aws-sigv4: url encode the canonical path

Refactors canon_query, so it could use the encoding part of the function
to use it in the path.

As the path doesn't encode '/', but encode '=', I had to add some
conditions to know If I was doing the query or path encoding.

Also, instead of adding a `bool in_path` variable, I use `bool
*found_equals` to know if the function was called for the query or path,
as found_equals is used only in query_encoding.

Test 472 verifies.

Reported-by: Alexander Shtuchkin
Fixes #13754
Closes #13814

Signed-off-by: Matthias Gatto <matthias.gatto@outscale.com>

diff --git a/lib/http_aws_sigv4.c b/lib/http_aws_sigv4.c
index 27c95d47fe8bd57247a502bd973ba90fed415e02..4cc397dfeda671652949e614e5a81f4623485aaa 100644 (file)
--- a/lib/http_aws_sigv4.c
+++ b/lib/http_aws_sigv4.c
@@ -423,6 +423,76 @@ static int compare_func(const void *a, const void *b)
 
 #define MAX_QUERYPAIRS 64
 
+/**
+ * found_equals have a double meaning,
+ * detect if an equal have been found when called from canon_query,
+ * and mark that this function is called to compute the path,
+ * if found_equals is NULL.
+ */
+static CURLcode canon_string(const char *q, size_t len,
+                             struct dynbuf *dq, bool *found_equals)
+{
+  CURLcode result = CURLE_OK;
+
+  for(; len && !result; q++, len--) {
+    if(ISALNUM(*q))
+      result = Curl_dyn_addn(dq, q, 1);
+    else {
+      switch(*q) {
+      case '-':
+      case '.':
+      case '_':
+      case '~':
+        /* allowed as-is */
+        result = Curl_dyn_addn(dq, q, 1);
+        break;
+      case '%':
+        /* uppercase the following if hexadecimal */
+        if(ISXDIGIT(q[1]) && ISXDIGIT(q[2])) {
+          char tmp[3]="%";
+          tmp[1] = Curl_raw_toupper(q[1]);
+          tmp[2] = Curl_raw_toupper(q[2]);
+          result = Curl_dyn_addn(dq, tmp, 3);
+          q += 2;
+          len -= 2;
+        }
+        else
+          /* '%' without a following two-digit hex, encode it */
+          result = Curl_dyn_addn(dq, "%25", 3);
+        break;
+      default: {
+        const char hex[] = "0123456789ABCDEF";
+        char out[3]={'%'};
+
+        if(!found_equals) {
+          /* if found_equals is NULL assuming, been in path */
+          if(*q == '/') {
+            /* allowed as if */
+            result = Curl_dyn_addn(dq, q, 1);
+            break;
+          }
+        }
+        else {
+          /* allowed as-is */
+          if(*q == '=') {
+            result = Curl_dyn_addn(dq, q, 1);
+            *found_equals = true;
+            break;
+          }
+        }
+        /* URL encode */
+        out[1] = hex[((unsigned char)*q)>>4];
+        out[2] = hex[*q & 0xf];
+        result = Curl_dyn_addn(dq, out, 3);
+        break;
+      }
+      }
+    }
+  }
+  return result;
+}
+
+
 static CURLcode canon_query(struct Curl_easy *data,
                             const char *query, struct dynbuf *dq)
 {
@@ -460,54 +530,11 @@ static CURLcode canon_query(struct Curl_easy *data,
 
   ap = &array[0];
   for(i = 0; !result && (i < entry); i++, ap++) {
-    size_t len;
     const char *q = ap->p;
     bool found_equals = false;
     if(!ap->len)
       continue;
-    for(len = ap->len; len && !result; q++, len--) {
-      if(ISALNUM(*q))
-        result = Curl_dyn_addn(dq, q, 1);
-      else {
-        switch(*q) {
-        case '-':
-        case '.':
-        case '_':
-        case '~':
-          /* allowed as-is */
-          result = Curl_dyn_addn(dq, q, 1);
-          break;
-        case '=':
-          /* allowed as-is */
-          result = Curl_dyn_addn(dq, q, 1);
-          found_equals = true;
-          break;
-        case '%':
-          /* uppercase the following if hexadecimal */
-          if(ISXDIGIT(q[1]) && ISXDIGIT(q[2])) {
-            char tmp[3]="%";
-            tmp[1] = Curl_raw_toupper(q[1]);
-            tmp[2] = Curl_raw_toupper(q[2]);
-            result = Curl_dyn_addn(dq, tmp, 3);
-            q += 2;
-            len -= 2;
-          }
-          else
-            /* '%' without a following two-digit hex, encode it */
-            result = Curl_dyn_addn(dq, "%25", 3);
-          break;
-        default: {
-          /* URL encode */
-          const char hex[] = "0123456789ABCDEF";
-          char out[3]={'%'};
-          out[1] = hex[((unsigned char)*q)>>4];
-          out[2] = hex[*q & 0xf];
-          result = Curl_dyn_addn(dq, out, 3);
-          break;
-        }
-        }
-      }
-    }
+    result = canon_string(q, ap->len, dq, &found_equals);
     if(!result && !found_equals) {
       /* queries without value still need an equals */
       result = Curl_dyn_addn(dq, "=", 1);
@@ -540,6 +567,7 @@ CURLcode Curl_output_aws_sigv4(struct Curl_easy *data, bool proxy)
   struct dynbuf canonical_headers;
   struct dynbuf signed_headers;
   struct dynbuf canonical_query;
+  struct dynbuf canonical_path;
   char *date_header = NULL;
   Curl_HttpReq httpreq;
   const char *method = NULL;
@@ -570,6 +598,7 @@ CURLcode Curl_output_aws_sigv4(struct Curl_easy *data, bool proxy)
   Curl_dyn_init(&canonical_headers, CURL_MAX_HTTP_HEADER);
   Curl_dyn_init(&canonical_query, CURL_MAX_HTTP_HEADER);
   Curl_dyn_init(&signed_headers, CURL_MAX_HTTP_HEADER);
+  Curl_dyn_init(&canonical_path, CURL_MAX_HTTP_HEADER);
 
   /*
    * Parameters parsing
@@ -698,6 +727,11 @@ CURLcode Curl_output_aws_sigv4(struct Curl_easy *data, bool proxy)
   result = canon_query(data, data->state.up.query, &canonical_query);
   if(result)
     goto fail;
+
+  result = canon_string(data->state.up.path, strlen(data->state.up.path),
+                        &canonical_path, NULL);
+  if(result)
+    goto fail;
   result = CURLE_OUT_OF_MEMORY;
 
   canonical_request =
@@ -708,7 +742,7 @@ CURLcode Curl_output_aws_sigv4(struct Curl_easy *data, bool proxy)
                   "%s\n" /* SignedHeaders */
                   "%.*s",  /* HashedRequestPayload in hex */
                   method,
-                  data->state.up.path,
+                  Curl_dyn_ptr(&canonical_path),
                   Curl_dyn_ptr(&canonical_query) ?
                   Curl_dyn_ptr(&canonical_query) : "",
                   Curl_dyn_ptr(&canonical_headers),
@@ -800,6 +834,7 @@ CURLcode Curl_output_aws_sigv4(struct Curl_easy *data, bool proxy)
 
 fail:
   Curl_dyn_free(&canonical_query);
+  Curl_dyn_free(&canonical_path);
   Curl_dyn_free(&canonical_headers);
   Curl_dyn_free(&signed_headers);
   free(canonical_request);

diff --git a/tests/data/Makefile.inc b/tests/data/Makefile.inc
index 079a33a9627ad6af45ce252af57b1ab6e64ddf9f..6352b44656bf9e2ab8bdaa53e5abd8d84da18a82 100644 (file)
--- a/tests/data/Makefile.inc
+++ b/tests/data/Makefile.inc
@@ -73,7 +73,7 @@ test426 test427 test428 test429 test430 test431 test432 test433 test434 \
 test435 test436 test437 test438 test439 test440 test441 test442 test443 \
 test444 test445 test446 test447 test448 test449 test450 test451 test452 \
 test453 test454 test455 test456 test457 test458 test459 test460 test461 \
-test462 test463 test467 test468 test469 test470 test471 \
+test462 test463 test467 test468 test469 test470 test471 test472 \
 \
 test490 test491 test492 test493 test494 test495 test496 test497 test498 \
 test499 test500 test501 test502 test503 test504 test505 test506 test507 \

diff --git a/tests/data/test472 b/tests/data/test472
new file mode 100644 (file)
index 0000000..721b9e8
--- /dev/null
+++ b/tests/data/test472
@@ -0,0 +1,59 @@
+<testcase>
+<info>
+<keywords>
+HTTP
+aws-sigv4
+</keywords>
+</info>
+
+#
+# Server-side
+<reply>
+<data crlf="yes">
+HTTP/1.1 200 OK
+Date: Tue, 09 Nov 2010 14:49:00 GMT
+Server: test-server/fake
+Last-Modified: Tue, 13 Jun 2000 12:10:00 GMT
+ETag: "21025-dc7-39462498"
+Accept-Ranges: bytes
+Content-Length: 6
+Connection: close
+Content-Type: text/html
+Funny-head: yesyes
+
+-foo-
+</data>
+</reply>
+
+#
+# Client-side
+<client>
+<server>
+http
+</server>
+<features>
+debug
+Unicode
+</features>
+<name>
+aws-sigv4 with query
+</name>
+<command>
+"http://fake.fake.fake:8000/%TESTNUMBER/a=あ" -u user:secret --aws-sigv4 "aws:amz:us-east-2:es" --connect-to fake.fake.fake:8000:%HOSTIP:%HTTPPORT
+</command>
+</client>
+
+#
+# Verify data after the test has been "shot"
+<verify>
+<protocol crlf="yes">
+GET /472/a=%e3%81%82 HTTP/1.1
+Host: fake.fake.fake:8000
+Authorization: AWS4-HMAC-SHA256 Credential=user/19700101/us-east-2/es/aws4_request, SignedHeaders=host;x-amz-date, Signature=c63315c199922f7ee00141869a250389405d19e205057249fb74726d940b1fc3
+X-Amz-Date: 19700101T000000Z
+User-Agent: curl/%VERSION
+Accept: */*
+
+</protocol>
+</verify>
+</testcase>