urlapi: add CURLU_NO_GUESS_SCHEME

Used for extracting:

- when used asking for a scheme, it will return CURLUE_NO_SCHEME if the
  stored information was a guess

- when used asking for a URL, the URL is returned without a scheme, like
  when previously given to the URL parser when it was asked to guess

- as soon as the scheme is set explicitly, it is no longer internally
  marked as guessed

The idea being:

1. allow a user to figure out if a URL's scheme was set as a result of
  guessing

2. extract the URL without a guessed scheme

3. this makes it work similar to how we already deal with port numbers

Extend test 1560 to verify.

Closes #13616

docs/libcurl/curl_url_get.md

@@ -125,6 +125,22 @@ nothing following the hash sign.
 
 (Added in curl 8.8.0)
 
+## CURLU_NO_GUESS_SCHEME
+
+When this flag is used in curl_url_get(), it treats the scheme as non-existing
+if it was set as a result of a previous guess; when CURLU_GUESS_SCHEME was
+used parsing a URL.
+
+Using this flag when getting CURLUPART_SCHEME if the scheme was set as the
+result of a guess makes curl_url_get() return CURLUE_NO_SCHEME.
+
+Using this flag when getting CURLUPART_URL if the scheme was set as the result
+of a guess makes curl_url_get() return the full URL without the scheme
+component. Such a URL can then only be parsed with curl_url_set() if
+CURLU_GUESS_SCHEME is used.
+
+(Added in curl 8.9.0)
+
 # PARTS
 
 ## CURLUPART_URL

docs/libcurl/symbols-in-versions

@@ -1068,6 +1068,7 @@ CURLU_GET_EMPTY                 8.8.0
 CURLU_GUESS_SCHEME              7.62.0
 CURLU_NO_AUTHORITY              7.67.0
 CURLU_NO_DEFAULT_PORT           7.62.0
+CURLU_NO_GUESS_SCHEME           8.9.0
 CURLU_NON_SUPPORT_SCHEME        7.62.0
 CURLU_PATH_AS_IS                7.62.0
 CURLU_PUNY2IDN                  8.3.0

include/curl/urlapi.h

@@ -102,6 +102,7 @@ typedef enum {
 #define CURLU_GET_EMPTY (1<<14)         /* allow empty queries and fragments
                                            when extracting the URL or the
                                            components */
+#define CURLU_NO_GUESS_SCHEME (1<<14)   /* for get, don't accept a guess */
 
 typedef struct Curl_URL CURLU;
 

lib/urlapi.c

@@ -82,6 +82,7 @@ struct Curl_URL {
   unsigned short portnum; /* the numerical version (if 'port' is set) */
   BIT(query_present);    /* to support blank */
   BIT(fragment_present); /* to support blank */
+  BIT(guessed_scheme);   /* when a URL without scheme is parsed */
 };
 
 #define DEFAULT_SCHEME "https"
@@ -1223,6 +1224,7 @@ static CURLUcode parseurl(const char *url, CURLU *u, unsigned int flags)
           result = CURLUE_OUT_OF_MEMORY;
           goto fail;
         }
+        u->guessed_scheme = TRUE;
       }
     }
     else if(flags & CURLU_NO_AUTHORITY) {
@@ -1437,6 +1439,8 @@ CURLUcode curl_url_get(const CURLU *u, CURLUPart what,
     ptr = u->scheme;
     ifmissing = CURLUE_NO_SCHEME;
     urldecode = FALSE; /* never for schemes */
+    if((flags & CURLU_NO_GUESS_SCHEME) && u->guessed_scheme)
+      return CURLUE_NO_SCHEME;
     break;
   case CURLUPART_USER:
     ptr = u->user;
@@ -1525,6 +1529,7 @@ CURLUcode curl_url_get(const CURLU *u, CURLUPart what,
       return CURLUE_NO_HOST;
     else {
       const struct Curl_handler *h = NULL;
+      char schemebuf[MAX_SCHEME_LEN + 5];
       if(u->scheme)
         scheme = u->scheme;
       else if(flags & CURLU_DEFAULT_SCHEME)
@@ -1595,8 +1600,13 @@ CURLUcode curl_url_get(const CURLU *u, CURLUPart what,
         }
       }
 
-      url = aprintf("%s://%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
-                    scheme,
+      if(!(flags & CURLU_NO_GUESS_SCHEME) || !u->guessed_scheme)
+        msnprintf(schemebuf, sizeof(schemebuf), "%s://", scheme);
+      else
+        schemebuf[0] = 0;
+
+      url = aprintf("%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
+                    schemebuf,
                     u->user ? u->user : "",
                     u->password ? ":": "",
                     u->password ? u->password : "",
@@ -1718,6 +1728,7 @@ CURLUcode curl_url_set(CURLU *u, CURLUPart what,
       break;
     case CURLUPART_SCHEME:
       storep = &u->scheme;
+      u->guessed_scheme = FALSE;
       break;
     case CURLUPART_USER:
       storep = &u->user;
@@ -1790,6 +1801,7 @@ CURLUcode curl_url_set(CURLU *u, CURLUPart what,
     }
     else
       return CURLUE_BAD_SCHEME;
+    u->guessed_scheme = FALSE;
     break;
   }
   case CURLUPART_USER:

tests/libtest/lib1560.c

@@ -151,6 +151,9 @@ struct clearurlcase {
 };
 
 static const struct testcase get_parts_list[] ={
+  {"curl.se",
+   "[10] | [11] | [12] | [13] | curl.se | [15] | / | [16] | [17]",
+   CURLU_GUESS_SCHEME, CURLU_NO_GUESS_SCHEME, CURLUE_OK},
   {"https://curl.se:0/#",
    "https | [11] | [12] | [13] | curl.se | 0 | / | [16] | ",
    0, CURLU_GET_EMPTY, CURLUE_OK},
@@ -526,6 +529,9 @@ static const struct testcase get_parts_list[] ={
 };
 
 static const struct urltestcase get_url_list[] = {
+  {"example.com",
+   "example.com/",
+   CURLU_GUESS_SCHEME, CURLU_NO_GUESS_SCHEME, CURLUE_OK},
   {"http://user@example.com?#",
    "http://user@example.com/?#",
    0, CURLU_GET_EMPTY, CURLUE_OK},