Upgrade http-parser.

ext/http-parser/http_parser.c

@@ -400,6 +400,8 @@ enum http_host_state
   , s_http_host
   , s_http_host_v6
   , s_http_host_v6_end
+  , s_http_host_v6_zone_start
+  , s_http_host_v6_zone
   , s_http_host_port_start
   , s_http_host_port
 };
@@ -433,6 +435,12 @@ enum http_host_state
   (IS_ALPHANUM(c) || (c) == '.' || (c) == '-' || (c) == '_')
 #endif
 
+/**
+ * Verify that a char is a valid visible (printable) US-ASCII
+ * character or %x80-FF
+ **/
+#define IS_HEADER_CHAR(ch)                                                     \
+  (ch == CR || ch == LF || ch == 9 || (ch > 31 && ch != 127))
 
 #define start_state (parser->type == HTTP_REQUEST ? s_start_req : s_start_res)
 
@@ -637,6 +645,7 @@ size_t http_parser_execute (http_parser *parser,
   const char *body_mark = 0;
   const char *status_mark = 0;
   enum state p_state = (enum state) parser->state;
+  const unsigned int lenient = parser->lenient_http_headers;
 
   /* We're in an error state. Don't bother doing anything. */
   if (HTTP_PARSER_ERRNO(parser) != HPE_OK) {
@@ -957,21 +966,23 @@ reexecute:
         parser->method = (enum http_method) 0;
         parser->index = 1;
         switch (ch) {
+          case 'A': parser->method = HTTP_ACL; break;
+          case 'B': parser->method = HTTP_BIND; break;
           case 'C': parser->method = HTTP_CONNECT; /* or COPY, CHECKOUT */ break;
           case 'D': parser->method = HTTP_DELETE; break;
           case 'G': parser->method = HTTP_GET; break;
           case 'H': parser->method = HTTP_HEAD; break;
-          case 'L': parser->method = HTTP_LOCK; break;
+          case 'L': parser->method = HTTP_LOCK; /* or LINK */ break;
           case 'M': parser->method = HTTP_MKCOL; /* or MOVE, MKACTIVITY, MERGE, M-SEARCH, MKCALENDAR */ break;
           case 'N': parser->method = HTTP_NOTIFY; break;
           case 'O': parser->method = HTTP_OPTIONS; break;
           case 'P': parser->method = HTTP_POST;
             /* or PROPFIND|PROPPATCH|PUT|PATCH|PURGE */
             break;
-          case 'R': parser->method = HTTP_REPORT; break;
+          case 'R': parser->method = HTTP_REPORT; /* or REBIND */ break;
           case 'S': parser->method = HTTP_SUBSCRIBE; /* or SEARCH */ break;
           case 'T': parser->method = HTTP_TRACE; break;
-          case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE */ break;
+          case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE, UNBIND, UNLINK */ break;
           default:
             SET_ERRNO(HPE_INVALID_METHOD);
             goto error;
@@ -1027,16 +1038,32 @@ reexecute:
             SET_ERRNO(HPE_INVALID_METHOD);
             goto error;
           }
-        } else if (parser->index == 1 && parser->method == HTTP_POST) {
-          if (ch == 'R') {
-            parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */
-          } else if (ch == 'U') {
-            parser->method = HTTP_PUT; /* or HTTP_PURGE */
-          } else if (ch == 'A') {
-            parser->method = HTTP_PATCH;
-          } else {
-            SET_ERRNO(HPE_INVALID_METHOD);
-            goto error;
+        } else if (parser->method == HTTP_REPORT) {
+            if (parser->index == 2 && ch == 'B') {
+              parser->method = HTTP_REBIND;
+            } else {
+              SET_ERRNO(HPE_INVALID_METHOD);
+              goto error;
+            }
+        } else if (parser->index == 1) {
+          if (parser->method == HTTP_POST) {
+            if (ch == 'R') {
+              parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */
+            } else if (ch == 'U') {
+              parser->method = HTTP_PUT; /* or HTTP_PURGE */
+            } else if (ch == 'A') {
+              parser->method = HTTP_PATCH;
+            } else {
+              SET_ERRNO(HPE_INVALID_METHOD);
+              goto error;
+            }
+          } else if (parser->method == HTTP_LOCK) {
+            if (ch == 'I') {
+              parser->method = HTTP_LINK;
+            } else {
+              SET_ERRNO(HPE_INVALID_METHOD);
+              goto error;
+            }
           }
         } else if (parser->index == 2) {
           if (parser->method == HTTP_PUT) {
@@ -1049,6 +1076,8 @@ reexecute:
           } else if (parser->method == HTTP_UNLOCK) {
             if (ch == 'S') {
               parser->method = HTTP_UNSUBSCRIBE;
+            } else if(ch == 'B') {
+              parser->method = HTTP_UNBIND;
             } else {
               SET_ERRNO(HPE_INVALID_METHOD);
               goto error;
@@ -1059,6 +1088,8 @@ reexecute:
           }
         } else if (parser->index == 4 && parser->method == HTTP_PROPFIND && ch == 'P') {
           parser->method = HTTP_PROPPATCH;
+        } else if (parser->index == 3 && parser->method == HTTP_UNLOCK && ch == 'I') {
+          parser->method = HTTP_UNLINK;
         } else {
           SET_ERRNO(HPE_INVALID_METHOD);
           goto error;
@@ -1384,7 +1415,12 @@ reexecute:
                   || c != CONTENT_LENGTH[parser->index]) {
                 parser->header_state = h_general;
               } else if (parser->index == sizeof(CONTENT_LENGTH)-2) {
+                if (parser->flags & F_CONTENTLENGTH) {
+                  SET_ERRNO(HPE_UNEXPECTED_CONTENT_LENGTH);
+                  goto error;
+                }
                 parser->header_state = h_content_length;
+                parser->flags |= F_CONTENTLENGTH;
               }
               break;
 
@@ -1536,6 +1572,11 @@ reexecute:
             REEXECUTE();
           }
 
+          if (!lenient && !IS_HEADER_CHAR(ch)) {
+            SET_ERRNO(HPE_INVALID_HEADER_TOKEN);
+            goto error;
+          }
+
           c = LOWER(ch);
 
           switch (h_state) {
@@ -1703,7 +1744,10 @@ reexecute:
 
       case s_header_almost_done:
       {
-        STRICT_CHECK(ch != LF);
+        if (UNLIKELY(ch != LF)) {
+          SET_ERRNO(HPE_LF_EXPECTED);
+          goto error;
+        }
 
         UPDATE_STATE(s_header_value_lws);
         break;
@@ -1782,9 +1826,17 @@ reexecute:
 
         if (parser->flags & F_TRAILING) {
           /* End of a chunked request */
-          UPDATE_STATE(NEW_MESSAGE());
-          CALLBACK_NOTIFY(message_complete);
-          break;
+          UPDATE_STATE(s_message_done);
+          CALLBACK_NOTIFY_NOADVANCE(chunk_complete);
+          REEXECUTE();
+        }
+
+        /* Cannot use chunked encoding and a content-length header together
+           per the HTTP specification. */
+        if ((parser->flags & F_CHUNKED) &&
+            (parser->flags & F_CONTENTLENGTH)) {
+          SET_ERRNO(HPE_UNEXPECTED_CONTENT_LENGTH);
+          goto error;
         }
 
         UPDATE_STATE(s_headers_done);
@@ -1828,12 +1880,16 @@ reexecute:
 
       case s_headers_done:
       {
+        int hasBody;
         STRICT_CHECK(ch != LF);
 
         parser->nread = 0;
 
-        /* Exit, the rest of the connect is in a different protocol. */
-        if (parser->upgrade) {
+        hasBody = parser->flags & F_CHUNKED ||
+          (parser->content_length > 0 && parser->content_length != ULLONG_MAX);
+        if (parser->upgrade && (parser->method == HTTP_CONNECT ||
+                                (parser->flags & F_SKIPBODY) || !hasBody)) {
+          /* Exit, the rest of the message is in a different protocol. */
           UPDATE_STATE(NEW_MESSAGE());
           CALLBACK_NOTIFY(message_complete);
           RETURN((p - data) + 1);
@@ -1854,8 +1910,7 @@ reexecute:
             /* Content-Length header given and non-zero */
             UPDATE_STATE(s_body_identity);
           } else {
-            if (parser->type == HTTP_REQUEST ||
-                !http_message_needs_eof(parser)) {
+            if (!http_message_needs_eof(parser)) {
               /* Assume content-length 0 - read the next */
               UPDATE_STATE(NEW_MESSAGE());
               CALLBACK_NOTIFY(message_complete);
@@ -1915,6 +1970,10 @@ reexecute:
       case s_message_done:
         UPDATE_STATE(NEW_MESSAGE());
         CALLBACK_NOTIFY(message_complete);
+        if (parser->upgrade) {
+          /* Exit, the rest of the message is in a different protocol. */
+          RETURN((p - data) + 1);
+        }
         break;
 
       case s_chunk_size_start:
@@ -1994,6 +2053,7 @@ reexecute:
         } else {
           UPDATE_STATE(s_chunk_data);
         }
+        CALLBACK_NOTIFY(chunk_header);
         break;
       }
 
@@ -2033,6 +2093,7 @@ reexecute:
         STRICT_CHECK(ch != LF);
         parser->nread = 0;
         UPDATE_STATE(s_chunk_size_start);
+        CALLBACK_NOTIFY(chunk_complete);
         break;
 
       default:
@@ -2144,13 +2205,13 @@ http_parser_settings_init(http_parser_settings *settings)
 
 const char *
 http_errno_name(enum http_errno err) {
-  assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0])));
+  assert(((size_t) err) < ARRAY_SIZE(http_strerror_tab));
   return http_strerror_tab[err].name;
 }
 
 const char *
 http_errno_description(enum http_errno err) {
-  assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0])));
+  assert(((size_t) err) < ARRAY_SIZE(http_strerror_tab));
   return http_strerror_tab[err].description;
 }
 
@@ -2203,6 +2264,23 @@ http_parse_host_char(enum http_host_state s, const char ch) {
         return s_http_host_v6;
       }
 
+      if (s == s_http_host_v6 && ch == '%') {
+        return s_http_host_v6_zone_start;
+      }
+      break;
+
+    case s_http_host_v6_zone:
+      if (ch == ']') {
+        return s_http_host_v6_end;
+      }
+
+    /* FALLTHROUGH */
+    case s_http_host_v6_zone_start:
+      /* RFC 6874 Zone ID consists of 1*( unreserved / pct-encoded) */
+      if (IS_ALPHANUM(ch) || ch == '%' || ch == '.' || ch == '-' || ch == '_' ||
+          ch == '~') {
+        return s_http_host_v6_zone;
+      }
       break;
 
     case s_http_host_port:
@@ -2221,6 +2299,7 @@ http_parse_host_char(enum http_host_state s, const char ch) {
 
 static int
 http_parse_host(const char * buf, struct http_parser_url *u, int found_at) {
+  assert(u->field_set & (1 << UF_HOST));
   enum http_host_state s;
 
   const char *p;
@@ -2252,6 +2331,11 @@ http_parse_host(const char * buf, struct http_parser_url *u, int found_at) {
         u->field_data[UF_HOST].len++;
         break;
 
+      case s_http_host_v6_zone_start:
+      case s_http_host_v6_zone:
+        u->field_data[UF_HOST].len++;
+        break;
+
       case s_http_host_port:
         if (s != s_http_host_port) {
           u->field_data[UF_PORT].off = p - buf;
@@ -2281,6 +2365,8 @@ http_parse_host(const char * buf, struct http_parser_url *u, int found_at) {
     case s_http_host_start:
     case s_http_host_v6_start:
     case s_http_host_v6:
+    case s_http_host_v6_zone_start:
+    case s_http_host_v6_zone:
     case s_http_host_port_start:
     case s_http_userinfo:
     case s_http_userinfo_start:
@@ -2292,6 +2378,11 @@ http_parse_host(const char * buf, struct http_parser_url *u, int found_at) {
   return 0;
 }
 
+void
+http_parser_url_init(struct http_parser_url *u) {
+  memset(u, 0, sizeof(*u));
+}
+
 int
 http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
                       struct http_parser_url *u)
@@ -2365,7 +2456,12 @@ http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
 
   /* host must be present if there is a schema */
   /* parsing http:///toto will fail */
-  if ((u->field_set & ((1 << UF_SCHEMA) | (1 << UF_HOST))) != 0) {
+  if ((u->field_set & (1 << UF_SCHEMA)) &&
+      (u->field_set & (1 << UF_HOST)) == 0) {
+    return 1;
+  }
+
+  if (u->field_set & (1 << UF_HOST)) {
     if (http_parse_host(buf, u, found_at) != 0) {
       return 1;
     }

ext/http-parser/http_parser.h

@@ -26,11 +26,12 @@ extern "C" {
 
 /* Also update SONAME in the Makefile whenever you change these. */
 #define HTTP_PARSER_VERSION_MAJOR 2
-#define HTTP_PARSER_VERSION_MINOR 4
-#define HTTP_PARSER_VERSION_PATCH 2
+#define HTTP_PARSER_VERSION_MINOR 6
+#define HTTP_PARSER_VERSION_PATCH 1
 
 #include <sys/types.h>
-#if defined(_WIN32) && !defined(__MINGW32__) && (!defined(_MSC_VER) || _MSC_VER<1600)
+#if defined(_WIN32) && !defined(__MINGW32__) && \
+  (!defined(_MSC_VER) || _MSC_VER<1600) && !defined(__WINE__)
 #include <BaseTsd.h>
 #include <stddef.h>
 typedef __int8 int8_t;
@@ -95,7 +96,7 @@ typedef int (*http_cb) (http_parser*);
   XX(5,  CONNECT,     CONNECT)      \
   XX(6,  OPTIONS,     OPTIONS)      \
   XX(7,  TRACE,       TRACE)        \
-  /* webdav */                      \
+  /* WebDAV */                      \
   XX(8,  COPY,        COPY)         \
   XX(9,  LOCK,        LOCK)         \
   XX(10, MKCOL,       MKCOL)        \
@@ -104,21 +105,28 @@ typedef int (*http_cb) (http_parser*);
   XX(13, PROPPATCH,   PROPPATCH)    \
   XX(14, SEARCH,      SEARCH)       \
   XX(15, UNLOCK,      UNLOCK)       \
+  XX(16, BIND,        BIND)         \
+  XX(17, REBIND,      REBIND)       \
+  XX(18, UNBIND,      UNBIND)       \
+  XX(19, ACL,         ACL)          \
   /* subversion */                  \
-  XX(16, REPORT,      REPORT)       \
-  XX(17, MKACTIVITY,  MKACTIVITY)   \
-  XX(18, CHECKOUT,    CHECKOUT)     \
-  XX(19, MERGE,       MERGE)        \
+  XX(20, REPORT,      REPORT)       \
+  XX(21, MKACTIVITY,  MKACTIVITY)   \
+  XX(22, CHECKOUT,    CHECKOUT)     \
+  XX(23, MERGE,       MERGE)        \
   /* upnp */                        \
-  XX(20, MSEARCH,     M-SEARCH)     \
-  XX(21, NOTIFY,      NOTIFY)       \
-  XX(22, SUBSCRIBE,   SUBSCRIBE)    \
-  XX(23, UNSUBSCRIBE, UNSUBSCRIBE)  \
+  XX(24, MSEARCH,     M-SEARCH)     \
+  XX(25, NOTIFY,      NOTIFY)       \
+  XX(26, SUBSCRIBE,   SUBSCRIBE)    \
+  XX(27, UNSUBSCRIBE, UNSUBSCRIBE)  \
   /* RFC-5789 */                    \
-  XX(24, PATCH,       PATCH)        \
-  XX(25, PURGE,       PURGE)        \
+  XX(28, PATCH,       PATCH)        \
+  XX(29, PURGE,       PURGE)        \
   /* CalDAV */                      \
-  XX(26, MKCALENDAR,  MKCALENDAR)   \
+  XX(30, MKCALENDAR,  MKCALENDAR)   \
+  /* RFC-2068, section 19.6.1.2 */  \
+  XX(31, LINK,        LINK)         \
+  XX(32, UNLINK,      UNLINK)       \
 
 enum http_method
   {
@@ -140,11 +148,12 @@ enum flags
   , F_TRAILING              = 1 << 4
   , F_UPGRADE               = 1 << 5
   , F_SKIPBODY              = 1 << 6
+  , F_CONTENTLENGTH         = 1 << 7
   };
 
 
 /* Map for errno-related constants
- * 
+ *
  * The provided argument should be a macro that takes 2 arguments.
  */
 #define HTTP_ERRNO_MAP(XX)                                           \
@@ -160,6 +169,8 @@ enum flags
   XX(CB_body, "the on_body callback failed")                         \
   XX(CB_message_complete, "the on_message_complete callback failed") \
   XX(CB_status, "the on_status callback failed")                     \
+  XX(CB_chunk_header, "the on_chunk_header callback failed")         \
+  XX(CB_chunk_complete, "the on_chunk_complete callback failed")     \
                                                                      \
   /* Parsing-related errors */                                       \
   XX(INVALID_EOF_STATE, "stream ended at an unexpected time")        \
@@ -180,6 +191,8 @@ enum flags
   XX(INVALID_HEADER_TOKEN, "invalid character in header")            \
   XX(INVALID_CONTENT_LENGTH,                                         \
      "invalid character in content-length header")                   \
+  XX(UNEXPECTED_CONTENT_LENGTH,                                      \
+     "unexpected content-length header")                             \
   XX(INVALID_CHUNK_SIZE,                                             \
      "invalid character in chunk size header")                       \
   XX(INVALID_CONSTANT, "invalid constant string")                    \
@@ -204,10 +217,11 @@ enum http_errno {
 struct http_parser {
   /** PRIVATE **/
   unsigned int type : 2;         /* enum http_parser_type */
-  unsigned int flags : 7;        /* F_* values from 'flags' enum; semi-public */
+  unsigned int flags : 8;        /* F_* values from 'flags' enum; semi-public */
   unsigned int state : 7;        /* enum state from http_parser.c */
-  unsigned int header_state : 8; /* enum header_state from http_parser.c */
-  unsigned int index : 8;        /* index into current matcher */
+  unsigned int header_state : 7; /* enum header_state from http_parser.c */
+  unsigned int index : 7;        /* index into current matcher */
+  unsigned int lenient_http_headers : 1;
 
   uint32_t nread;          /* # bytes read in various scenarios */
   uint64_t content_length; /* # bytes in body (0 if no Content-Length header) */
@@ -240,6 +254,11 @@ struct http_parser_settings {
   http_cb      on_headers_complete;
   http_data_cb on_body;
   http_cb      on_message_complete;
+  /* When on_chunk_header is called, the current chunk length is stored
+   * in parser->content_length.
+   */
+  http_cb      on_chunk_header;
+  http_cb      on_chunk_complete;
 };
 
 
@@ -318,6 +337,9 @@ const char *http_errno_name(enum http_errno err);
 /* Return a string description of the given error */
 const char *http_errno_description(enum http_errno err);
 
+/* Initialize all http_parser_url members to 0 */
+void http_parser_url_init(struct http_parser_url *u);
+
 /* Parse a URL; return nonzero on failure */
 int http_parser_parse_url(const char *buf, size_t buflen,
                           int is_connect,