Skip to content
This repository has been archived by the owner on Nov 6, 2022. It is now read-only.

Support opaque urls #485

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Support opaque urls #485

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
2f77284
add cmake
ekoby Jun 27, 2019
c7edd84
support file, opague, and query-only(magnet:) URLS
ekoby Jun 27, 2019
d1afe57
support full scheme spec https://tools.ietf.org/html/rfc3986#section-3.1
ekoby Jul 2, 2019
81b9438
Rename UF_OPAQUE -> UF_OPAQ due to conflict on MacOS/BSD (defined in …
ekoby Jul 3, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 20 additions & 0 deletions CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
cmake_minimum_required(VERSION 3.14)
project(http_parser C)

set(CMAKE_C_STANDARD 11)

include_directories(.)

add_library(http_parser STATIC
http_parser.c)
target_include_directories(http_parser
PUBLIC ./)

add_executable(test test.c)
target_link_libraries(test PUBLIC http_parser)

add_executable(bench bench.c)
target_link_libraries(bench PUBLIC http_parser)

add_executable(url_parser contrib/url_parser.c)
target_link_libraries(url_parser PUBLIC http_parser)
31 changes: 25 additions & 6 deletions http_parser.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -311,6 +311,7 @@ enum state
, s_req_query_string
, s_req_fragment_start
, s_req_fragment
, s_req_opague
, s_req_http_start
, s_req_http_H
, s_req_http_HT
Expand Down Expand Up @@ -425,6 +426,8 @@ enum http_host_state
(c) == ';' || (c) == ':' || (c) == '&' || (c) == '=' || (c) == '+' || \
(c) == '$' || (c) == ',')

#define IS_SCHEME_CHAR(c) (IS_ALPHANUM(c) || c == '.' || c == '+' || c == '-')

#define STRICT_TOKEN(c) ((c == ' ') ? 0 : tokens[(unsigned char)c])

#if HTTP_PARSER_STRICT
Expand Down Expand Up @@ -517,7 +520,8 @@ parse_url_char(enum state s, const char ch)
break;

case s_req_schema:
if (IS_ALPHA(ch)) {
// scheme spec: https://tools.ietf.org/html/rfc3986#section-3.1
if (IS_SCHEME_CHAR(ch)) {
return s;
}

Expand All @@ -532,7 +536,18 @@ parse_url_char(enum state s, const char ch)
return s_req_schema_slash_slash;
}

break;
if (ch == '?') {
return s_req_query_string_start;
}

return s_req_opague;

case s_req_opague:
if (ch == '?') {
return s_req_query_string_start;
}

return s;

case s_req_schema_slash_slash:
if (ch == '/') {
Expand Down Expand Up @@ -2399,6 +2414,10 @@ http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
uf = UF_FRAGMENT;
break;

case s_req_opague:
uf = UF_OPAQ;
break;

default:
assert(!"Unexpected state");
return 1;
Expand All @@ -2419,10 +2438,10 @@ 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)) &&
(u->field_set & (1 << UF_HOST)) == 0) {
return 1;
}
// 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) {
Expand Down
3 changes: 2 additions & 1 deletion http_parser.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -346,7 +346,8 @@ enum http_parser_url_fields
, UF_QUERY = 4
, UF_FRAGMENT = 5
, UF_USERINFO = 6
, UF_MAX = 7
, UF_OPAQ = 7
, UF_MAX = 8
};


Expand Down
135 changes: 126 additions & 9 deletions test.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2801,6 +2801,7 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2820,6 +2821,7 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2839,16 +2841,17 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
}

, {.name="CONNECT request but not connect"
,.url="hostname:443"
,.is_connect=0
,.rv=1
}
//, {.name="CONNECT request but not connect"
// ,.url="hostname:443"
// ,.is_connect=0
// ,.rv=1
// }

, {.name="proxy ipv6 request"
,.url="http://[1:2::3:4]/"
Expand All @@ -2864,6 +2867,7 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2883,6 +2887,7 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2902,6 +2907,7 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2921,6 +2927,7 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2942,6 +2949,7 @@ const struct url_test url_tests[] =
,{ 30,187 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2961,6 +2969,7 @@ const struct url_test url_tests[] =
,{ 11, 10 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -2981,6 +2990,7 @@ const struct url_test url_tests[] =
,{ 0, 0 } /* UF_QUERY */
,{ 11, 4 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -3002,6 +3012,7 @@ const struct url_test url_tests[] =
,{ 36, 69 } /* UF_QUERY */
,{106, 7 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -3022,6 +3033,7 @@ const struct url_test url_tests[] =
,{ 29, 12 } /* UF_QUERY */
,{ 42, 4 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
Expand All @@ -3042,11 +3054,116 @@ const struct url_test url_tests[] =
,{ 33, 12 } /* UF_QUERY */
,{ 46, 4 } /* UF_FRAGMENT */
,{ 7, 3 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
}
, {.name="opaque URL: see https://golang.org/src/net/url/url_test.go#L136"
,.url="http:www.google.com/?q=go+language"
,.is_connect=0
,.u=
{.field_set= (1<<UF_SCHEMA) | (1<<UF_QUERY) | (1<<UF_OPAQ)
,.port=0
,.field_data=
{{ 0, 4 } /* UF_SCHEMA */
,{ 0, 0 } /* UF_HOST */
,{ 0, 0 } /* UF_PORT */
,{ 0, 0 } /* UF_PATH */
,{ 21, 13 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 5, 15 } /* UF_OPAQ */
}
}
,.rv=0
}

, {.name="opaque URL: see https://golang.org/src/net/url/url_test.go#L136"
,.url="mailto:[email protected]"
,.is_connect=0
,.u=
{.field_set= (1<<UF_SCHEMA) | (1<<UF_OPAQ)
,.port=0
,.field_data=
{{ 0, 6 } /* UF_SCHEMA */
,{ 0, 0 } /* UF_HOST */
,{ 0, 0 } /* UF_PORT */
,{ 0, 0 } /* UF_PATH */
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 7, 17 } /* UF_OPAQ */
}
}
,.rv=0
}
, {.name="opaque URL: see https://golang.org/src/net/url/url_test.go#L136"
,.url="magnet:?xt=urn:btih:c12fe1c06bba254a9dc9f519b335aa7c1367a88a&dn"
,.is_connect=0
,.u=
{.field_set= (1<<UF_SCHEMA) | (1<<UF_QUERY)
,.port=0
,.field_data=
{{ 0, 6 } /* UF_SCHEMA */
,{ 0, 0 } /* UF_HOST */
,{ 0, 0 } /* UF_PORT */
,{ 0, 0 } /* UF_PATH */
,{ 8, 55 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
}
, {.name="file url"
,.url="file:///tmp/data"
,.is_connect=0
,.u=
{.field_set= (1<<UF_SCHEMA) | (1<<UF_PATH)
,.port=0
,.field_data=
{{ 0, 4 } /* UF_SCHEMA */
,{ 0, 0 } /* UF_HOST */
,{ 0, 0 } /* UF_PORT */
,{ 7, 9 } /* UF_PATH */
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 0, 0 } /* UF_OPAQ */
}
}
,.rv=0
}

, {.name="scheme full test. https://tools.ietf.org/html/rfc3986#section-3.1"
,.url="sch3m3+full-test.v21:somethig_here"
,.is_connect=0
,.u=
{.field_set= (1<<UF_SCHEMA) | (1<<UF_OPAQ)
,.port=0
,.field_data=
{{ 0, 20 } /* UF_SCHEMA */
,{ 0, 0 } /* UF_HOST */
,{ 0, 0 } /* UF_PORT */
,{ 0, 0 } /* UF_PATH */
,{ 0, 0 } /* UF_QUERY */
,{ 0, 0 } /* UF_FRAGMENT */
,{ 0, 0 } /* UF_USERINFO */
,{ 21, 13 } /* UF_OPAQ */
}
}
,.rv=0
}

, {.name="scheme start with num. https://tools.ietf.org/html/rfc3986#section-3.1"
,.url="5ch3m3+full-test.v21:somethig_here"
,.is_connect=0
,.rv=1
}


, {.name="double @"
,.url="http://a:b@@hostname:443/"
,.is_connect=0
Expand Down Expand Up @@ -3208,10 +3325,10 @@ const struct url_test url_tests[] =
,.rv=1 /* s_dead */
}

, {.name="proxy emtpy hostname"
,.url="http:///fo"
,.rv=1 /* s_dead */
}
//, {.name="proxy emtpy hostname"
// ,.url="http:///fo"
// ,.rv=1 /* s_dead */
// }

, {.name="proxy = in URL"
,.url="http://host=ame/fo"
Expand Down