epub: Decode URL escapes in epub paths.

4 files changed, 47 insertions, 1 deletions

diff --git a/include/mupdf/fitz/string.h b/include/mupdf/fitz/string.h
index 4de882a6..010dc46a 100644
--- a/include/mupdf/fitz/string.h
+++ b/include/mupdf/fitz/string.h
@@ -58,6 +58,11 @@ int fz_strlcat(char *dst, const char *src, int n);
 void fz_dirname(char *dir, const char *path, int dirsize);
 
 /*
+	fz_urldecode: decode url escapes.
+*/
+char *fz_urldecode(char *url);
+
+/*
 	fz_cleanname: rewrite path to the shortest string that names the same path.
 
 	Eliminates multiple and trailing slashes, interprets "." and "..".
diff --git a/source/fitz/string.c b/source/fitz/string.c
index 224eeff7..820a4426 100644
--- a/source/fitz/string.c
+++ b/source/fitz/string.c
@@ -84,6 +84,44 @@ fz_dirname(char *dir, const char *path, int n)
 	dir[i+1] = 0;
 }
 
+static int ishex(int a)
+{
+	return (a >= 'A' && a <= 'F') ||
+		(a >= 'a' && a <= 'f') ||
+		(a >= '0' && a <= '9');
+}
+
+static int tohex(int c)
+{
+	if (c >= '0' && c <= '9') return c - '0';
+	if (c >= 'a' && c <= 'f') return c - 'a' + 0xA;
+	if (c >= 'A' && c <= 'F') return c - 'A' + 0xA;
+	return 0;
+}
+
+char *
+fz_urldecode(char *url)
+{
+	char *s = url;
+	char *p = url;
+	while (*s)
+	{
+		int c = (unsigned char) *s++;
+		if (c == '%' && ishex(s[0]) && ishex(s[1]))
+		{
+			int a = tohex(*s++);
+			int b = tohex(*s++);
+			*p++ = a << 4 | b;
+		}
+		else
+		{
+			*p++ = c;
+		}
+	}
+	*p = 0;
+	return url;
+}
+
 #define SEP(x) ((x)=='/' || (x) == 0)
 
 char *
@@ -139,6 +177,7 @@ fz_cleanname(char *name)
 	*q = '\0';
 	return name;
 }
+
 enum
 {
 	UTFmax = 4, /* maximum bytes per rune */
diff --git a/source/html/epub-doc.c b/source/html/epub-doc.c
index 2ea9f298..e77770ac 100644
--- a/source/html/epub-doc.c
+++ b/source/html/epub-doc.c
@@ -164,7 +164,7 @@ path_from_idref(char *path, fz_xml *manifest, const char *base_uri, const char *
 	fz_strlcpy(path, base_uri, n);
 	fz_strlcat(path, "/", n);
 	fz_strlcat(path, rel_path, n);
-	return fz_cleanname(path);
+	return fz_cleanname(fz_urldecode(path));
 }
 
 static epub_chapter *
diff --git a/source/html/html-layout.c b/source/html/html-layout.c
index a64fccc4..4926dc51 100644
--- a/source/html/html-layout.c
+++ b/source/html/html-layout.c
@@ -129,6 +129,7 @@ static void generate_image(fz_context *ctx, fz_archive *zip, const char *base_ur
 	fz_strlcpy(path, base_uri, sizeof path);
 	fz_strlcat(path, "/", sizeof path);
 	fz_strlcat(path, src, sizeof path);
+	fz_urldecode(path);
 	fz_cleanname(path);
 
 	fz_try(ctx)
@@ -1024,6 +1025,7 @@ html_load_css(fz_context *ctx, fz_archive *zip, const char *base_uri, fz_css_rul
 						fz_strlcpy(path, base_uri, sizeof path);
 						fz_strlcat(path, "/", sizeof path);
 						fz_strlcat(path, href, sizeof path);
+						fz_urldecode(path);
 						fz_cleanname(path);
 
 						buf = fz_read_archive_entry(ctx, zip, path);