From c1253d22193eae7937c7f0a223485be7bc9aa5b3 Mon Sep 17 00:00:00 2001
From: Michael Brown <mcb30@etherboot.org>
Date: Thu, 28 Apr 2005 13:20:02 +0000
Subject: [PATCH] Much, much smaller.  :)

---
 src/core/url.c | 183 +++++++++++++++----------------------------------
 1 file changed, 55 insertions(+), 128 deletions(-)

diff --git a/src/core/url.c b/src/core/url.c
index dd5d1a94..2a156e8b 100644
--- a/src/core/url.c
+++ b/src/core/url.c
@@ -1,144 +1,71 @@
-#include "stdlib.h"
 #include "string.h"
-#include "proto.h"
-#include "resolv.h"
 #include "url.h"
 
-static struct protocol protocols[0] __protocol_start;
-static struct protocol default_protocols[0] __default_protocol_start;
-static struct protocol protocols_end[0] __protocol_end;
-
 /*
- * Parse protocol portion of a URL.  Return 0 if no "proto://" is
- * present.
- *
- */
-static inline int parse_protocol ( struct url_info *info, const char **p ) {
-	const char *q = *p;
-
-	info->protocol = q;
-	for ( ; *q ; q++ ) {
-		if ( memcmp ( q, "://", 3 ) == 0 ) {
-			info->protocol_len = q - info->protocol;
-			*p = q + 3;
-			return 1;
-		}
-	}
-	return 0;
-}
-
-/*
- * Parse the host:port portion of a URL.  Also fills in sin_port.
- *
- */
-static inline void parse_host_port ( struct url_info *info, const char **p ) {
-	info->host = *p;
-	for ( ; **p && ( **p != '/' ) ; (*p)++ ) {
-		if ( **p == ':' ) {
-			info->host_len = *p - info->host;
-			info->port = ++(*p);
-			info->sin.sin_port = strtoul ( *p, p, 10 );
-			info->port_len = *p - info->port;
-			return;
-		}
-	}
-	/* No ':' separator seen; it's all the host part */
-	info->host_len = *p - info->host;
-}
-
-/*
- * Identify the protocol
- *
- */
-static inline int identify_protocol ( struct url_info *info ) {
-	struct protocol *proto;
-
-	if ( info->protocol_len ) {
-		char *terminator;
-		char temp;
-
-		/* Explcitly specified protocol */
-		terminator = ( char * ) &info->protocol[info->protocol_len];
-		temp = *terminator;
-		*terminator = '\0';
-		for ( proto = protocols ; proto < protocols_end ; proto++ ) {
-			if ( memcmp ( proto->name, info->protocol,
-				      info->protocol_len + 1 ) == 0 ) {
-				info->proto = proto;
-				break;
-			}
-		}
-		*terminator = temp;
-	} else {
-		/* No explicitly specified protocol */
-		if ( default_protocols < protocols_end )
-			info->proto = default_protocols;
-	}
-	return ( ( int ) info->proto ); /* NULL indicates failure */
-}
-
-/*
- * Resolve the host portion of the URL
- *
- */
-static inline int resolve_host ( struct url_info *info ) {
-	char *terminator;
-	char temp;
-	int success;
-
-	if ( ! info->host_len ) {
-		/* No host specified - leave sin.sin_addr empty to
-		 * indicate use of DHCP-supplied next-server
-		 */
-		return 1;
-	}
-
-	terminator = ( char * ) &info->host[info->host_len];
-	temp = *terminator;
-	*terminator = '\0';
-	success = resolv ( &info->sin.sin_addr, info->host );
-	*terminator = temp;
-	return success;
-}
-
-/*
- * Parse a URL string into its constituent parts.  Perform name
- * resolution if required (and if resolver code is linked in), and
- * identify the protocol.
+ * Parse a URL string into its constituent parts.
  *
  * We accept URLs of the form
  *
  *   [protocol://[host][:port]/]path/to/file
  *
  * We return true for success, 0 for failure (e.g. unknown protocol).
- * Note that the "/" before path/to/file *will* be counted as part of
- * the filename, if it is present.
+ * The URL string will be modified by having NULs inserted after
+ * "protocol", "host" and "port".  The original URL can be
+ * reconstructed by calling unparse_url.
  *
  */
-int parse_url ( struct url_info *info, const char *url ) {
-	const char *p;
+void parse_url ( struct url_info *info, char *url ) {
+	char *p;
 
-	/* Fill in initial values */
+	/* Zero the structure */
 	memset ( info, 0, sizeof ( *info ) );
-	info->url = url;
-	info->protocol = url;
-	info->host = url;
-	info->port = url;
+
+	/* Search for a protocol delimiter */
+	for ( p = url ; *p ; p++ ) {
+		if ( memcmp ( p, "://", 3 ) != 0 )
+			continue;
+
+		/* URL has an explicit protocol */
+		info->protocol = url;
+		*p = '\0';
+		p += 3;
+		info->host = p;
+
+		/* Search for port or file delimiter */
+		for ( ; *p ; p++ ) {
+			if ( *p == ':' ) {
+				*p = '\0';
+				info->port = p + 1;
+				continue;
+			}
+			if ( *p == '/' ) {
+				*(p++) = '\0';
+				break;
+			}
+		}
+		info->file = p;
+		return;
+	}
+
+	/* URL has no explicit protocol; is just a filename */
 	info->file = url;
-
-	/* Split the URL into substrings, and fill in sin.sin_port */
-	p = url;
-	if ( parse_protocol ( info, &p ) )
-		parse_host_port ( info, &p );
-	info->file = p;
-
-	/* Identify the protocol */
-	if ( ! identify_protocol ( info ) )
-		return 0;
-
-	/* Resolve the host name to an IP address */
-	if ( ! resolve_host ( info ) )
-		return 0;
-
-	return 1;
+}
+
+/*
+ * Restore a parsed URL to its original pristine form.
+ *
+ */
+char * unparse_url ( struct url_info *info ) {
+	if ( info->protocol ) {
+		/* URL had a protocol: fill in the deleted separators */
+		info->file[-1] = '/';
+		if ( info->port ) {
+			info->port[-1] = ':';
+		}
+		info->host[-3] = ':';
+		return info->protocol;
+	} else {
+		/* URL had no protocol; was just a filename */
+		return info->file;
+	}
 }