tech@mandoc.bsd.lv
 help / color / mirror / Atom feed
* man.cgi: don't add query strings to search results
@ 2015-12-24 14:16 Anthony J. Bentley
  2015-12-25 14:58 ` Ingo Schwarze
  0 siblings, 1 reply; 2+ messages in thread
From: Anthony J. Bentley @ 2015-12-24 14:16 UTC (permalink / raw)
  To: tech

Hi,

After performing a search, man.cgi adds the query string to the
resulting URL (if there's a single result) or to each link in the
search results (if more than one).

I believe that here the query string is just clutter. If I want the
query string after clicking a link in the multiple-results case, I
can just hit the back button to return to the results page, which
still contains the query in its own URL and in the search box. If
I want the query string after getting an unambiguous result... well,
I can't imagine wanting that. But even so, browsers like Firefox
will still retain the query in that case too if I hit back. No benefit
to keeping it on the destination page.

And I really think the benefit of clean, simple URLs by default for
the most common case--linking to individual manuals--can't be
overstated.

Index: cgi.c
===================================================================
RCS file: /cvs/mdocml/cgi.c,v
retrieving revision 1.115
diff -u -p -r1.115 cgi.c
--- cgi.c	7 Nov 2015 17:58:55 -0000	1.115
+++ cgi.c	24 Dec 2015 13:49:40 -0000
@@ -63,9 +63,6 @@ static	void		 html_print(const char *);
 static	void		 html_putchar(char);
 static	int		 http_decode(char *);
 static	void		 http_parse(struct req *, const char *);
-static	void		 http_print(const char *);
-static	void		 http_putchar(char);
-static	void		 http_printquery(const struct req *, const char *);
 static	void		 pathgen(struct req *);
 static	void		 pg_error_badrequest(const char *);
 static	void		 pg_error_internal(void);
@@ -149,40 +146,6 @@ html_putchar(char c)
 	}
 }
 
-static void
-http_printquery(const struct req *req, const char *sep)
-{
-
-	if (NULL != req->q.query) {
-		printf("query=");
-		http_print(req->q.query);
-	}
-	if (0 == req->q.equal)
-		printf("%sapropos=1", sep);
-	if (NULL != req->q.sec) {
-		printf("%ssec=", sep);
-		http_print(req->q.sec);
-	}
-	if (NULL != req->q.arch) {
-		printf("%sarch=", sep);
-		http_print(req->q.arch);
-	}
-	if (strcmp(req->q.manpath, req->p[0])) {
-		printf("%smanpath=", sep);
-		http_print(req->q.manpath);
-	}
-}
-
-static void
-http_print(const char *p)
-{
-
-	if (NULL == p)
-		return;
-	while ('\0' != *p)
-		http_putchar(*p++);
-}
-
 /*
  * Call through to html_putchar().
  * Accepts NULL strings.
@@ -303,20 +266,6 @@ next:
 	}
 }
 
-static void
-http_putchar(char c)
-{
-
-	if (isalnum((unsigned char)c)) {
-		putchar((unsigned char)c);
-		return;
-	} else if (' ' == c) {
-		putchar('+');
-		return;
-	}
-	printf("%%%.2x", c);
-}
-
 /*
  * HTTP-decode a string.  The standard explanation is that this turns
  * "%4e+foo" into "n foo" in the regular way.  This is done in-place
@@ -623,9 +572,8 @@ pg_searchres(const struct req *req, stru
 		 * without any delay.
 		 */
 		printf("Status: 303 See Other\r\n");
-		printf("Location: http://%s%s/%s/%s?",
+		printf("Location: http://%s%s/%s/%s",
 		    HTTP_HOST, scriptname, req->q.manpath, r[0].file);
-		http_printquery(req, "&");
 		printf("\r\n"
 		     "Content-Type: text/html; charset=utf-8\r\n"
 		     "\r\n");
@@ -640,9 +588,8 @@ pg_searchres(const struct req *req, stru
 	for (i = 0; i < sz; i++) {
 		printf("<TR>\n"
 		       "<TD CLASS=\"title\">\n"
-		       "<A HREF=\"%s/%s/%s?",
+		       "<A HREF=\"%s/%s/%s",
 		    scriptname, req->q.manpath, r[i].file);
-		http_printquery(req, "&amp;");
 		printf("\">");
 		html_print(r[i].names);
 		printf("</A>\n"
--
 To unsubscribe send an email to tech+unsubscribe@mdocml.bsd.lv

^ permalink raw reply	[flat|nested] 2+ messages in thread
end of thread, other threads:[~2015-12-25 14:58 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-12-24 14:16 man.cgi: don't add query strings to search results Anthony J. Bentley
2015-12-25 14:58 ` Ingo Schwarze

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).