[jira] [Commented] (TS-4312) Adding config to parse urls according to RFC

["ASF GitHub Bot (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TS-4312?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15225467#comment-15225467 ] 

ASF GitHub Bot commented on TS-4312:
------------------------------------

Github user shenzhang920 commented on a diff in the pull request:

    https://github.com/apache/trafficserver/pull/541#discussion_r58474731
  
    --- Diff: proxy/hdrs/URL.cc ---
    @@ -1158,9 +1158,51 @@ url_parse_scheme(HdrHeap *heap, URLImpl *url, const char **start, const char *en
       return PARSE_ERROR; // no non-whitespace found
     }
     
    +static bool
    +url_init_non_encoded_char_array(char *non_encoded_char, size_t size)
    +{
    +  const char *allowed_char = ":/?#[]@"                     // gen-delims
    +                              "!$&'()*+,;="                // sub-delims
    +                              "ABCDEFGHIJKLMNOPQRSTUVWXYZ" // ALPHA
    +                              "abcdefghijklmnopqrstuvwxyz"
    +                              "0123456789"                 // DIGIT
    +                              "-._~"                       // other unreserved
    +                              "%";                         // '%' should also not be encoded
    +
    +  memset(non_encoded_char, 0, size);
    +  for (size_t i = 0; i < strlen(allowed_char); ++i) {
    +    ink_assert((unsigned char)allowed_char[i] < size);
    +    non_encoded_char[(unsigned char)allowed_char[i]] = 1;
    +  }
    +
    +  return true;
    +}
    +
    +static char non_encoded_char[256];
    +static bool url_init_non_encoded_char_array_done = url_init_non_encoded_char_array(non_encoded_char, sizeof(non_encoded_char));
    --- End diff --
    
    Hardcoding would look like as below, it is error prone.  Should I change it?
    
    static const char non_encoded_char[256] = {
        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
        0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // ! # $ % & ' ( ) * + , - . /
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, // 0-9 : ; = ?
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // @ A-O
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, // P-Z [ ] _
        0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // a-o
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0  // p-z ~
        // values of indices from 128 to 255 are all 0
      };



> Adding config to parse urls according to RFC
> --------------------------------------------
>
>                 Key: TS-4312
>                 URL: https://issues.apache.org/jira/browse/TS-4312
>             Project: Traffic Server
>          Issue Type: Bug
>            Reporter: Shen Zhang
>            Assignee: Brian Geffon
>             Fix For: 6.2.0
>
>
> Adding a config option "proxy.config.http.strict_uri_parsing" to sends http status code 400 back to client if the URL includes non-RFC 3986 compliant character



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)