From a0f2c4514e969d0a127227201cbdb8e57f71df63 Mon Sep 17 00:00:00 2001 From: Eric Wong Date: Sun, 5 Jul 2009 15:46:52 -0700 Subject: Rename unicorn/http11 => unicorn_http We couldn't do proper namespacing for the C module so there was a potential conflict with Init_http11() in Mongrel. This was needed because Mongrel's HTTP parser could be used in some applications and we may be unfortunate enough need to support them. --- .gitignore | 2 +- GNUmakefile | 22 +- Manifest | 12 +- Rakefile | 10 +- ext/unicorn/http11/ext_help.h | 12 - ext/unicorn/http11/extconf.rb | 5 - ext/unicorn/http11/http11.c | 457 ---------- ext/unicorn/http11/http11_parser.h | 1289 ---------------------------- ext/unicorn/http11/http11_parser.rl | 158 ---- ext/unicorn/http11/http11_parser_common.rl | 58 -- ext/unicorn_http/ext_help.h | 12 + ext/unicorn_http/extconf.rb | 5 + ext/unicorn_http/unicorn_http.c | 457 ++++++++++ ext/unicorn_http/unicorn_http.h | 1289 ++++++++++++++++++++++++++++ ext/unicorn_http/unicorn_http.rl | 158 ++++ ext/unicorn_http/unicorn_http_common.rl | 58 ++ lib/unicorn/chunked_reader.rb | 2 +- lib/unicorn/http_request.rb | 2 +- lib/unicorn/trailer_parser.rb | 2 +- test/test_helper.rb | 2 +- test/unit/test_chunked_reader.rb | 2 +- test/unit/test_trailer_parser.rb | 2 +- 22 files changed, 2008 insertions(+), 2008 deletions(-) delete mode 100644 ext/unicorn/http11/ext_help.h delete mode 100644 ext/unicorn/http11/extconf.rb delete mode 100644 ext/unicorn/http11/http11.c delete mode 100644 ext/unicorn/http11/http11_parser.h delete mode 100644 ext/unicorn/http11/http11_parser.rl delete mode 100644 ext/unicorn/http11/http11_parser_common.rl create mode 100644 ext/unicorn_http/ext_help.h create mode 100644 ext/unicorn_http/extconf.rb create mode 100644 ext/unicorn_http/unicorn_http.c create mode 100644 ext/unicorn_http/unicorn_http.h create mode 100644 ext/unicorn_http/unicorn_http.rl create mode 100644 ext/unicorn_http/unicorn_http_common.rl diff --git a/.gitignore b/.gitignore index 8588a5c..04258bb 100644 --- a/.gitignore +++ b/.gitignore @@ -7,7 +7,7 @@ /doc /local.mk /test/install-* -ext/unicorn/http11/Makefile +ext/unicorn_http/Makefile log/ pkg/ /vendor diff --git a/GNUmakefile b/GNUmakefile index 6a9bd7a..72984b6 100644 --- a/GNUmakefile +++ b/GNUmakefile @@ -27,30 +27,30 @@ T_n_log := $(subst .n,$(log_suffix),$(T_n)) T_r_log := $(subst .r,$(log_suffix),$(T_r)) test_prefix = $(CURDIR)/test/install-$(RUBY_VERSION) -ext := ext/unicorn/http11 -c_files := $(addprefix $(ext)/,ext_help.h http11.c http11_parser.h) -rl_files := $(addprefix $(ext)/,http11_parser.rl http11_parser_common.rl) +ext := ext/unicorn_http +c_files := $(addprefix $(ext)/,ext_help.h unicorn_http.c unicorn_http.h) +rl_files := $(addprefix $(ext)/,unicorn_http.rl unicorn_http_common.rl) rb_files := $(shell grep '^\(bin\|lib\)' Manifest) inst_deps := $(c_files) $(rb_files) -ragel: $(ext)/http11_parser.h -$(ext)/http11_parser.h: $(rl_files) - cd $(@D) && $(ragel) http11_parser.rl -C $(RLFLAGS) -o $(@F) +ragel: $(ext)/unicorn_http.h +$(ext)/unicorn_http.h: $(rl_files) + cd $(@D) && $(ragel) unicorn_http.rl -C $(RLFLAGS) -o $(@F) $(ruby) -i -p -e '$$_.gsub!(%r{[ \t]*$$},"")' $@ $(ext)/Makefile: $(ext)/extconf.rb $(c_files) cd $(@D) && $(ruby) extconf.rb -$(ext)/http11.$(DLEXT): $(ext)/Makefile +$(ext)/unicorn_http.$(DLEXT): $(ext)/Makefile $(MAKE) -C $(@D) -lib/unicorn/http11.$(DLEXT): $(ext)/http11.$(DLEXT) +lib/unicorn_http.$(DLEXT): $(ext)/unicorn_http.$(DLEXT) @mkdir -p lib install -m644 $< $@ -http11: lib/unicorn/http11.$(DLEXT) +http: lib/unicorn_http.$(DLEXT) $(test_prefix)/.stamp: $(inst_deps) mkdir -p $(test_prefix)/.ccache tar c bin ext lib GNUmakefile Manifest | (cd $(test_prefix) && tar x) $(MAKE) -C $(test_prefix) clean - $(MAKE) -C $(test_prefix) http11 shebang + $(MAKE) -C $(test_prefix) http shebang > $@ bins := $(wildcard bin/*) @@ -119,7 +119,7 @@ prep_setup_rb := @-$(RM) $(setup_rb_files);$(MAKE) -C $(ext) clean clean: -$(MAKE) -C $(ext) clean - $(RM) $(ext)/Makefile lib/unicorn/http11.$(DLEXT) + $(RM) $(ext)/Makefile lib/unicorn_http.$(DLEXT) $(RM) $(setup_rb_files) $(t_log) $(RM) -r $(test_prefix) diff --git a/Manifest b/Manifest index 72ff78d..73d1a2b 100644 --- a/Manifest +++ b/Manifest @@ -18,12 +18,12 @@ bin/unicorn_rails examples/echo.ru examples/git.ru examples/init.sh -ext/unicorn/http11/ext_help.h -ext/unicorn/http11/extconf.rb -ext/unicorn/http11/http11.c -ext/unicorn/http11/http11_parser.h -ext/unicorn/http11/http11_parser.rl -ext/unicorn/http11/http11_parser_common.rl +ext/unicorn_http/ext_help.h +ext/unicorn_http/extconf.rb +ext/unicorn_http/unicorn_http.c +ext/unicorn_http/unicorn_http.h +ext/unicorn_http/unicorn_http.rl +ext/unicorn_http/unicorn_http_common.rl lib/unicorn.rb lib/unicorn/app/exec_cgi.rb lib/unicorn/app/inetd.rb diff --git a/Rakefile b/Rakefile index e7b4c36..29e2d41 100644 --- a/Rakefile +++ b/Rakefile @@ -6,9 +6,9 @@ Echoe.new("unicorn") do |p| p.summary = "Rack HTTP server for Unix, fast clients and nothing else" p.author = "Eric Wong" p.email = "normalperson@yhbt.net" - p.clean_pattern = ['ext/unicorn/http11/*.{bundle,so,o,obj,pdb,lib,def,exp}', + p.clean_pattern = ['ext/unicorn_http/*.{bundle,so,o,obj,pdb,lib,def,exp}', 'lib/*.{bundle,so,o,obj,pdb,lib,def,exp}', - 'ext/unicorn/http11/Makefile', + 'ext/unicorn_http/Makefile', 'pkg', 'lib/*.bundle', '*.gem', 'site/output', '.config', 'coverage', 'test_*.log', 'log', 'doc'] @@ -29,10 +29,10 @@ end desc "Rebuild the Ragel sources" task :ragel do - Dir.chdir "ext/unicorn/http11" do - target = "http11_parser.c" + Dir.chdir "ext/unicorn_http" do + target = "unicorn_http.c" File.unlink target if File.exist? target - sh "ragel http11_parser.rl -C -G2 -o #{target}" + sh "ragel unicorn_http.rl -C -G2 -o #{target}" raise "Failed to build C source" unless File.exist? target end end diff --git a/ext/unicorn/http11/ext_help.h b/ext/unicorn/http11/ext_help.h deleted file mode 100644 index 17f7b01..0000000 --- a/ext/unicorn/http11/ext_help.h +++ /dev/null @@ -1,12 +0,0 @@ -#ifndef ext_help_h -#define ext_help_h - -#define ARRAY_SIZE(x) (sizeof(x)/sizeof(x[0])) - -#ifdef DEBUG -#define TRACE() fprintf(stderr, "> %s:%d:%s\n", __FILE__, __LINE__, __FUNCTION__) -#else -#define TRACE() -#endif - -#endif diff --git a/ext/unicorn/http11/extconf.rb b/ext/unicorn/http11/extconf.rb deleted file mode 100644 index 1b16dff..0000000 --- a/ext/unicorn/http11/extconf.rb +++ /dev/null @@ -1,5 +0,0 @@ -require 'mkmf' - -dir_config("unicorn/http11") -have_library("c", "main") -create_makefile("unicorn/http11") diff --git a/ext/unicorn/http11/http11.c b/ext/unicorn/http11/http11.c deleted file mode 100644 index f640a08..0000000 --- a/ext/unicorn/http11/http11.c +++ /dev/null @@ -1,457 +0,0 @@ -/** - * Copyright (c) 2009 Eric Wong (all bugs are Eric's fault) - * Copyright (c) 2005 Zed A. Shaw - * You can redistribute it and/or modify it under the same terms as Ruby. - */ -#include "ruby.h" -#include "ext_help.h" -#include -#include -#include "http11_parser.h" - -static http_parser *data_get(VALUE self) -{ - http_parser *http; - - Data_Get_Struct(self, http_parser, http); - if (!http) - rb_raise(rb_eArgError, "NULL found for http when shouldn't be."); - return http; -} - -#ifndef RSTRING_PTR -#define RSTRING_PTR(s) (RSTRING(s)->ptr) -#endif -#ifndef RSTRING_LEN -#define RSTRING_LEN(s) (RSTRING(s)->len) -#endif - -static VALUE mUnicorn; -static VALUE cHttpParser; -static VALUE eHttpParserError; -static VALUE sym_http_body; - -#define HTTP_PREFIX "HTTP_" -#define HTTP_PREFIX_LEN (sizeof(HTTP_PREFIX) - 1) - -static VALUE global_rack_url_scheme; -static VALUE global_request_method; -static VALUE global_request_uri; -static VALUE global_fragment; -static VALUE global_query_string; -static VALUE global_http_version; -static VALUE global_request_path; -static VALUE global_path_info; -static VALUE global_server_name; -static VALUE global_server_port; -static VALUE global_server_protocol; -static VALUE global_server_protocol_value; -static VALUE global_http_host; -static VALUE global_http_x_forwarded_proto; -static VALUE global_port_80; -static VALUE global_port_443; -static VALUE global_localhost; -static VALUE global_http; - -/** Defines common length and error messages for input length validation. */ -#define DEF_MAX_LENGTH(N, length) \ - static const size_t MAX_##N##_LENGTH = length; \ - static const char * const MAX_##N##_LENGTH_ERR = \ - "HTTP element " # N " is longer than the " # length " allowed length." - -/** - * Validates the max length of given input and throws an HttpParserError - * exception if over. - */ -#define VALIDATE_MAX_LENGTH(len, N) do { \ - if (len > MAX_##N##_LENGTH) \ - rb_raise(eHttpParserError, MAX_##N##_LENGTH_ERR); \ -} while (0) - -/** Defines global strings in the init method. */ -#define DEF_GLOBAL(N, val) do { \ - global_##N = rb_obj_freeze(rb_str_new(val, sizeof(val) - 1)); \ - rb_global_variable(&global_##N); \ -} while (0) - -/* Defines the maximum allowed lengths for various input elements.*/ -DEF_MAX_LENGTH(FIELD_NAME, 256); -DEF_MAX_LENGTH(FIELD_VALUE, 80 * 1024); -DEF_MAX_LENGTH(REQUEST_URI, 1024 * 12); -DEF_MAX_LENGTH(FRAGMENT, 1024); /* Don't know if this length is specified somewhere or not */ -DEF_MAX_LENGTH(REQUEST_PATH, 1024); -DEF_MAX_LENGTH(QUERY_STRING, (1024 * 10)); -DEF_MAX_LENGTH(HEADER, (1024 * (80 + 32))); - -struct common_field { - const signed long len; - const char *name; - VALUE value; -}; - -/* - * A list of common HTTP headers we expect to receive. - * This allows us to avoid repeatedly creating identical string - * objects to be used with rb_hash_aset(). - */ -static struct common_field common_http_fields[] = { -# define f(N) { (sizeof(N) - 1), N, Qnil } - f("ACCEPT"), - f("ACCEPT_CHARSET"), - f("ACCEPT_ENCODING"), - f("ACCEPT_LANGUAGE"), - f("ALLOW"), - f("AUTHORIZATION"), - f("CACHE_CONTROL"), - f("CONNECTION"), - f("CONTENT_ENCODING"), - f("CONTENT_LENGTH"), - f("CONTENT_TYPE"), - f("COOKIE"), - f("DATE"), - f("EXPECT"), - f("FROM"), - f("HOST"), - f("IF_MATCH"), - f("IF_MODIFIED_SINCE"), - f("IF_NONE_MATCH"), - f("IF_RANGE"), - f("IF_UNMODIFIED_SINCE"), - f("KEEP_ALIVE"), /* Firefox sends this */ - f("MAX_FORWARDS"), - f("PRAGMA"), - f("PROXY_AUTHORIZATION"), - f("RANGE"), - f("REFERER"), - f("TE"), - f("TRAILER"), - f("TRANSFER_ENCODING"), - f("UPGRADE"), - f("USER_AGENT"), - f("VIA"), - f("X_FORWARDED_FOR"), /* common for proxies */ - f("X_FORWARDED_PROTO"), /* common for proxies */ - f("X_REAL_IP"), /* common for proxies */ - f("WARNING") -# undef f -}; - -/* this function is not performance-critical */ -static void init_common_fields(void) -{ - int i; - struct common_field *cf = common_http_fields; - char tmp[256]; /* MAX_FIELD_NAME_LENGTH */ - memcpy(tmp, HTTP_PREFIX, HTTP_PREFIX_LEN); - - for(i = 0; i < ARRAY_SIZE(common_http_fields); cf++, i++) { - /* Rack doesn't like certain headers prefixed with "HTTP_" */ - if (!strcmp("CONTENT_LENGTH", cf->name) || - !strcmp("CONTENT_TYPE", cf->name)) { - cf->value = rb_str_new(cf->name, cf->len); - } else { - memcpy(tmp + HTTP_PREFIX_LEN, cf->name, cf->len + 1); - cf->value = rb_str_new(tmp, HTTP_PREFIX_LEN + cf->len); - } - cf->value = rb_obj_freeze(cf->value); - rb_global_variable(&cf->value); - } -} - -static VALUE find_common_field_value(const char *field, size_t flen) -{ - int i; - struct common_field *cf = common_http_fields; - for(i = 0; i < ARRAY_SIZE(common_http_fields); i++, cf++) { - if (cf->len == flen && !memcmp(cf->name, field, flen)) - return cf->value; - } - return Qnil; -} - -static void http_field(void *data, const char *field, - size_t flen, const char *value, size_t vlen) -{ - VALUE req = (VALUE)data; - VALUE f = Qnil; - - VALIDATE_MAX_LENGTH(flen, FIELD_NAME); - VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE); - - f = find_common_field_value(field, flen); - - if (f == Qnil) { - /* - * We got a strange header that we don't have a memoized value for. - * Fallback to creating a new string to use as a hash key. - * - * using rb_str_new(NULL, len) here is faster than rb_str_buf_new(len) - * in my testing, because: there's no minimum allocation length (and - * no check for it, either), RSTRING_LEN(f) does not need to be - * written twice, and and RSTRING_PTR(f) will already be - * null-terminated for us. - */ - f = rb_str_new(NULL, HTTP_PREFIX_LEN + flen); - memcpy(RSTRING_PTR(f), HTTP_PREFIX, HTTP_PREFIX_LEN); - memcpy(RSTRING_PTR(f) + HTTP_PREFIX_LEN, field, flen); - assert(*(RSTRING_PTR(f) + RSTRING_LEN(f)) == '\0'); /* paranoia */ - /* fprintf(stderr, "UNKNOWN HEADER <%s>\n", RSTRING_PTR(f)); */ - } else if (f == global_http_host && rb_hash_aref(req, f) != Qnil) { - return; - } - - rb_hash_aset(req, f, rb_str_new(value, vlen)); -} - -static void request_method(void *data, const char *at, size_t length) -{ - VALUE req = (VALUE)data; - VALUE val = Qnil; - - val = rb_str_new(at, length); - rb_hash_aset(req, global_request_method, val); -} - -static void scheme(void *data, const char *at, size_t length) -{ - rb_hash_aset((VALUE)data, global_rack_url_scheme, rb_str_new(at, length)); -} - -static void host(void *data, const char *at, size_t length) -{ - rb_hash_aset((VALUE)data, global_http_host, rb_str_new(at, length)); -} - -static void request_uri(void *data, const char *at, size_t length) -{ - VALUE req = (VALUE)data; - VALUE val = Qnil; - - VALIDATE_MAX_LENGTH(length, REQUEST_URI); - - val = rb_str_new(at, length); - rb_hash_aset(req, global_request_uri, val); - - /* "OPTIONS * HTTP/1.1\r\n" is a valid request */ - if (length == 1 && *at == '*') { - val = rb_str_new(NULL, 0); - rb_hash_aset(req, global_request_path, val); - rb_hash_aset(req, global_path_info, val); - } -} - -static void fragment(void *data, const char *at, size_t length) -{ - VALUE req = (VALUE)data; - VALUE val = Qnil; - - VALIDATE_MAX_LENGTH(length, FRAGMENT); - - val = rb_str_new(at, length); - rb_hash_aset(req, global_fragment, val); -} - -static void request_path(void *data, const char *at, size_t length) -{ - VALUE req = (VALUE)data; - VALUE val = Qnil; - - VALIDATE_MAX_LENGTH(length, REQUEST_PATH); - - val = rb_str_new(at, length); - rb_hash_aset(req, global_request_path, val); - - /* rack says PATH_INFO must start with "/" or be empty */ - if (!(length == 1 && *at == '*')) - rb_hash_aset(req, global_path_info, val); -} - -static void query_string(void *data, const char *at, size_t length) -{ - VALUE req = (VALUE)data; - VALUE val = Qnil; - - VALIDATE_MAX_LENGTH(length, QUERY_STRING); - - val = rb_str_new(at, length); - rb_hash_aset(req, global_query_string, val); -} - -static void http_version(void *data, const char *at, size_t length) -{ - VALUE req = (VALUE)data; - VALUE val = rb_str_new(at, length); - rb_hash_aset(req, global_http_version, val); -} - -/** Finalizes the request header to have a bunch of stuff that's needed. */ -static void header_done(void *data, const char *at, size_t length) -{ - VALUE req = (VALUE)data; - VALUE server_name = global_localhost; - VALUE server_port = global_port_80; - VALUE temp; - - /* rack requires QUERY_STRING */ - if (rb_hash_aref(req, global_query_string) == Qnil) - rb_hash_aset(req, global_query_string, rb_str_new(NULL, 0)); - - /* set rack.url_scheme to "https" or "http", no others are allowed by Rack */ - if ((temp = rb_hash_aref(req, global_rack_url_scheme)) == Qnil) { - if ((temp = rb_hash_aref(req, global_http_x_forwarded_proto)) != Qnil && - RSTRING_LEN(temp) == 5 && - !memcmp("https", RSTRING_PTR(temp), 5)) - server_port = global_port_443; - else - temp = global_http; - rb_hash_aset(req, global_rack_url_scheme, temp); - } else if (RSTRING_LEN(temp) == 5 && !memcmp("https", RSTRING_PTR(temp), 5)) { - server_port = global_port_443; - } - - /* parse and set the SERVER_NAME and SERVER_PORT variables */ - if ((temp = rb_hash_aref(req, global_http_host)) != Qnil) { - char *colon = memchr(RSTRING_PTR(temp), ':', RSTRING_LEN(temp)); - if (colon) { - long port_start = colon - RSTRING_PTR(temp) + 1; - - server_name = rb_str_substr(temp, 0, colon - RSTRING_PTR(temp)); - if ((RSTRING_LEN(temp) - port_start) > 0) - server_port = rb_str_substr(temp, port_start, RSTRING_LEN(temp)); - } else { - server_name = temp; - } - } - rb_hash_aset(req, global_server_name, server_name); - rb_hash_aset(req, global_server_port, server_port); - rb_hash_aset(req, global_server_protocol, global_server_protocol_value); - - /* grab the initial body and stuff it into the hash */ - temp = rb_hash_aref(req, global_request_method); - if (temp != Qnil) { - long len = RSTRING_LEN(temp); - char *ptr = RSTRING_PTR(temp); - - if (memcmp(ptr, "HEAD", len) && memcmp(ptr, "GET", len)) - rb_hash_aset(req, sym_http_body, rb_str_new(at, length)); - } -} - -static void HttpParser_free(void *data) { - TRACE(); - - if(data) { - free(data); - } -} - - -static VALUE HttpParser_alloc(VALUE klass) -{ - VALUE obj; - http_parser *hp = ALLOC_N(http_parser, 1); - TRACE(); - http_parser_init(hp); - - obj = Data_Wrap_Struct(klass, NULL, HttpParser_free, hp); - - return obj; -} - - -/** - * call-seq: - * parser.new -> parser - * - * Creates a new parser. - */ -static VALUE HttpParser_init(VALUE self) -{ - http_parser_init(data_get(self)); - - return self; -} - - -/** - * call-seq: - * parser.reset -> nil - * - * Resets the parser to it's initial state so that you can reuse it - * rather than making new ones. - */ -static VALUE HttpParser_reset(VALUE self) -{ - http_parser_init(data_get(self)); - - return Qnil; -} - - -/** - * call-seq: - * parser.execute(req_hash, data) -> true/false - * - * Takes a Hash and a String of data, parses the String of data filling - * in the Hash returning a boolean to indicate whether or not parsing - * is finished. - * - * This function now throws an exception when there is a parsing error. - * This makes the logic for working with the parser much easier. You - * will need to wrap the parser with an exception handling block. - */ - -static VALUE HttpParser_execute(VALUE self, VALUE req_hash, VALUE data) -{ - http_parser *http = data_get(self); - char *dptr = RSTRING_PTR(data); - long dlen = RSTRING_LEN(data); - - if (http->nread < dlen) { - http->data = (void *)req_hash; - http_parser_execute(http, dptr, dlen); - - VALIDATE_MAX_LENGTH(http->nread, HEADER); - - if (!http_parser_has_error(http)) - return http_parser_is_finished(http) ? Qtrue : Qfalse; - - rb_raise(eHttpParserError, "Invalid HTTP format, parsing fails."); - } - rb_raise(eHttpParserError, "Requested start is after data buffer end."); -} - -void Init_http11(void) -{ - mUnicorn = rb_define_module("Unicorn"); - - DEF_GLOBAL(rack_url_scheme, "rack.url_scheme"); - DEF_GLOBAL(request_method, "REQUEST_METHOD"); - DEF_GLOBAL(request_uri, "REQUEST_URI"); - DEF_GLOBAL(fragment, "FRAGMENT"); - DEF_GLOBAL(query_string, "QUERY_STRING"); - DEF_GLOBAL(http_version, "HTTP_VERSION"); - DEF_GLOBAL(request_path, "REQUEST_PATH"); - DEF_GLOBAL(path_info, "PATH_INFO"); - DEF_GLOBAL(server_name, "SERVER_NAME"); - DEF_GLOBAL(server_port, "SERVER_PORT"); - DEF_GLOBAL(server_protocol, "SERVER_PROTOCOL"); - DEF_GLOBAL(server_protocol_value, "HTTP/1.1"); - DEF_GLOBAL(http_x_forwarded_proto, "HTTP_X_FORWARDED_PROTO"); - DEF_GLOBAL(port_80, "80"); - DEF_GLOBAL(port_443, "443"); - DEF_GLOBAL(localhost, "localhost"); - DEF_GLOBAL(http, "http"); - - eHttpParserError = rb_define_class_under(mUnicorn, "HttpParserError", rb_eIOError); - - cHttpParser = rb_define_class_under(mUnicorn, "HttpParser", rb_cObject); - rb_define_alloc_func(cHttpParser, HttpParser_alloc); - rb_define_method(cHttpParser, "initialize", HttpParser_init,0); - rb_define_method(cHttpParser, "reset", HttpParser_reset,0); - rb_define_method(cHttpParser, "execute", HttpParser_execute,2); - sym_http_body = ID2SYM(rb_intern("http_body")); - init_common_fields(); - global_http_host = find_common_field_value("HOST", 4); - assert(global_http_host != Qnil); -} diff --git a/ext/unicorn/http11/http11_parser.h b/ext/unicorn/http11/http11_parser.h deleted file mode 100644 index 8d95c59..0000000 --- a/ext/unicorn/http11/http11_parser.h +++ /dev/null @@ -1,1289 +0,0 @@ - -#line 1 "http11_parser.rl" -/** - * Copyright (c) 2005 Zed A. Shaw - * You can redistribute it and/or modify it under the same terms as Ruby. - */ -#ifndef http11_parser_h -#define http11_parser_h - -#include - -static void http_field(void *data, const char *field, - size_t flen, const char *value, size_t vlen); -static void request_method(void *data, const char *at, size_t length); -static void scheme(void *data, const char *at, size_t length); -static void host(void *data, const char *at, size_t length); -static void request_uri(void *data, const char *at, size_t length); -static void fragment(void *data, const char *at, size_t length); -static void request_path(void *data, const char *at, size_t length); -static void query_string(void *data, const char *at, size_t length); -static void http_version(void *data, const char *at, size_t length); -static void header_done(void *data, const char *at, size_t length); - -typedef struct http_parser { - int cs; - size_t body_start; - size_t nread; - size_t mark; - size_t field_start; - size_t field_len; - size_t query_start; - - void *data; -} http_parser; - -static int http_parser_has_error(http_parser *parser); -static int http_parser_is_finished(http_parser *parser); - -/* - * capitalizes all lower-case ASCII characters, - * converts dashes to underscores. - */ -static void snake_upcase_char(char *c) -{ - if (*c >= 'a' && *c <= 'z') - *c &= ~0x20; - else if (*c == '-') - *c = '_'; -} - -static void downcase_char(char *c) -{ - if (*c >= 'A' && *c <= 'Z') - *c |= 0x20; -} - -#define LEN(AT, FPC) (FPC - buffer - parser->AT) -#define MARK(M,FPC) (parser->M = (FPC) - buffer) -#define PTR_TO(F) (buffer + parser->F) - -/** Machine **/ - - -#line 109 "http11_parser.rl" - - -/** Data **/ - -#line 70 "http11_parser.h" -static const int http_parser_start = 1; -static const int http_parser_first_final = 63; -static const int http_parser_error = 0; - -static const int http_parser_en_main = 1; - - -#line 113 "http11_parser.rl" - -static void http_parser_init(http_parser *parser) { - int cs = 0; - memset(parser, 0, sizeof(*parser)); - -#line 84 "http11_parser.h" - { - cs = http_parser_start; - } - -#line 118 "http11_parser.rl" - parser->cs = cs; -} - -/** exec **/ -static void http_parser_execute( - http_parser *parser, const char *buffer, size_t len) -{ - const char *p, *pe; - int cs = parser->cs; - size_t off = parser->nread; - - assert(off <= len && "offset past end of buffer"); - - p = buffer+off; - pe = buffer+len; - - assert(*pe == '\0' && "pointer does not end on NUL"); - assert(pe - p == len - off && "pointers aren't same distance"); - - -#line 110 "http11_parser.h" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -case 1: - switch( (*p) ) { - case 36: goto tr0; - case 95: goto tr0; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto tr0; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto tr0; - } else - goto tr0; - goto st0; -st0: -cs = 0; - goto _out; -tr0: -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st2; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: -#line 141 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st44; - case 95: goto st44; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st44; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st44; - } else - goto st44; - goto st0; -tr2: -#line 77 "http11_parser.rl" - { - request_method(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st3; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: -#line 166 "http11_parser.h" - switch( (*p) ) { - case 42: goto tr4; - case 47: goto tr5; - case 72: goto tr6; - case 104: goto tr6; - } - goto st0; -tr4: -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st4; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: -#line 182 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr7; - case 35: goto tr8; - } - goto st0; -tr7: -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st5; -tr30: -#line 64 "http11_parser.rl" - {MARK(mark, p); } -#line 85 "http11_parser.rl" - { - fragment(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st5; -tr33: -#line 85 "http11_parser.rl" - { - fragment(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st5; -tr37: -#line 98 "http11_parser.rl" - { - request_path(parser->data, PTR_TO(mark), LEN(mark,p)); - } -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st5; -tr48: -#line 89 "http11_parser.rl" - {MARK(query_start, p); } -#line 90 "http11_parser.rl" - { - query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); - } -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st5; -tr52: -#line 90 "http11_parser.rl" - { - query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); - } -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st5; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: -#line 244 "http11_parser.h" - if ( (*p) == 72 ) - goto tr9; - goto st0; -tr9: -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st6; -st6: - if ( ++p == pe ) - goto _test_eof6; -case 6: -#line 256 "http11_parser.h" - if ( (*p) == 84 ) - goto st7; - goto st0; -st7: - if ( ++p == pe ) - goto _test_eof7; -case 7: - if ( (*p) == 84 ) - goto st8; - goto st0; -st8: - if ( ++p == pe ) - goto _test_eof8; -case 8: - if ( (*p) == 80 ) - goto st9; - goto st0; -st9: - if ( ++p == pe ) - goto _test_eof9; -case 9: - if ( (*p) == 47 ) - goto st10; - goto st0; -st10: - if ( ++p == pe ) - goto _test_eof10; -case 10: - if ( 48 <= (*p) && (*p) <= 57 ) - goto st11; - goto st0; -st11: - if ( ++p == pe ) - goto _test_eof11; -case 11: - if ( (*p) == 46 ) - goto st12; - if ( 48 <= (*p) && (*p) <= 57 ) - goto st11; - goto st0; -st12: - if ( ++p == pe ) - goto _test_eof12; -case 12: - if ( 48 <= (*p) && (*p) <= 57 ) - goto st13; - goto st0; -st13: - if ( ++p == pe ) - goto _test_eof13; -case 13: - if ( (*p) == 13 ) - goto tr17; - if ( 48 <= (*p) && (*p) <= 57 ) - goto st13; - goto st0; -tr17: -#line 94 "http11_parser.rl" - { - http_version(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st14; -tr25: -#line 73 "http11_parser.rl" - { MARK(mark, p); } -#line 74 "http11_parser.rl" - { - http_field(parser->data, PTR_TO(field_start), parser->field_len, PTR_TO(mark), LEN(mark, p)); - } - goto st14; -tr28: -#line 74 "http11_parser.rl" - { - http_field(parser->data, PTR_TO(field_start), parser->field_len, PTR_TO(mark), LEN(mark, p)); - } - goto st14; -st14: - if ( ++p == pe ) - goto _test_eof14; -case 14: -#line 337 "http11_parser.h" - if ( (*p) == 10 ) - goto st15; - goto st0; -st15: - if ( ++p == pe ) - goto _test_eof15; -case 15: - switch( (*p) ) { - case 13: goto st16; - case 33: goto tr20; - case 124: goto tr20; - case 126: goto tr20; - } - if ( (*p) < 45 ) { - if ( (*p) > 39 ) { - if ( 42 <= (*p) && (*p) <= 43 ) - goto tr20; - } else if ( (*p) >= 35 ) - goto tr20; - } else if ( (*p) > 46 ) { - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto tr20; - } else if ( (*p) > 90 ) { - if ( 94 <= (*p) && (*p) <= 122 ) - goto tr20; - } else - goto tr20; - } else - goto tr20; - goto st0; -st16: - if ( ++p == pe ) - goto _test_eof16; -case 16: - if ( (*p) == 10 ) - goto tr21; - goto st0; -tr21: -#line 102 "http11_parser.rl" - { - parser->body_start = p - buffer + 1; - header_done(parser->data, p + 1, pe - p - 1); - {p++; cs = 63; goto _out;} - } - goto st63; -st63: - if ( ++p == pe ) - goto _test_eof63; -case 63: -#line 388 "http11_parser.h" - goto st0; -tr20: -#line 66 "http11_parser.rl" - { MARK(field_start, p); } -#line 67 "http11_parser.rl" - { snake_upcase_char((char *)p); } - goto st17; -tr22: -#line 67 "http11_parser.rl" - { snake_upcase_char((char *)p); } - goto st17; -st17: - if ( ++p == pe ) - goto _test_eof17; -case 17: -#line 404 "http11_parser.h" - switch( (*p) ) { - case 33: goto tr22; - case 58: goto tr23; - case 124: goto tr22; - case 126: goto tr22; - } - if ( (*p) < 45 ) { - if ( (*p) > 39 ) { - if ( 42 <= (*p) && (*p) <= 43 ) - goto tr22; - } else if ( (*p) >= 35 ) - goto tr22; - } else if ( (*p) > 46 ) { - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto tr22; - } else if ( (*p) > 90 ) { - if ( 94 <= (*p) && (*p) <= 122 ) - goto tr22; - } else - goto tr22; - } else - goto tr22; - goto st0; -tr23: -#line 69 "http11_parser.rl" - { - parser->field_len = LEN(field_start, p); - } - goto st18; -tr26: -#line 73 "http11_parser.rl" - { MARK(mark, p); } - goto st18; -st18: - if ( ++p == pe ) - goto _test_eof18; -case 18: -#line 443 "http11_parser.h" - switch( (*p) ) { - case 13: goto tr25; - case 32: goto tr26; - } - goto tr24; -tr24: -#line 73 "http11_parser.rl" - { MARK(mark, p); } - goto st19; -st19: - if ( ++p == pe ) - goto _test_eof19; -case 19: -#line 457 "http11_parser.h" - if ( (*p) == 13 ) - goto tr28; - goto st19; -tr8: -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st20; -tr38: -#line 98 "http11_parser.rl" - { - request_path(parser->data, PTR_TO(mark), LEN(mark,p)); - } -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st20; -tr49: -#line 89 "http11_parser.rl" - {MARK(query_start, p); } -#line 90 "http11_parser.rl" - { - query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); - } -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st20; -tr53: -#line 90 "http11_parser.rl" - { - query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); - } -#line 82 "http11_parser.rl" - { - request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); - } - goto st20; -st20: - if ( ++p == pe ) - goto _test_eof20; -case 20: -#line 503 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr30; - case 35: goto st0; - case 37: goto tr31; - case 127: goto st0; - } - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto tr29; -tr29: -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st21; -st21: - if ( ++p == pe ) - goto _test_eof21; -case 21: -#line 521 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr33; - case 35: goto st0; - case 37: goto st22; - case 127: goto st0; - } - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto st21; -tr31: -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st22; -st22: - if ( ++p == pe ) - goto _test_eof22; -case 22: -#line 539 "http11_parser.h" - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st23; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st23; - } else - goto st23; - goto st0; -st23: - if ( ++p == pe ) - goto _test_eof23; -case 23: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st21; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st21; - } else - goto st21; - goto st0; -tr5: -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st24; -tr65: -#line 81 "http11_parser.rl" - { host(parser->data, PTR_TO(mark), LEN(mark, p)); } -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st24; -st24: - if ( ++p == pe ) - goto _test_eof24; -case 24: -#line 576 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr37; - case 35: goto tr38; - case 37: goto st25; - case 59: goto tr40; - case 63: goto tr41; - case 127: goto st0; - } - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto st24; -st25: - if ( ++p == pe ) - goto _test_eof25; -case 25: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st26; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st26; - } else - goto st26; - goto st0; -st26: - if ( ++p == pe ) - goto _test_eof26; -case 26: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st24; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st24; - } else - goto st24; - goto st0; -tr40: -#line 98 "http11_parser.rl" - { - request_path(parser->data, PTR_TO(mark), LEN(mark,p)); - } - goto st27; -st27: - if ( ++p == pe ) - goto _test_eof27; -case 27: -#line 624 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr7; - case 35: goto tr8; - case 37: goto st28; - case 63: goto st30; - case 127: goto st0; - } - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto st27; -st28: - if ( ++p == pe ) - goto _test_eof28; -case 28: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st29; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st29; - } else - goto st29; - goto st0; -st29: - if ( ++p == pe ) - goto _test_eof29; -case 29: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st27; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st27; - } else - goto st27; - goto st0; -tr41: -#line 98 "http11_parser.rl" - { - request_path(parser->data, PTR_TO(mark), LEN(mark,p)); - } - goto st30; -st30: - if ( ++p == pe ) - goto _test_eof30; -case 30: -#line 671 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr48; - case 35: goto tr49; - case 37: goto tr50; - case 127: goto st0; - } - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto tr47; -tr47: -#line 89 "http11_parser.rl" - {MARK(query_start, p); } - goto st31; -st31: - if ( ++p == pe ) - goto _test_eof31; -case 31: -#line 689 "http11_parser.h" - switch( (*p) ) { - case 32: goto tr52; - case 35: goto tr53; - case 37: goto st32; - case 127: goto st0; - } - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto st31; -tr50: -#line 89 "http11_parser.rl" - {MARK(query_start, p); } - goto st32; -st32: - if ( ++p == pe ) - goto _test_eof32; -case 32: -#line 707 "http11_parser.h" - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st33; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st33; - } else - goto st33; - goto st0; -st33: - if ( ++p == pe ) - goto _test_eof33; -case 33: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st31; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st31; - } else - goto st31; - goto st0; -tr6: -#line 64 "http11_parser.rl" - {MARK(mark, p); } -#line 68 "http11_parser.rl" - { downcase_char((char *)p); } - goto st34; -st34: - if ( ++p == pe ) - goto _test_eof34; -case 34: -#line 740 "http11_parser.h" - switch( (*p) ) { - case 84: goto tr56; - case 116: goto tr56; - } - goto st0; -tr56: -#line 68 "http11_parser.rl" - { downcase_char((char *)p); } - goto st35; -st35: - if ( ++p == pe ) - goto _test_eof35; -case 35: -#line 754 "http11_parser.h" - switch( (*p) ) { - case 84: goto tr57; - case 116: goto tr57; - } - goto st0; -tr57: -#line 68 "http11_parser.rl" - { downcase_char((char *)p); } - goto st36; -st36: - if ( ++p == pe ) - goto _test_eof36; -case 36: -#line 768 "http11_parser.h" - switch( (*p) ) { - case 80: goto tr58; - case 112: goto tr58; - } - goto st0; -tr58: -#line 68 "http11_parser.rl" - { downcase_char((char *)p); } - goto st37; -st37: - if ( ++p == pe ) - goto _test_eof37; -case 37: -#line 782 "http11_parser.h" - switch( (*p) ) { - case 58: goto tr59; - case 83: goto tr60; - case 115: goto tr60; - } - goto st0; -tr59: -#line 80 "http11_parser.rl" - { scheme(parser->data, PTR_TO(mark), LEN(mark, p)); } - goto st38; -st38: - if ( ++p == pe ) - goto _test_eof38; -case 38: -#line 797 "http11_parser.h" - if ( (*p) == 47 ) - goto st39; - goto st0; -st39: - if ( ++p == pe ) - goto _test_eof39; -case 39: - if ( (*p) == 47 ) - goto st40; - goto st0; -st40: - if ( ++p == pe ) - goto _test_eof40; -case 40: - if ( (*p) == 95 ) - goto tr63; - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto tr63; - } else if ( (*p) > 57 ) { - if ( (*p) > 90 ) { - if ( 97 <= (*p) && (*p) <= 122 ) - goto tr63; - } else if ( (*p) >= 65 ) - goto tr63; - } else - goto tr63; - goto st0; -tr63: -#line 64 "http11_parser.rl" - {MARK(mark, p); } - goto st41; -st41: - if ( ++p == pe ) - goto _test_eof41; -case 41: -#line 834 "http11_parser.h" - switch( (*p) ) { - case 47: goto tr65; - case 58: goto st42; - case 95: goto st41; - } - if ( (*p) < 65 ) { - if ( 45 <= (*p) && (*p) <= 57 ) - goto st41; - } else if ( (*p) > 90 ) { - if ( 97 <= (*p) && (*p) <= 122 ) - goto st41; - } else - goto st41; - goto st0; -st42: - if ( ++p == pe ) - goto _test_eof42; -case 42: - if ( (*p) == 47 ) - goto tr65; - if ( 48 <= (*p) && (*p) <= 57 ) - goto st42; - goto st0; -tr60: -#line 68 "http11_parser.rl" - { downcase_char((char *)p); } - goto st43; -st43: - if ( ++p == pe ) - goto _test_eof43; -case 43: -#line 866 "http11_parser.h" - if ( (*p) == 58 ) - goto tr59; - goto st0; -st44: - if ( ++p == pe ) - goto _test_eof44; -case 44: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st45; - case 95: goto st45; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st45; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st45; - } else - goto st45; - goto st0; -st45: - if ( ++p == pe ) - goto _test_eof45; -case 45: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st46; - case 95: goto st46; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st46; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st46; - } else - goto st46; - goto st0; -st46: - if ( ++p == pe ) - goto _test_eof46; -case 46: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st47; - case 95: goto st47; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st47; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st47; - } else - goto st47; - goto st0; -st47: - if ( ++p == pe ) - goto _test_eof47; -case 47: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st48; - case 95: goto st48; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st48; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st48; - } else - goto st48; - goto st0; -st48: - if ( ++p == pe ) - goto _test_eof48; -case 48: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st49; - case 95: goto st49; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st49; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st49; - } else - goto st49; - goto st0; -st49: - if ( ++p == pe ) - goto _test_eof49; -case 49: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st50; - case 95: goto st50; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st50; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st50; - } else - goto st50; - goto st0; -st50: - if ( ++p == pe ) - goto _test_eof50; -case 50: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st51; - case 95: goto st51; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st51; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st51; - } else - goto st51; - goto st0; -st51: - if ( ++p == pe ) - goto _test_eof51; -case 51: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st52; - case 95: goto st52; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st52; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st52; - } else - goto st52; - goto st0; -st52: - if ( ++p == pe ) - goto _test_eof52; -case 52: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st53; - case 95: goto st53; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st53; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st53; - } else - goto st53; - goto st0; -st53: - if ( ++p == pe ) - goto _test_eof53; -case 53: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st54; - case 95: goto st54; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st54; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st54; - } else - goto st54; - goto st0; -st54: - if ( ++p == pe ) - goto _test_eof54; -case 54: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st55; - case 95: goto st55; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st55; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st55; - } else - goto st55; - goto st0; -st55: - if ( ++p == pe ) - goto _test_eof55; -case 55: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st56; - case 95: goto st56; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st56; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st56; - } else - goto st56; - goto st0; -st56: - if ( ++p == pe ) - goto _test_eof56; -case 56: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st57; - case 95: goto st57; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st57; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st57; - } else - goto st57; - goto st0; -st57: - if ( ++p == pe ) - goto _test_eof57; -case 57: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st58; - case 95: goto st58; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st58; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st58; - } else - goto st58; - goto st0; -st58: - if ( ++p == pe ) - goto _test_eof58; -case 58: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st59; - case 95: goto st59; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st59; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st59; - } else - goto st59; - goto st0; -st59: - if ( ++p == pe ) - goto _test_eof59; -case 59: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st60; - case 95: goto st60; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st60; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st60; - } else - goto st60; - goto st0; -st60: - if ( ++p == pe ) - goto _test_eof60; -case 60: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st61; - case 95: goto st61; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st61; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st61; - } else - goto st61; - goto st0; -st61: - if ( ++p == pe ) - goto _test_eof61; -case 61: - switch( (*p) ) { - case 32: goto tr2; - case 36: goto st62; - case 95: goto st62; - } - if ( (*p) < 48 ) { - if ( 45 <= (*p) && (*p) <= 46 ) - goto st62; - } else if ( (*p) > 57 ) { - if ( 65 <= (*p) && (*p) <= 90 ) - goto st62; - } else - goto st62; - goto st0; -st62: - if ( ++p == pe ) - goto _test_eof62; -case 62: - if ( (*p) == 32 ) - goto tr2; - goto st0; - } - _test_eof2: cs = 2; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof6: cs = 6; goto _test_eof; - _test_eof7: cs = 7; goto _test_eof; - _test_eof8: cs = 8; goto _test_eof; - _test_eof9: cs = 9; goto _test_eof; - _test_eof10: cs = 10; goto _test_eof; - _test_eof11: cs = 11; goto _test_eof; - _test_eof12: cs = 12; goto _test_eof; - _test_eof13: cs = 13; goto _test_eof; - _test_eof14: cs = 14; goto _test_eof; - _test_eof15: cs = 15; goto _test_eof; - _test_eof16: cs = 16; goto _test_eof; - _test_eof63: cs = 63; goto _test_eof; - _test_eof17: cs = 17; goto _test_eof; - _test_eof18: cs = 18; goto _test_eof; - _test_eof19: cs = 19; goto _test_eof; - _test_eof20: cs = 20; goto _test_eof; - _test_eof21: cs = 21; goto _test_eof; - _test_eof22: cs = 22; goto _test_eof; - _test_eof23: cs = 23; goto _test_eof; - _test_eof24: cs = 24; goto _test_eof; - _test_eof25: cs = 25; goto _test_eof; - _test_eof26: cs = 26; goto _test_eof; - _test_eof27: cs = 27; goto _test_eof; - _test_eof28: cs = 28; goto _test_eof; - _test_eof29: cs = 29; goto _test_eof; - _test_eof30: cs = 30; goto _test_eof; - _test_eof31: cs = 31; goto _test_eof; - _test_eof32: cs = 32; goto _test_eof; - _test_eof33: cs = 33; goto _test_eof; - _test_eof34: cs = 34; goto _test_eof; - _test_eof35: cs = 35; goto _test_eof; - _test_eof36: cs = 36; goto _test_eof; - _test_eof37: cs = 37; goto _test_eof; - _test_eof38: cs = 38; goto _test_eof; - _test_eof39: cs = 39; goto _test_eof; - _test_eof40: cs = 40; goto _test_eof; - _test_eof41: cs = 41; goto _test_eof; - _test_eof42: cs = 42; goto _test_eof; - _test_eof43: cs = 43; goto _test_eof; - _test_eof44: cs = 44; goto _test_eof; - _test_eof45: cs = 45; goto _test_eof; - _test_eof46: cs = 46; goto _test_eof; - _test_eof47: cs = 47; goto _test_eof; - _test_eof48: cs = 48; goto _test_eof; - _test_eof49: cs = 49; goto _test_eof; - _test_eof50: cs = 50; goto _test_eof; - _test_eof51: cs = 51; goto _test_eof; - _test_eof52: cs = 52; goto _test_eof; - _test_eof53: cs = 53; goto _test_eof; - _test_eof54: cs = 54; goto _test_eof; - _test_eof55: cs = 55; goto _test_eof; - _test_eof56: cs = 56; goto _test_eof; - _test_eof57: cs = 57; goto _test_eof; - _test_eof58: cs = 58; goto _test_eof; - _test_eof59: cs = 59; goto _test_eof; - _test_eof60: cs = 60; goto _test_eof; - _test_eof61: cs = 61; goto _test_eof; - _test_eof62: cs = 62; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 138 "http11_parser.rl" - - if (!http_parser_has_error(parser)) - parser->cs = cs; - parser->nread += p - (buffer + off); - - assert(p <= pe && "buffer overflow after parsing execute"); - assert(parser->nread <= len && "nread longer than length"); - assert(parser->body_start <= len && "body starts after buffer end"); - assert(parser->mark < len && "mark is after buffer end"); - assert(parser->field_len <= len && "field has length longer than whole buffer"); - assert(parser->field_start < len && "field starts after buffer end"); -} - -static int http_parser_has_error(http_parser *parser) { - return parser->cs == http_parser_error; -} - -static int http_parser_is_finished(http_parser *parser) { - return parser->cs == http_parser_first_final; -} -#endif /* http11_parser_h */ diff --git a/ext/unicorn/http11/http11_parser.rl b/ext/unicorn/http11/http11_parser.rl deleted file mode 100644 index 9894276..0000000 --- a/ext/unicorn/http11/http11_parser.rl +++ /dev/null @@ -1,158 +0,0 @@ -/** - * Copyright (c) 2005 Zed A. Shaw - * You can redistribute it and/or modify it under the same terms as Ruby. - */ -#ifndef http11_parser_h -#define http11_parser_h - -#include - -static void http_field(void *data, const char *field, - size_t flen, const char *value, size_t vlen); -static void request_method(void *data, const char *at, size_t length); -static void scheme(void *data, const char *at, size_t length); -static void host(void *data, const char *at, size_t length); -static void request_uri(void *data, const char *at, size_t length); -static void fragment(void *data, const char *at, size_t length); -static void request_path(void *data, const char *at, size_t length); -static void query_string(void *data, const char *at, size_t length); -static void http_version(void *data, const char *at, size_t length); -static void header_done(void *data, const char *at, size_t length); - -typedef struct http_parser { - int cs; - size_t body_start; - size_t nread; - size_t mark; - size_t field_start; - size_t field_len; - size_t query_start; - - void *data; -} http_parser; - -static int http_parser_has_error(http_parser *parser); -static int http_parser_is_finished(http_parser *parser); - -/* - * capitalizes all lower-case ASCII characters, - * converts dashes to underscores. - */ -static void snake_upcase_char(char *c) -{ - if (*c >= 'a' && *c <= 'z') - *c &= ~0x20; - else if (*c == '-') - *c = '_'; -} - -static void downcase_char(char *c) -{ - if (*c >= 'A' && *c <= 'Z') - *c |= 0x20; -} - -#define LEN(AT, FPC) (FPC - buffer - parser->AT) -#define MARK(M,FPC) (parser->M = (FPC) - buffer) -#define PTR_TO(F) (buffer + parser->F) - -/** Machine **/ - -%%{ - machine http_parser; - - action mark {MARK(mark, fpc); } - - action start_field { MARK(field_start, fpc); } - action snake_upcase_field { snake_upcase_char((char *)fpc); } - action downcase_char { downcase_char((char *)fpc); } - action write_field { - parser->field_len = LEN(field_start, fpc); - } - - action start_value { MARK(mark, fpc); } - action write_value { - http_field(parser->data, PTR_TO(field_start), parser->field_len, PTR_TO(mark), LEN(mark, fpc)); - } - action request_method { - request_method(parser->data, PTR_TO(mark), LEN(mark, fpc)); - } - action scheme { scheme(parser->data, PTR_TO(mark), LEN(mark, fpc)); } - action host { host(parser->data, PTR_TO(mark), LEN(mark, fpc)); } - action request_uri { - request_uri(parser->data, PTR_TO(mark), LEN(mark, fpc)); - } - action fragment { - fragment(parser->data, PTR_TO(mark), LEN(mark, fpc)); - } - - action start_query {MARK(query_start, fpc); } - action query_string { - query_string(parser->data, PTR_TO(query_start), LEN(query_start, fpc)); - } - - action http_version { - http_version(parser->data, PTR_TO(mark), LEN(mark, fpc)); - } - - action request_path { - request_path(parser->data, PTR_TO(mark), LEN(mark,fpc)); - } - - action done { - parser->body_start = fpc - buffer + 1; - header_done(parser->data, fpc + 1, pe - fpc - 1); - fbreak; - } - - include http_parser_common "http11_parser_common.rl"; -}%% - -/** Data **/ -%% write data; - -static void http_parser_init(http_parser *parser) { - int cs = 0; - memset(parser, 0, sizeof(*parser)); - %% write init; - parser->cs = cs; -} - -/** exec **/ -static void http_parser_execute( - http_parser *parser, const char *buffer, size_t len) -{ - const char *p, *pe; - int cs = parser->cs; - size_t off = parser->nread; - - assert(off <= len && "offset past end of buffer"); - - p = buffer+off; - pe = buffer+len; - - assert(*pe == '\0' && "pointer does not end on NUL"); - assert(pe - p == len - off && "pointers aren't same distance"); - - %% write exec; - - if (!http_parser_has_error(parser)) - parser->cs = cs; - parser->nread += p - (buffer + off); - - assert(p <= pe && "buffer overflow after parsing execute"); - assert(parser->nread <= len && "nread longer than length"); - assert(parser->body_start <= len && "body starts after buffer end"); - assert(parser->mark < len && "mark is after buffer end"); - assert(parser->field_len <= len && "field has length longer than whole buffer"); - assert(parser->field_start < len && "field starts after buffer end"); -} - -static int http_parser_has_error(http_parser *parser) { - return parser->cs == http_parser_error; -} - -static int http_parser_is_finished(http_parser *parser) { - return parser->cs == http_parser_first_final; -} -#endif /* http11_parser_h */ diff --git a/ext/unicorn/http11/http11_parser_common.rl b/ext/unicorn/http11/http11_parser_common.rl deleted file mode 100644 index ae01a55..0000000 --- a/ext/unicorn/http11/http11_parser_common.rl +++ /dev/null @@ -1,58 +0,0 @@ -%%{ - - machine http_parser_common; - -#### HTTP PROTOCOL GRAMMAR -# line endings - CRLF = "\r\n"; - -# character types - CTL = (cntrl | 127); - safe = ("$" | "-" | "_" | "."); - extra = ("!" | "*" | "'" | "(" | ")" | ","); - reserved = (";" | "/" | "?" | ":" | "@" | "&" | "=" | "+"); - sorta_safe = ("\"" | "<" | ">"); - unsafe = (CTL | " " | "#" | "%" | sorta_safe); - national = any -- (alpha | digit | reserved | extra | safe | unsafe); - unreserved = (alpha | digit | safe | extra | national); - escape = ("%" xdigit xdigit); - uchar = (unreserved | escape | sorta_safe); - pchar = (uchar | ":" | "@" | "&" | "=" | "+"); - tspecials = ("(" | ")" | "<" | ">" | "@" | "," | ";" | ":" | "\\" | "\"" | "/" | "[" | "]" | "?" | "=" | "{" | "}" | " " | "\t"); - -# elements - token = (ascii -- (CTL | tspecials)); - -# URI schemes and absolute paths - scheme = ( "http"i ("s"i)? ) $downcase_char >mark %scheme; - hostname = (alnum | "-" | "." | "_")+; - host_with_port = (hostname (":" digit*)?) >mark %host; - - path = ( pchar+ ( "/" pchar* )* ) ; - query = ( uchar | reserved )* %query_string ; - param = ( pchar | "/" )* ; - params = ( param ( ";" param )* ) ; - rel_path = ( path? %request_path (";" params)? ) ("?" %start_query query)?; - absolute_path = ( "/"+ rel_path ); - path_uri = absolute_path > mark %request_uri; - Absolute_URI = (scheme "://" host_with_port path_uri); - - Request_URI = ((absolute_path | "*") >mark %request_uri) | Absolute_URI; - Fragment = ( uchar | reserved )* >mark %fragment; - Method = ( upper | digit | safe ){1,20} >mark %request_method; - - http_number = ( digit+ "." digit+ ) ; - HTTP_Version = ( "HTTP/" http_number ) >mark %http_version ; - Request_Line = ( Method " " Request_URI ("#" Fragment){0,1} " " HTTP_Version CRLF ) ; - - field_name = ( token -- ":" )+ >start_field $snake_upcase_field %write_field; - - field_value = any* >start_value %write_value; - - message_header = field_name ":" " "* field_value :> CRLF; - - Request = Request_Line ( message_header )* ( CRLF @done ); - -main := Request; - -}%% diff --git a/ext/unicorn_http/ext_help.h b/ext/unicorn_http/ext_help.h new file mode 100644 index 0000000..17f7b01 --- /dev/null +++ b/ext/unicorn_http/ext_help.h @@ -0,0 +1,12 @@ +#ifndef ext_help_h +#define ext_help_h + +#define ARRAY_SIZE(x) (sizeof(x)/sizeof(x[0])) + +#ifdef DEBUG +#define TRACE() fprintf(stderr, "> %s:%d:%s\n", __FILE__, __LINE__, __FUNCTION__) +#else +#define TRACE() +#endif + +#endif diff --git a/ext/unicorn_http/extconf.rb b/ext/unicorn_http/extconf.rb new file mode 100644 index 0000000..29d9334 --- /dev/null +++ b/ext/unicorn_http/extconf.rb @@ -0,0 +1,5 @@ +require 'mkmf' + +dir_config("unicorn_http") +have_library("c", "main") +create_makefile("unicorn_http") diff --git a/ext/unicorn_http/unicorn_http.c b/ext/unicorn_http/unicorn_http.c new file mode 100644 index 0000000..41936cd --- /dev/null +++ b/ext/unicorn_http/unicorn_http.c @@ -0,0 +1,457 @@ +/** + * Copyright (c) 2009 Eric Wong (all bugs are Eric's fault) + * Copyright (c) 2005 Zed A. Shaw + * You can redistribute it and/or modify it under the same terms as Ruby. + */ +#include "ruby.h" +#include "ext_help.h" +#include +#include +#include "unicorn_http.h" + +static http_parser *data_get(VALUE self) +{ + http_parser *http; + + Data_Get_Struct(self, http_parser, http); + if (!http) + rb_raise(rb_eArgError, "NULL found for http when shouldn't be."); + return http; +} + +#ifndef RSTRING_PTR +#define RSTRING_PTR(s) (RSTRING(s)->ptr) +#endif +#ifndef RSTRING_LEN +#define RSTRING_LEN(s) (RSTRING(s)->len) +#endif + +static VALUE mUnicorn; +static VALUE cHttpParser; +static VALUE eHttpParserError; +static VALUE sym_http_body; + +#define HTTP_PREFIX "HTTP_" +#define HTTP_PREFIX_LEN (sizeof(HTTP_PREFIX) - 1) + +static VALUE global_rack_url_scheme; +static VALUE global_request_method; +static VALUE global_request_uri; +static VALUE global_fragment; +static VALUE global_query_string; +static VALUE global_http_version; +static VALUE global_request_path; +static VALUE global_path_info; +static VALUE global_server_name; +static VALUE global_server_port; +static VALUE global_server_protocol; +static VALUE global_server_protocol_value; +static VALUE global_http_host; +static VALUE global_http_x_forwarded_proto; +static VALUE global_port_80; +static VALUE global_port_443; +static VALUE global_localhost; +static VALUE global_http; + +/** Defines common length and error messages for input length validation. */ +#define DEF_MAX_LENGTH(N, length) \ + static const size_t MAX_##N##_LENGTH = length; \ + static const char * const MAX_##N##_LENGTH_ERR = \ + "HTTP element " # N " is longer than the " # length " allowed length." + +/** + * Validates the max length of given input and throws an HttpParserError + * exception if over. + */ +#define VALIDATE_MAX_LENGTH(len, N) do { \ + if (len > MAX_##N##_LENGTH) \ + rb_raise(eHttpParserError, MAX_##N##_LENGTH_ERR); \ +} while (0) + +/** Defines global strings in the init method. */ +#define DEF_GLOBAL(N, val) do { \ + global_##N = rb_obj_freeze(rb_str_new(val, sizeof(val) - 1)); \ + rb_global_variable(&global_##N); \ +} while (0) + +/* Defines the maximum allowed lengths for various input elements.*/ +DEF_MAX_LENGTH(FIELD_NAME, 256); +DEF_MAX_LENGTH(FIELD_VALUE, 80 * 1024); +DEF_MAX_LENGTH(REQUEST_URI, 1024 * 12); +DEF_MAX_LENGTH(FRAGMENT, 1024); /* Don't know if this length is specified somewhere or not */ +DEF_MAX_LENGTH(REQUEST_PATH, 1024); +DEF_MAX_LENGTH(QUERY_STRING, (1024 * 10)); +DEF_MAX_LENGTH(HEADER, (1024 * (80 + 32))); + +struct common_field { + const signed long len; + const char *name; + VALUE value; +}; + +/* + * A list of common HTTP headers we expect to receive. + * This allows us to avoid repeatedly creating identical string + * objects to be used with rb_hash_aset(). + */ +static struct common_field common_http_fields[] = { +# define f(N) { (sizeof(N) - 1), N, Qnil } + f("ACCEPT"), + f("ACCEPT_CHARSET"), + f("ACCEPT_ENCODING"), + f("ACCEPT_LANGUAGE"), + f("ALLOW"), + f("AUTHORIZATION"), + f("CACHE_CONTROL"), + f("CONNECTION"), + f("CONTENT_ENCODING"), + f("CONTENT_LENGTH"), + f("CONTENT_TYPE"), + f("COOKIE"), + f("DATE"), + f("EXPECT"), + f("FROM"), + f("HOST"), + f("IF_MATCH"), + f("IF_MODIFIED_SINCE"), + f("IF_NONE_MATCH"), + f("IF_RANGE"), + f("IF_UNMODIFIED_SINCE"), + f("KEEP_ALIVE"), /* Firefox sends this */ + f("MAX_FORWARDS"), + f("PRAGMA"), + f("PROXY_AUTHORIZATION"), + f("RANGE"), + f("REFERER"), + f("TE"), + f("TRAILER"), + f("TRANSFER_ENCODING"), + f("UPGRADE"), + f("USER_AGENT"), + f("VIA"), + f("X_FORWARDED_FOR"), /* common for proxies */ + f("X_FORWARDED_PROTO"), /* common for proxies */ + f("X_REAL_IP"), /* common for proxies */ + f("WARNING") +# undef f +}; + +/* this function is not performance-critical */ +static void init_common_fields(void) +{ + int i; + struct common_field *cf = common_http_fields; + char tmp[256]; /* MAX_FIELD_NAME_LENGTH */ + memcpy(tmp, HTTP_PREFIX, HTTP_PREFIX_LEN); + + for(i = 0; i < ARRAY_SIZE(common_http_fields); cf++, i++) { + /* Rack doesn't like certain headers prefixed with "HTTP_" */ + if (!strcmp("CONTENT_LENGTH", cf->name) || + !strcmp("CONTENT_TYPE", cf->name)) { + cf->value = rb_str_new(cf->name, cf->len); + } else { + memcpy(tmp + HTTP_PREFIX_LEN, cf->name, cf->len + 1); + cf->value = rb_str_new(tmp, HTTP_PREFIX_LEN + cf->len); + } + cf->value = rb_obj_freeze(cf->value); + rb_global_variable(&cf->value); + } +} + +static VALUE find_common_field_value(const char *field, size_t flen) +{ + int i; + struct common_field *cf = common_http_fields; + for(i = 0; i < ARRAY_SIZE(common_http_fields); i++, cf++) { + if (cf->len == flen && !memcmp(cf->name, field, flen)) + return cf->value; + } + return Qnil; +} + +static void http_field(void *data, const char *field, + size_t flen, const char *value, size_t vlen) +{ + VALUE req = (VALUE)data; + VALUE f = Qnil; + + VALIDATE_MAX_LENGTH(flen, FIELD_NAME); + VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE); + + f = find_common_field_value(field, flen); + + if (f == Qnil) { + /* + * We got a strange header that we don't have a memoized value for. + * Fallback to creating a new string to use as a hash key. + * + * using rb_str_new(NULL, len) here is faster than rb_str_buf_new(len) + * in my testing, because: there's no minimum allocation length (and + * no check for it, either), RSTRING_LEN(f) does not need to be + * written twice, and and RSTRING_PTR(f) will already be + * null-terminated for us. + */ + f = rb_str_new(NULL, HTTP_PREFIX_LEN + flen); + memcpy(RSTRING_PTR(f), HTTP_PREFIX, HTTP_PREFIX_LEN); + memcpy(RSTRING_PTR(f) + HTTP_PREFIX_LEN, field, flen); + assert(*(RSTRING_PTR(f) + RSTRING_LEN(f)) == '\0'); /* paranoia */ + /* fprintf(stderr, "UNKNOWN HEADER <%s>\n", RSTRING_PTR(f)); */ + } else if (f == global_http_host && rb_hash_aref(req, f) != Qnil) { + return; + } + + rb_hash_aset(req, f, rb_str_new(value, vlen)); +} + +static void request_method(void *data, const char *at, size_t length) +{ + VALUE req = (VALUE)data; + VALUE val = Qnil; + + val = rb_str_new(at, length); + rb_hash_aset(req, global_request_method, val); +} + +static void scheme(void *data, const char *at, size_t length) +{ + rb_hash_aset((VALUE)data, global_rack_url_scheme, rb_str_new(at, length)); +} + +static void host(void *data, const char *at, size_t length) +{ + rb_hash_aset((VALUE)data, global_http_host, rb_str_new(at, length)); +} + +static void request_uri(void *data, const char *at, size_t length) +{ + VALUE req = (VALUE)data; + VALUE val = Qnil; + + VALIDATE_MAX_LENGTH(length, REQUEST_URI); + + val = rb_str_new(at, length); + rb_hash_aset(req, global_request_uri, val); + + /* "OPTIONS * HTTP/1.1\r\n" is a valid request */ + if (length == 1 && *at == '*') { + val = rb_str_new(NULL, 0); + rb_hash_aset(req, global_request_path, val); + rb_hash_aset(req, global_path_info, val); + } +} + +static void fragment(void *data, const char *at, size_t length) +{ + VALUE req = (VALUE)data; + VALUE val = Qnil; + + VALIDATE_MAX_LENGTH(length, FRAGMENT); + + val = rb_str_new(at, length); + rb_hash_aset(req, global_fragment, val); +} + +static void request_path(void *data, const char *at, size_t length) +{ + VALUE req = (VALUE)data; + VALUE val = Qnil; + + VALIDATE_MAX_LENGTH(length, REQUEST_PATH); + + val = rb_str_new(at, length); + rb_hash_aset(req, global_request_path, val); + + /* rack says PATH_INFO must start with "/" or be empty */ + if (!(length == 1 && *at == '*')) + rb_hash_aset(req, global_path_info, val); +} + +static void query_string(void *data, const char *at, size_t length) +{ + VALUE req = (VALUE)data; + VALUE val = Qnil; + + VALIDATE_MAX_LENGTH(length, QUERY_STRING); + + val = rb_str_new(at, length); + rb_hash_aset(req, global_query_string, val); +} + +static void http_version(void *data, const char *at, size_t length) +{ + VALUE req = (VALUE)data; + VALUE val = rb_str_new(at, length); + rb_hash_aset(req, global_http_version, val); +} + +/** Finalizes the request header to have a bunch of stuff that's needed. */ +static void header_done(void *data, const char *at, size_t length) +{ + VALUE req = (VALUE)data; + VALUE server_name = global_localhost; + VALUE server_port = global_port_80; + VALUE temp; + + /* rack requires QUERY_STRING */ + if (rb_hash_aref(req, global_query_string) == Qnil) + rb_hash_aset(req, global_query_string, rb_str_new(NULL, 0)); + + /* set rack.url_scheme to "https" or "http", no others are allowed by Rack */ + if ((temp = rb_hash_aref(req, global_rack_url_scheme)) == Qnil) { + if ((temp = rb_hash_aref(req, global_http_x_forwarded_proto)) != Qnil && + RSTRING_LEN(temp) == 5 && + !memcmp("https", RSTRING_PTR(temp), 5)) + server_port = global_port_443; + else + temp = global_http; + rb_hash_aset(req, global_rack_url_scheme, temp); + } else if (RSTRING_LEN(temp) == 5 && !memcmp("https", RSTRING_PTR(temp), 5)) { + server_port = global_port_443; + } + + /* parse and set the SERVER_NAME and SERVER_PORT variables */ + if ((temp = rb_hash_aref(req, global_http_host)) != Qnil) { + char *colon = memchr(RSTRING_PTR(temp), ':', RSTRING_LEN(temp)); + if (colon) { + long port_start = colon - RSTRING_PTR(temp) + 1; + + server_name = rb_str_substr(temp, 0, colon - RSTRING_PTR(temp)); + if ((RSTRING_LEN(temp) - port_start) > 0) + server_port = rb_str_substr(temp, port_start, RSTRING_LEN(temp)); + } else { + server_name = temp; + } + } + rb_hash_aset(req, global_server_name, server_name); + rb_hash_aset(req, global_server_port, server_port); + rb_hash_aset(req, global_server_protocol, global_server_protocol_value); + + /* grab the initial body and stuff it into the hash */ + temp = rb_hash_aref(req, global_request_method); + if (temp != Qnil) { + long len = RSTRING_LEN(temp); + char *ptr = RSTRING_PTR(temp); + + if (memcmp(ptr, "HEAD", len) && memcmp(ptr, "GET", len)) + rb_hash_aset(req, sym_http_body, rb_str_new(at, length)); + } +} + +static void HttpParser_free(void *data) { + TRACE(); + + if(data) { + free(data); + } +} + + +static VALUE HttpParser_alloc(VALUE klass) +{ + VALUE obj; + http_parser *hp = ALLOC_N(http_parser, 1); + TRACE(); + http_parser_init(hp); + + obj = Data_Wrap_Struct(klass, NULL, HttpParser_free, hp); + + return obj; +} + + +/** + * call-seq: + * parser.new -> parser + * + * Creates a new parser. + */ +static VALUE HttpParser_init(VALUE self) +{ + http_parser_init(data_get(self)); + + return self; +} + + +/** + * call-seq: + * parser.reset -> nil + * + * Resets the parser to it's initial state so that you can reuse it + * rather than making new ones. + */ +static VALUE HttpParser_reset(VALUE self) +{ + http_parser_init(data_get(self)); + + return Qnil; +} + + +/** + * call-seq: + * parser.execute(req_hash, data) -> true/false + * + * Takes a Hash and a String of data, parses the String of data filling + * in the Hash returning a boolean to indicate whether or not parsing + * is finished. + * + * This function now throws an exception when there is a parsing error. + * This makes the logic for working with the parser much easier. You + * will need to wrap the parser with an exception handling block. + */ + +static VALUE HttpParser_execute(VALUE self, VALUE req_hash, VALUE data) +{ + http_parser *http = data_get(self); + char *dptr = RSTRING_PTR(data); + long dlen = RSTRING_LEN(data); + + if (http->nread < dlen) { + http->data = (void *)req_hash; + http_parser_execute(http, dptr, dlen); + + VALIDATE_MAX_LENGTH(http->nread, HEADER); + + if (!http_parser_has_error(http)) + return http_parser_is_finished(http) ? Qtrue : Qfalse; + + rb_raise(eHttpParserError, "Invalid HTTP format, parsing fails."); + } + rb_raise(eHttpParserError, "Requested start is after data buffer end."); +} + +void Init_unicorn_http(void) +{ + mUnicorn = rb_define_module("Unicorn"); + + DEF_GLOBAL(rack_url_scheme, "rack.url_scheme"); + DEF_GLOBAL(request_method, "REQUEST_METHOD"); + DEF_GLOBAL(request_uri, "REQUEST_URI"); + DEF_GLOBAL(fragment, "FRAGMENT"); + DEF_GLOBAL(query_string, "QUERY_STRING"); + DEF_GLOBAL(http_version, "HTTP_VERSION"); + DEF_GLOBAL(request_path, "REQUEST_PATH"); + DEF_GLOBAL(path_info, "PATH_INFO"); + DEF_GLOBAL(server_name, "SERVER_NAME"); + DEF_GLOBAL(server_port, "SERVER_PORT"); + DEF_GLOBAL(server_protocol, "SERVER_PROTOCOL"); + DEF_GLOBAL(server_protocol_value, "HTTP/1.1"); + DEF_GLOBAL(http_x_forwarded_proto, "HTTP_X_FORWARDED_PROTO"); + DEF_GLOBAL(port_80, "80"); + DEF_GLOBAL(port_443, "443"); + DEF_GLOBAL(localhost, "localhost"); + DEF_GLOBAL(http, "http"); + + eHttpParserError = rb_define_class_under(mUnicorn, "HttpParserError", rb_eIOError); + + cHttpParser = rb_define_class_under(mUnicorn, "HttpParser", rb_cObject); + rb_define_alloc_func(cHttpParser, HttpParser_alloc); + rb_define_method(cHttpParser, "initialize", HttpParser_init,0); + rb_define_method(cHttpParser, "reset", HttpParser_reset,0); + rb_define_method(cHttpParser, "execute", HttpParser_execute,2); + sym_http_body = ID2SYM(rb_intern("http_body")); + init_common_fields(); + global_http_host = find_common_field_value("HOST", 4); + assert(global_http_host != Qnil); +} diff --git a/ext/unicorn_http/unicorn_http.h b/ext/unicorn_http/unicorn_http.h new file mode 100644 index 0000000..d39a22b --- /dev/null +++ b/ext/unicorn_http/unicorn_http.h @@ -0,0 +1,1289 @@ + +#line 1 "unicorn_http.rl" +/** + * Copyright (c) 2005 Zed A. Shaw + * You can redistribute it and/or modify it under the same terms as Ruby. + */ +#ifndef unicorn_http_h +#define unicorn_http_h + +#include + +static void http_field(void *data, const char *field, + size_t flen, const char *value, size_t vlen); +static void request_method(void *data, const char *at, size_t length); +static void scheme(void *data, const char *at, size_t length); +static void host(void *data, const char *at, size_t length); +static void request_uri(void *data, const char *at, size_t length); +static void fragment(void *data, const char *at, size_t length); +static void request_path(void *data, const char *at, size_t length); +static void query_string(void *data, const char *at, size_t length); +static void http_version(void *data, const char *at, size_t length); +static void header_done(void *data, const char *at, size_t length); + +typedef struct http_parser { + int cs; + size_t body_start; + size_t nread; + size_t mark; + size_t field_start; + size_t field_len; + size_t query_start; + + void *data; +} http_parser; + +static int http_parser_has_error(http_parser *parser); +static int http_parser_is_finished(http_parser *parser); + +/* + * capitalizes all lower-case ASCII characters, + * converts dashes to underscores. + */ +static void snake_upcase_char(char *c) +{ + if (*c >= 'a' && *c <= 'z') + *c &= ~0x20; + else if (*c == '-') + *c = '_'; +} + +static void downcase_char(char *c) +{ + if (*c >= 'A' && *c <= 'Z') + *c |= 0x20; +} + +#define LEN(AT, FPC) (FPC - buffer - parser->AT) +#define MARK(M,FPC) (parser->M = (FPC) - buffer) +#define PTR_TO(F) (buffer + parser->F) + +/** Machine **/ + + +#line 109 "unicorn_http.rl" + + +/** Data **/ + +#line 70 "unicorn_http.h" +static const int http_parser_start = 1; +static const int http_parser_first_final = 63; +static const int http_parser_error = 0; + +static const int http_parser_en_main = 1; + + +#line 113 "unicorn_http.rl" + +static void http_parser_init(http_parser *parser) { + int cs = 0; + memset(parser, 0, sizeof(*parser)); + +#line 84 "unicorn_http.h" + { + cs = http_parser_start; + } + +#line 118 "unicorn_http.rl" + parser->cs = cs; +} + +/** exec **/ +static void http_parser_execute( + http_parser *parser, const char *buffer, size_t len) +{ + const char *p, *pe; + int cs = parser->cs; + size_t off = parser->nread; + + assert(off <= len && "offset past end of buffer"); + + p = buffer+off; + pe = buffer+len; + + assert(*pe == '\0' && "pointer does not end on NUL"); + assert(pe - p == len - off && "pointers aren't same distance"); + + +#line 110 "unicorn_http.h" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +case 1: + switch( (*p) ) { + case 36: goto tr0; + case 95: goto tr0; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto tr0; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto tr0; + } else + goto tr0; + goto st0; +st0: +cs = 0; + goto _out; +tr0: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st2; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: +#line 141 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st44; + case 95: goto st44; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st44; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st44; + } else + goto st44; + goto st0; +tr2: +#line 77 "unicorn_http.rl" + { + request_method(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st3; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: +#line 166 "unicorn_http.h" + switch( (*p) ) { + case 42: goto tr4; + case 47: goto tr5; + case 72: goto tr6; + case 104: goto tr6; + } + goto st0; +tr4: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st4; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: +#line 182 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr7; + case 35: goto tr8; + } + goto st0; +tr7: +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st5; +tr30: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } +#line 85 "unicorn_http.rl" + { + fragment(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st5; +tr33: +#line 85 "unicorn_http.rl" + { + fragment(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st5; +tr37: +#line 98 "unicorn_http.rl" + { + request_path(parser->data, PTR_TO(mark), LEN(mark,p)); + } +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st5; +tr48: +#line 89 "unicorn_http.rl" + {MARK(query_start, p); } +#line 90 "unicorn_http.rl" + { + query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); + } +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st5; +tr52: +#line 90 "unicorn_http.rl" + { + query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); + } +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st5; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: +#line 244 "unicorn_http.h" + if ( (*p) == 72 ) + goto tr9; + goto st0; +tr9: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st6; +st6: + if ( ++p == pe ) + goto _test_eof6; +case 6: +#line 256 "unicorn_http.h" + if ( (*p) == 84 ) + goto st7; + goto st0; +st7: + if ( ++p == pe ) + goto _test_eof7; +case 7: + if ( (*p) == 84 ) + goto st8; + goto st0; +st8: + if ( ++p == pe ) + goto _test_eof8; +case 8: + if ( (*p) == 80 ) + goto st9; + goto st0; +st9: + if ( ++p == pe ) + goto _test_eof9; +case 9: + if ( (*p) == 47 ) + goto st10; + goto st0; +st10: + if ( ++p == pe ) + goto _test_eof10; +case 10: + if ( 48 <= (*p) && (*p) <= 57 ) + goto st11; + goto st0; +st11: + if ( ++p == pe ) + goto _test_eof11; +case 11: + if ( (*p) == 46 ) + goto st12; + if ( 48 <= (*p) && (*p) <= 57 ) + goto st11; + goto st0; +st12: + if ( ++p == pe ) + goto _test_eof12; +case 12: + if ( 48 <= (*p) && (*p) <= 57 ) + goto st13; + goto st0; +st13: + if ( ++p == pe ) + goto _test_eof13; +case 13: + if ( (*p) == 13 ) + goto tr17; + if ( 48 <= (*p) && (*p) <= 57 ) + goto st13; + goto st0; +tr17: +#line 94 "unicorn_http.rl" + { + http_version(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st14; +tr25: +#line 73 "unicorn_http.rl" + { MARK(mark, p); } +#line 74 "unicorn_http.rl" + { + http_field(parser->data, PTR_TO(field_start), parser->field_len, PTR_TO(mark), LEN(mark, p)); + } + goto st14; +tr28: +#line 74 "unicorn_http.rl" + { + http_field(parser->data, PTR_TO(field_start), parser->field_len, PTR_TO(mark), LEN(mark, p)); + } + goto st14; +st14: + if ( ++p == pe ) + goto _test_eof14; +case 14: +#line 337 "unicorn_http.h" + if ( (*p) == 10 ) + goto st15; + goto st0; +st15: + if ( ++p == pe ) + goto _test_eof15; +case 15: + switch( (*p) ) { + case 13: goto st16; + case 33: goto tr20; + case 124: goto tr20; + case 126: goto tr20; + } + if ( (*p) < 45 ) { + if ( (*p) > 39 ) { + if ( 42 <= (*p) && (*p) <= 43 ) + goto tr20; + } else if ( (*p) >= 35 ) + goto tr20; + } else if ( (*p) > 46 ) { + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto tr20; + } else if ( (*p) > 90 ) { + if ( 94 <= (*p) && (*p) <= 122 ) + goto tr20; + } else + goto tr20; + } else + goto tr20; + goto st0; +st16: + if ( ++p == pe ) + goto _test_eof16; +case 16: + if ( (*p) == 10 ) + goto tr21; + goto st0; +tr21: +#line 102 "unicorn_http.rl" + { + parser->body_start = p - buffer + 1; + header_done(parser->data, p + 1, pe - p - 1); + {p++; cs = 63; goto _out;} + } + goto st63; +st63: + if ( ++p == pe ) + goto _test_eof63; +case 63: +#line 388 "unicorn_http.h" + goto st0; +tr20: +#line 66 "unicorn_http.rl" + { MARK(field_start, p); } +#line 67 "unicorn_http.rl" + { snake_upcase_char((char *)p); } + goto st17; +tr22: +#line 67 "unicorn_http.rl" + { snake_upcase_char((char *)p); } + goto st17; +st17: + if ( ++p == pe ) + goto _test_eof17; +case 17: +#line 404 "unicorn_http.h" + switch( (*p) ) { + case 33: goto tr22; + case 58: goto tr23; + case 124: goto tr22; + case 126: goto tr22; + } + if ( (*p) < 45 ) { + if ( (*p) > 39 ) { + if ( 42 <= (*p) && (*p) <= 43 ) + goto tr22; + } else if ( (*p) >= 35 ) + goto tr22; + } else if ( (*p) > 46 ) { + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto tr22; + } else if ( (*p) > 90 ) { + if ( 94 <= (*p) && (*p) <= 122 ) + goto tr22; + } else + goto tr22; + } else + goto tr22; + goto st0; +tr23: +#line 69 "unicorn_http.rl" + { + parser->field_len = LEN(field_start, p); + } + goto st18; +tr26: +#line 73 "unicorn_http.rl" + { MARK(mark, p); } + goto st18; +st18: + if ( ++p == pe ) + goto _test_eof18; +case 18: +#line 443 "unicorn_http.h" + switch( (*p) ) { + case 13: goto tr25; + case 32: goto tr26; + } + goto tr24; +tr24: +#line 73 "unicorn_http.rl" + { MARK(mark, p); } + goto st19; +st19: + if ( ++p == pe ) + goto _test_eof19; +case 19: +#line 457 "unicorn_http.h" + if ( (*p) == 13 ) + goto tr28; + goto st19; +tr8: +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st20; +tr38: +#line 98 "unicorn_http.rl" + { + request_path(parser->data, PTR_TO(mark), LEN(mark,p)); + } +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st20; +tr49: +#line 89 "unicorn_http.rl" + {MARK(query_start, p); } +#line 90 "unicorn_http.rl" + { + query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); + } +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st20; +tr53: +#line 90 "unicorn_http.rl" + { + query_string(parser->data, PTR_TO(query_start), LEN(query_start, p)); + } +#line 82 "unicorn_http.rl" + { + request_uri(parser->data, PTR_TO(mark), LEN(mark, p)); + } + goto st20; +st20: + if ( ++p == pe ) + goto _test_eof20; +case 20: +#line 503 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr30; + case 35: goto st0; + case 37: goto tr31; + case 127: goto st0; + } + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto tr29; +tr29: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st21; +st21: + if ( ++p == pe ) + goto _test_eof21; +case 21: +#line 521 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr33; + case 35: goto st0; + case 37: goto st22; + case 127: goto st0; + } + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto st21; +tr31: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st22; +st22: + if ( ++p == pe ) + goto _test_eof22; +case 22: +#line 539 "unicorn_http.h" + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st23; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st23; + } else + goto st23; + goto st0; +st23: + if ( ++p == pe ) + goto _test_eof23; +case 23: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st21; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st21; + } else + goto st21; + goto st0; +tr5: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st24; +tr65: +#line 81 "unicorn_http.rl" + { host(parser->data, PTR_TO(mark), LEN(mark, p)); } +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st24; +st24: + if ( ++p == pe ) + goto _test_eof24; +case 24: +#line 576 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr37; + case 35: goto tr38; + case 37: goto st25; + case 59: goto tr40; + case 63: goto tr41; + case 127: goto st0; + } + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto st24; +st25: + if ( ++p == pe ) + goto _test_eof25; +case 25: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st26; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st26; + } else + goto st26; + goto st0; +st26: + if ( ++p == pe ) + goto _test_eof26; +case 26: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st24; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st24; + } else + goto st24; + goto st0; +tr40: +#line 98 "unicorn_http.rl" + { + request_path(parser->data, PTR_TO(mark), LEN(mark,p)); + } + goto st27; +st27: + if ( ++p == pe ) + goto _test_eof27; +case 27: +#line 624 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr7; + case 35: goto tr8; + case 37: goto st28; + case 63: goto st30; + case 127: goto st0; + } + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto st27; +st28: + if ( ++p == pe ) + goto _test_eof28; +case 28: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st29; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st29; + } else + goto st29; + goto st0; +st29: + if ( ++p == pe ) + goto _test_eof29; +case 29: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st27; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st27; + } else + goto st27; + goto st0; +tr41: +#line 98 "unicorn_http.rl" + { + request_path(parser->data, PTR_TO(mark), LEN(mark,p)); + } + goto st30; +st30: + if ( ++p == pe ) + goto _test_eof30; +case 30: +#line 671 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr48; + case 35: goto tr49; + case 37: goto tr50; + case 127: goto st0; + } + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto tr47; +tr47: +#line 89 "unicorn_http.rl" + {MARK(query_start, p); } + goto st31; +st31: + if ( ++p == pe ) + goto _test_eof31; +case 31: +#line 689 "unicorn_http.h" + switch( (*p) ) { + case 32: goto tr52; + case 35: goto tr53; + case 37: goto st32; + case 127: goto st0; + } + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto st31; +tr50: +#line 89 "unicorn_http.rl" + {MARK(query_start, p); } + goto st32; +st32: + if ( ++p == pe ) + goto _test_eof32; +case 32: +#line 707 "unicorn_http.h" + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st33; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st33; + } else + goto st33; + goto st0; +st33: + if ( ++p == pe ) + goto _test_eof33; +case 33: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st31; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st31; + } else + goto st31; + goto st0; +tr6: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } +#line 68 "unicorn_http.rl" + { downcase_char((char *)p); } + goto st34; +st34: + if ( ++p == pe ) + goto _test_eof34; +case 34: +#line 740 "unicorn_http.h" + switch( (*p) ) { + case 84: goto tr56; + case 116: goto tr56; + } + goto st0; +tr56: +#line 68 "unicorn_http.rl" + { downcase_char((char *)p); } + goto st35; +st35: + if ( ++p == pe ) + goto _test_eof35; +case 35: +#line 754 "unicorn_http.h" + switch( (*p) ) { + case 84: goto tr57; + case 116: goto tr57; + } + goto st0; +tr57: +#line 68 "unicorn_http.rl" + { downcase_char((char *)p); } + goto st36; +st36: + if ( ++p == pe ) + goto _test_eof36; +case 36: +#line 768 "unicorn_http.h" + switch( (*p) ) { + case 80: goto tr58; + case 112: goto tr58; + } + goto st0; +tr58: +#line 68 "unicorn_http.rl" + { downcase_char((char *)p); } + goto st37; +st37: + if ( ++p == pe ) + goto _test_eof37; +case 37: +#line 782 "unicorn_http.h" + switch( (*p) ) { + case 58: goto tr59; + case 83: goto tr60; + case 115: goto tr60; + } + goto st0; +tr59: +#line 80 "unicorn_http.rl" + { scheme(parser->data, PTR_TO(mark), LEN(mark, p)); } + goto st38; +st38: + if ( ++p == pe ) + goto _test_eof38; +case 38: +#line 797 "unicorn_http.h" + if ( (*p) == 47 ) + goto st39; + goto st0; +st39: + if ( ++p == pe ) + goto _test_eof39; +case 39: + if ( (*p) == 47 ) + goto st40; + goto st0; +st40: + if ( ++p == pe ) + goto _test_eof40; +case 40: + if ( (*p) == 95 ) + goto tr63; + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto tr63; + } else if ( (*p) > 57 ) { + if ( (*p) > 90 ) { + if ( 97 <= (*p) && (*p) <= 122 ) + goto tr63; + } else if ( (*p) >= 65 ) + goto tr63; + } else + goto tr63; + goto st0; +tr63: +#line 64 "unicorn_http.rl" + {MARK(mark, p); } + goto st41; +st41: + if ( ++p == pe ) + goto _test_eof41; +case 41: +#line 834 "unicorn_http.h" + switch( (*p) ) { + case 47: goto tr65; + case 58: goto st42; + case 95: goto st41; + } + if ( (*p) < 65 ) { + if ( 45 <= (*p) && (*p) <= 57 ) + goto st41; + } else if ( (*p) > 90 ) { + if ( 97 <= (*p) && (*p) <= 122 ) + goto st41; + } else + goto st41; + goto st0; +st42: + if ( ++p == pe ) + goto _test_eof42; +case 42: + if ( (*p) == 47 ) + goto tr65; + if ( 48 <= (*p) && (*p) <= 57 ) + goto st42; + goto st0; +tr60: +#line 68 "unicorn_http.rl" + { downcase_char((char *)p); } + goto st43; +st43: + if ( ++p == pe ) + goto _test_eof43; +case 43: +#line 866 "unicorn_http.h" + if ( (*p) == 58 ) + goto tr59; + goto st0; +st44: + if ( ++p == pe ) + goto _test_eof44; +case 44: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st45; + case 95: goto st45; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st45; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st45; + } else + goto st45; + goto st0; +st45: + if ( ++p == pe ) + goto _test_eof45; +case 45: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st46; + case 95: goto st46; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st46; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st46; + } else + goto st46; + goto st0; +st46: + if ( ++p == pe ) + goto _test_eof46; +case 46: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st47; + case 95: goto st47; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st47; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st47; + } else + goto st47; + goto st0; +st47: + if ( ++p == pe ) + goto _test_eof47; +case 47: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st48; + case 95: goto st48; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st48; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st48; + } else + goto st48; + goto st0; +st48: + if ( ++p == pe ) + goto _test_eof48; +case 48: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st49; + case 95: goto st49; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st49; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st49; + } else + goto st49; + goto st0; +st49: + if ( ++p == pe ) + goto _test_eof49; +case 49: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st50; + case 95: goto st50; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st50; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st50; + } else + goto st50; + goto st0; +st50: + if ( ++p == pe ) + goto _test_eof50; +case 50: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st51; + case 95: goto st51; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st51; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st51; + } else + goto st51; + goto st0; +st51: + if ( ++p == pe ) + goto _test_eof51; +case 51: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st52; + case 95: goto st52; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st52; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st52; + } else + goto st52; + goto st0; +st52: + if ( ++p == pe ) + goto _test_eof52; +case 52: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st53; + case 95: goto st53; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st53; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st53; + } else + goto st53; + goto st0; +st53: + if ( ++p == pe ) + goto _test_eof53; +case 53: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st54; + case 95: goto st54; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st54; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st54; + } else + goto st54; + goto st0; +st54: + if ( ++p == pe ) + goto _test_eof54; +case 54: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st55; + case 95: goto st55; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st55; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st55; + } else + goto st55; + goto st0; +st55: + if ( ++p == pe ) + goto _test_eof55; +case 55: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st56; + case 95: goto st56; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st56; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st56; + } else + goto st56; + goto st0; +st56: + if ( ++p == pe ) + goto _test_eof56; +case 56: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st57; + case 95: goto st57; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st57; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st57; + } else + goto st57; + goto st0; +st57: + if ( ++p == pe ) + goto _test_eof57; +case 57: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st58; + case 95: goto st58; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st58; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st58; + } else + goto st58; + goto st0; +st58: + if ( ++p == pe ) + goto _test_eof58; +case 58: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st59; + case 95: goto st59; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st59; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st59; + } else + goto st59; + goto st0; +st59: + if ( ++p == pe ) + goto _test_eof59; +case 59: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st60; + case 95: goto st60; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st60; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st60; + } else + goto st60; + goto st0; +st60: + if ( ++p == pe ) + goto _test_eof60; +case 60: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st61; + case 95: goto st61; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st61; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st61; + } else + goto st61; + goto st0; +st61: + if ( ++p == pe ) + goto _test_eof61; +case 61: + switch( (*p) ) { + case 32: goto tr2; + case 36: goto st62; + case 95: goto st62; + } + if ( (*p) < 48 ) { + if ( 45 <= (*p) && (*p) <= 46 ) + goto st62; + } else if ( (*p) > 57 ) { + if ( 65 <= (*p) && (*p) <= 90 ) + goto st62; + } else + goto st62; + goto st0; +st62: + if ( ++p == pe ) + goto _test_eof62; +case 62: + if ( (*p) == 32 ) + goto tr2; + goto st0; + } + _test_eof2: cs = 2; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof6: cs = 6; goto _test_eof; + _test_eof7: cs = 7; goto _test_eof; + _test_eof8: cs = 8; goto _test_eof; + _test_eof9: cs = 9; goto _test_eof; + _test_eof10: cs = 10; goto _test_eof; + _test_eof11: cs = 11; goto _test_eof; + _test_eof12: cs = 12; goto _test_eof; + _test_eof13: cs = 13; goto _test_eof; + _test_eof14: cs = 14; goto _test_eof; + _test_eof15: cs = 15; goto _test_eof; + _test_eof16: cs = 16; goto _test_eof; + _test_eof63: cs = 63; goto _test_eof; + _test_eof17: cs = 17; goto _test_eof; + _test_eof18: cs = 18; goto _test_eof; + _test_eof19: cs = 19; goto _test_eof; + _test_eof20: cs = 20; goto _test_eof; + _test_eof21: cs = 21; goto _test_eof; + _test_eof22: cs = 22; goto _test_eof; + _test_eof23: cs = 23; goto _test_eof; + _test_eof24: cs = 24; goto _test_eof; + _test_eof25: cs = 25; goto _test_eof; + _test_eof26: cs = 26; goto _test_eof; + _test_eof27: cs = 27; goto _test_eof; + _test_eof28: cs = 28; goto _test_eof; + _test_eof29: cs = 29; goto _test_eof; + _test_eof30: cs = 30; goto _test_eof; + _test_eof31: cs = 31; goto _test_eof; + _test_eof32: cs = 32; goto _test_eof; + _test_eof33: cs = 33; goto _test_eof; + _test_eof34: cs = 34; goto _test_eof; + _test_eof35: cs = 35; goto _test_eof; + _test_eof36: cs = 36; goto _test_eof; + _test_eof37: cs = 37; goto _test_eof; + _test_eof38: cs = 38; goto _test_eof; + _test_eof39: cs = 39; goto _test_eof; + _test_eof40: cs = 40; goto _test_eof; + _test_eof41: cs = 41; goto _test_eof; + _test_eof42: cs = 42; goto _test_eof; + _test_eof43: cs = 43; goto _test_eof; + _test_eof44: cs = 44; goto _test_eof; + _test_eof45: cs = 45; goto _test_eof; + _test_eof46: cs = 46; goto _test_eof; + _test_eof47: cs = 47; goto _test_eof; + _test_eof48: cs = 48; goto _test_eof; + _test_eof49: cs = 49; goto _test_eof; + _test_eof50: cs = 50; goto _test_eof; + _test_eof51: cs = 51; goto _test_eof; + _test_eof52: cs = 52; goto _test_eof; + _test_eof53: cs = 53; goto _test_eof; + _test_eof54: cs = 54; goto _test_eof; + _test_eof55: cs = 55; goto _test_eof; + _test_eof56: cs = 56; goto _test_eof; + _test_eof57: cs = 57; goto _test_eof; + _test_eof58: cs = 58; goto _test_eof; + _test_eof59: cs = 59; goto _test_eof; + _test_eof60: cs = 60; goto _test_eof; + _test_eof61: cs = 61; goto _test_eof; + _test_eof62: cs = 62; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 138 "unicorn_http.rl" + + if (!http_parser_has_error(parser)) + parser->cs = cs; + parser->nread += p - (buffer + off); + + assert(p <= pe && "buffer overflow after parsing execute"); + assert(parser->nread <= len && "nread longer than length"); + assert(parser->body_start <= len && "body starts after buffer end"); + assert(parser->mark < len && "mark is after buffer end"); + assert(parser->field_len <= len && "field has length longer than whole buffer"); + assert(parser->field_start < len && "field starts after buffer end"); +} + +static int http_parser_has_error(http_parser *parser) { + return parser->cs == http_parser_error; +} + +static int http_parser_is_finished(http_parser *parser) { + return parser->cs == http_parser_first_final; +} +#endif /* unicorn_http_h */ diff --git a/ext/unicorn_http/unicorn_http.rl b/ext/unicorn_http/unicorn_http.rl new file mode 100644 index 0000000..d487a83 --- /dev/null +++ b/ext/unicorn_http/unicorn_http.rl @@ -0,0 +1,158 @@ +/** + * Copyright (c) 2005 Zed A. Shaw + * You can redistribute it and/or modify it under the same terms as Ruby. + */ +#ifndef unicorn_http_h +#define unicorn_http_h + +#include + +static void http_field(void *data, const char *field, + size_t flen, const char *value, size_t vlen); +static void request_method(void *data, const char *at, size_t length); +static void scheme(void *data, const char *at, size_t length); +static void host(void *data, const char *at, size_t length); +static void request_uri(void *data, const char *at, size_t length); +static void fragment(void *data, const char *at, size_t length); +static void request_path(void *data, const char *at, size_t length); +static void query_string(void *data, const char *at, size_t length); +static void http_version(void *data, const char *at, size_t length); +static void header_done(void *data, const char *at, size_t length); + +typedef struct http_parser { + int cs; + size_t body_start; + size_t nread; + size_t mark; + size_t field_start; + size_t field_len; + size_t query_start; + + void *data; +} http_parser; + +static int http_parser_has_error(http_parser *parser); +static int http_parser_is_finished(http_parser *parser); + +/* + * capitalizes all lower-case ASCII characters, + * converts dashes to underscores. + */ +static void snake_upcase_char(char *c) +{ + if (*c >= 'a' && *c <= 'z') + *c &= ~0x20; + else if (*c == '-') + *c = '_'; +} + +static void downcase_char(char *c) +{ + if (*c >= 'A' && *c <= 'Z') + *c |= 0x20; +} + +#define LEN(AT, FPC) (FPC - buffer - parser->AT) +#define MARK(M,FPC) (parser->M = (FPC) - buffer) +#define PTR_TO(F) (buffer + parser->F) + +/** Machine **/ + +%%{ + machine http_parser; + + action mark {MARK(mark, fpc); } + + action start_field { MARK(field_start, fpc); } + action snake_upcase_field { snake_upcase_char((char *)fpc); } + action downcase_char { downcase_char((char *)fpc); } + action write_field { + parser->field_len = LEN(field_start, fpc); + } + + action start_value { MARK(mark, fpc); } + action write_value { + http_field(parser->data, PTR_TO(field_start), parser->field_len, PTR_TO(mark), LEN(mark, fpc)); + } + action request_method { + request_method(parser->data, PTR_TO(mark), LEN(mark, fpc)); + } + action scheme { scheme(parser->data, PTR_TO(mark), LEN(mark, fpc)); } + action host { host(parser->data, PTR_TO(mark), LEN(mark, fpc)); } + action request_uri { + request_uri(parser->data, PTR_TO(mark), LEN(mark, fpc)); + } + action fragment { + fragment(parser->data, PTR_TO(mark), LEN(mark, fpc)); + } + + action start_query {MARK(query_start, fpc); } + action query_string { + query_string(parser->data, PTR_TO(query_start), LEN(query_start, fpc)); + } + + action http_version { + http_version(parser->data, PTR_TO(mark), LEN(mark, fpc)); + } + + action request_path { + request_path(parser->data, PTR_TO(mark), LEN(mark,fpc)); + } + + action done { + parser->body_start = fpc - buffer + 1; + header_done(parser->data, fpc + 1, pe - fpc - 1); + fbreak; + } + + include unicorn_http_common "unicorn_http_common.rl"; +}%% + +/** Data **/ +%% write data; + +static void http_parser_init(http_parser *parser) { + int cs = 0; + memset(parser, 0, sizeof(*parser)); + %% write init; + parser->cs = cs; +} + +/** exec **/ +static void http_parser_execute( + http_parser *parser, const char *buffer, size_t len) +{ + const char *p, *pe; + int cs = parser->cs; + size_t off = parser->nread; + + assert(off <= len && "offset past end of buffer"); + + p = buffer+off; + pe = buffer+len; + + assert(*pe == '\0' && "pointer does not end on NUL"); + assert(pe - p == len - off && "pointers aren't same distance"); + + %% write exec; + + if (!http_parser_has_error(parser)) + parser->cs = cs; + parser->nread += p - (buffer + off); + + assert(p <= pe && "buffer overflow after parsing execute"); + assert(parser->nread <= len && "nread longer than length"); + assert(parser->body_start <= len && "body starts after buffer end"); + assert(parser->mark < len && "mark is after buffer end"); + assert(parser->field_len <= len && "field has length longer than whole buffer"); + assert(parser->field_start < len && "field starts after buffer end"); +} + +static int http_parser_has_error(http_parser *parser) { + return parser->cs == http_parser_error; +} + +static int http_parser_is_finished(http_parser *parser) { + return parser->cs == http_parser_first_final; +} +#endif /* unicorn_http_h */ diff --git a/ext/unicorn_http/unicorn_http_common.rl b/ext/unicorn_http/unicorn_http_common.rl new file mode 100644 index 0000000..61e6d52 --- /dev/null +++ b/ext/unicorn_http/unicorn_http_common.rl @@ -0,0 +1,58 @@ +%%{ + + machine unicorn_http_common; + +#### HTTP PROTOCOL GRAMMAR +# line endings + CRLF = "\r\n"; + +# character types + CTL = (cntrl | 127); + safe = ("$" | "-" | "_" | "."); + extra = ("!" | "*" | "'" | "(" | ")" | ","); + reserved = (";" | "/" | "?" | ":" | "@" | "&" | "=" | "+"); + sorta_safe = ("\"" | "<" | ">"); + unsafe = (CTL | " " | "#" | "%" | sorta_safe); + national = any -- (alpha | digit | reserved | extra | safe | unsafe); + unreserved = (alpha | digit | safe | extra | national); + escape = ("%" xdigit xdigit); + uchar = (unreserved | escape | sorta_safe); + pchar = (uchar | ":" | "@" | "&" | "=" | "+"); + tspecials = ("(" | ")" | "<" | ">" | "@" | "," | ";" | ":" | "\\" | "\"" | "/" | "[" | "]" | "?" | "=" | "{" | "}" | " " | "\t"); + +# elements + token = (ascii -- (CTL | tspecials)); + +# URI schemes and absolute paths + scheme = ( "http"i ("s"i)? ) $downcase_char >mark %scheme; + hostname = (alnum | "-" | "." | "_")+; + host_with_port = (hostname (":" digit*)?) >mark %host; + + path = ( pchar+ ( "/" pchar* )* ) ; + query = ( uchar | reserved )* %query_string ; + param = ( pchar | "/" )* ; + params = ( param ( ";" param )* ) ; + rel_path = ( path? %request_path (";" params)? ) ("?" %start_query query)?; + absolute_path = ( "/"+ rel_path ); + path_uri = absolute_path > mark %request_uri; + Absolute_URI = (scheme "://" host_with_port path_uri); + + Request_URI = ((absolute_path | "*") >mark %request_uri) | Absolute_URI; + Fragment = ( uchar | reserved )* >mark %fragment; + Method = ( upper | digit | safe ){1,20} >mark %request_method; + + http_number = ( digit+ "." digit+ ) ; + HTTP_Version = ( "HTTP/" http_number ) >mark %http_version ; + Request_Line = ( Method " " Request_URI ("#" Fragment){0,1} " " HTTP_Version CRLF ) ; + + field_name = ( token -- ":" )+ >start_field $snake_upcase_field %write_field; + + field_value = any* >start_value %write_value; + + message_header = field_name ":" " "* field_value :> CRLF; + + Request = Request_Line ( message_header )* ( CRLF @done ); + +main := Request; + +}%% diff --git a/lib/unicorn/chunked_reader.rb b/lib/unicorn/chunked_reader.rb index 606e4a6..539b350 100644 --- a/lib/unicorn/chunked_reader.rb +++ b/lib/unicorn/chunked_reader.rb @@ -2,7 +2,7 @@ # You can redistribute it and/or modify it under the same terms as Ruby. require 'unicorn' -require 'unicorn/http11' +require 'unicorn_http' module Unicorn class ChunkedReader diff --git a/lib/unicorn/http_request.rb b/lib/unicorn/http_request.rb index b8df403..f4cc77c 100644 --- a/lib/unicorn/http_request.rb +++ b/lib/unicorn/http_request.rb @@ -1,7 +1,7 @@ require 'stringio' # compiled extension -require 'unicorn/http11' +require 'unicorn_http' module Unicorn class HttpRequest diff --git a/lib/unicorn/trailer_parser.rb b/lib/unicorn/trailer_parser.rb index 9431331..22f2e1d 100644 --- a/lib/unicorn/trailer_parser.rb +++ b/lib/unicorn/trailer_parser.rb @@ -1,7 +1,7 @@ # Copyright (c) 2009 Eric Wong # You can redistribute it and/or modify it under the same terms as Ruby. require 'unicorn' -require 'unicorn/http11' +require 'unicorn_http' # Eventually I should integrate this into HttpParser... module Unicorn diff --git a/test/test_helper.rb b/test/test_helper.rb index 0f2f311..29800e9 100644 --- a/test/test_helper.rb +++ b/test/test_helper.rb @@ -27,7 +27,7 @@ require 'tempfile' require 'fileutils' require 'logger' require 'unicorn' -require 'unicorn/http11' +require 'unicorn_http' if ENV['DEBUG'] require 'ruby-debug' diff --git a/test/unit/test_chunked_reader.rb b/test/unit/test_chunked_reader.rb index 67fe43b..6aaa72a 100644 --- a/test/unit/test_chunked_reader.rb +++ b/test/unit/test_chunked_reader.rb @@ -1,6 +1,6 @@ require 'test/unit' require 'unicorn' -require 'unicorn/http11' +require 'unicorn_http' require 'tempfile' require 'io/nonblock' require 'digest/sha1' diff --git a/test/unit/test_trailer_parser.rb b/test/unit/test_trailer_parser.rb index 840e9ad..5f3b16d 100644 --- a/test/unit/test_trailer_parser.rb +++ b/test/unit/test_trailer_parser.rb @@ -1,6 +1,6 @@ require 'test/unit' require 'unicorn' -require 'unicorn/http11' +require 'unicorn_http' require 'unicorn/trailer_parser' class TestTrailerParser < Test::Unit::TestCase -- cgit v1.2.3-24-ge0c7