid codepoint */
+	if (errno || e == end || *end != ';' || l < 0 || l > 0x10ffff)
+		return -1;
 	len = codepointtoutf8(l, buf);
 	buf[len] = '\0';
 
@@ -304,13 +296,13 @@ numericentitytostr(const char *e, char *buf, size_t bufsiz)
 }
 
 /* convert named- or numeric entity string to buffer string
- * returns byte-length of string. */
+ * returns byte-length of string or -1 on failure. */
 int
 xml_entitytostr(const char *e, char *buf, size_t bufsiz)
 {
 	/* doesn't start with & */
 	if (e[0] != '&')
-		return 0;
+		return -1;
 	/* numeric entity */
 	if (e[1] == '#')
 		return numericentitytostr(e + 2, buf, bufsiz);
diff --git a/xml.h b/xml.h
@@ -1,6 +1,8 @@
 #ifndef _XML_H
 #define _XML_H
 
+#include <stdio.h>
+
 typedef struct xmlparser {
 	/* handlers */
 	void (*xmlattr)(struct xmlparser *, const char *, size_t,