thin 0.4.1 → 0.5.0

data/ext/thin_parser/parser.h

@@ -0,0 +1,49 @@
+/**
+ * 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 <sys/types.h>
+
+#if defined(_WIN32)
+#include <stddef.h>
+#endif
+
+typedef void (*element_cb)(void *data, const char *at, size_t length);
+typedef void (*field_cb)(void *data, const char *field, size_t flen, const char *value, size_t vlen);
+
+typedef struct http_parser { 
+  int cs;
+  size_t body_start;
+  int content_len;
+  size_t nread;
+  size_t mark;
+  size_t field_start;
+  size_t field_len;
+  size_t query_start;
+
+  void *data;
+
+  field_cb http_field;
+  element_cb request_method;
+  element_cb request_uri;
+  element_cb fragment;
+  element_cb request_path;
+  element_cb query_string;
+  element_cb http_version;
+  element_cb header_done;
+  
+} http_parser;
+
+int http_parser_init(http_parser *parser);
+int http_parser_finish(http_parser *parser);
+size_t http_parser_execute(http_parser *parser, const char *data, size_t len, size_t off);
+int http_parser_has_error(http_parser *parser);
+int http_parser_is_finished(http_parser *parser);
+
+#define http_parser_nread(parser) (parser)->nread 
+
+#endif

data/ext/thin_parser/parser.rl

@@ -0,0 +1,143 @@
+/**
+ * Copyright (c) 2005 Zed A. Shaw
+ * You can redistribute it and/or modify it under the same terms as Ruby.
+ */
+#include "parser.h"
+#include <stdio.h>
+#include <assert.h>
+#include <stdlib.h>
+#include <ctype.h>
+#include <string.h>
+
+#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 write_field { 
+    parser->field_len = LEN(field_start, fpc);
+  }
+
+  action start_value { MARK(mark, fpc); }
+  action write_value { 
+    parser->http_field(parser->data, PTR_TO(field_start), parser->field_len, PTR_TO(mark), LEN(mark, fpc));
+  }
+  action request_method { 
+    parser->request_method(parser->data, PTR_TO(mark), LEN(mark, fpc));
+  }
+  action request_uri { 
+    parser->request_uri(parser->data, PTR_TO(mark), LEN(mark, fpc));
+  }
+  action fragment { 
+    parser->fragment(parser->data, PTR_TO(mark), LEN(mark, fpc));
+  }
+
+  action start_query {MARK(query_start, fpc); }
+  action query_string { 
+    parser->query_string(parser->data, PTR_TO(query_start), LEN(query_start, fpc));
+  }
+
+  action http_version {	
+    parser->http_version(parser->data, PTR_TO(mark), LEN(mark, fpc));
+  }
+
+  action request_path {
+    parser->request_path(parser->data, PTR_TO(mark), LEN(mark,fpc));
+  }
+
+  action done { 
+    parser->body_start = fpc - buffer + 1; 
+    parser->header_done(parser->data, fpc + 1, pe - fpc - 1);
+    fbreak;
+  }
+
+  include http_parser_common "common.rl";
+
+}%%
+
+/** Data **/
+%% write data;
+
+int http_parser_init(http_parser *parser)  {
+  int cs = 0;
+  %% write init;
+  parser->cs = cs;
+  parser->body_start = 0;
+  parser->content_len = 0;
+  parser->mark = 0;
+  parser->nread = 0;
+  parser->field_len = 0;
+  parser->field_start = 0;    
+
+  return(1);
+}
+
+
+/** exec **/
+size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len, size_t off)  {
+  const char *p, *pe;
+  int cs = parser->cs;
+
+  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;
+
+  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");
+
+  if(parser->body_start) {
+    /* final \r\n combo encountered so stop right here */
+    %%write eof;
+    parser->nread++;
+  }
+
+  return(parser->nread);
+}
+
+int http_parser_finish(http_parser *parser)
+{
+  int cs = parser->cs;
+
+  %%write eof;
+
+  parser->cs = cs;
+
+  if (http_parser_has_error(parser) ) {
+    return -1;
+  } else if (http_parser_is_finished(parser) ) {
+    return 1;
+  } else {
+    return 0;
+  }
+}
+
+int http_parser_has_error(http_parser *parser) {
+  return parser->cs == http_parser_error;
+}
+
+int http_parser_is_finished(http_parser *parser) {
+  return parser->cs == http_parser_first_final;
+}

data/ext/thin_parser/thin.c

@@ -0,0 +1,424 @@
+/**
+ * Mongrel Parser adpated to Thin and to play more nicely with Rack specs.
+ * 
+ * Orignal version 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 <assert.h>
+#include <string.h>
+#include "parser.h"
+#include <ctype.h>
+
+static VALUE mThin;
+static VALUE cHttpParser;
+static VALUE eHttpParserError;
+
+static VALUE global_empty;
+static VALUE global_http_prefix;
+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_content_length;
+static VALUE global_http_content_length;
+static VALUE global_request_path;
+static VALUE global_content_type;
+static VALUE global_http_content_type;
+static VALUE global_gateway_interface;
+static VALUE global_gateway_interface_value;
+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_port_80;
+static VALUE global_http_body;
+static VALUE global_url_scheme;
+static VALUE global_url_scheme_value;
+static VALUE global_script_name;
+static VALUE global_path_info;
+
+#define TRIE_INCREASE 30
+
+/** Defines common length and error messages for input length validation. */
+#define DEF_MAX_LENGTH(N,length) const size_t MAX_##N##_LENGTH = length; const char *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) if(len > MAX_##N##_LENGTH) { rb_raise(eHttpParserError, MAX_##N##_LENGTH_ERR); }
+
+/** Defines global strings in the init method. */
+#define DEF_GLOBAL(N, val)   global_##N = rb_obj_freeze(rb_str_new2(val)); rb_global_variable(&global_##N)
+
+
+/* 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)));
+
+
+void http_field(void *data, const char *field, size_t flen, const char *value, size_t vlen)
+{
+  char *ch, *end;
+  VALUE req = (VALUE)data;
+  VALUE v = Qnil;
+  VALUE f = Qnil;
+
+  VALIDATE_MAX_LENGTH(flen, FIELD_NAME);
+  VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE);
+
+  v = rb_str_new(value, vlen);
+  f = rb_str_dup(global_http_prefix);
+  f = rb_str_buf_cat(f, field, flen); 
+
+  for(ch = RSTRING(f)->ptr, end = ch + RSTRING(f)->len; ch < end; ch++) {
+    if(*ch == '-') {
+      *ch = '_';
+    } else {
+      *ch = toupper(*ch);
+    }
+  }
+
+  rb_hash_aset(req, f, v);
+}
+
+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);
+}
+
+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);
+}
+
+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);
+}
+
+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);
+  rb_hash_aset(req, global_path_info, val);
+}
+
+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);
+}
+
+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. */
+
+void header_done(void *data, const char *at, size_t length)
+{
+  VALUE req = (VALUE)data;
+  VALUE temp = Qnil;
+  VALUE ctype = Qnil;
+  VALUE clen = Qnil;
+  VALUE body = Qnil;
+  char *colon = NULL;
+
+  clen = rb_hash_aref(req, global_http_content_length);
+  if(clen != Qnil) {
+    rb_hash_aset(req, global_content_length, clen);
+    rb_hash_delete(req, global_http_content_length);
+  }
+
+  ctype = rb_hash_aref(req, global_http_content_type);
+  if(ctype != Qnil) {
+    rb_hash_aset(req, global_content_type, ctype);
+    rb_hash_delete(req, global_http_content_type);
+  }
+
+  rb_hash_aset(req, global_gateway_interface, global_gateway_interface_value);
+  if((temp = rb_hash_aref(req, global_http_host)) != Qnil) {
+    /* ruby better close strings off with a '\0' dammit */
+    colon = strchr(RSTRING(temp)->ptr, ':');
+    if(colon != NULL) {
+      rb_hash_aset(req, global_server_name, rb_str_substr(temp, 0, colon - RSTRING(temp)->ptr));
+      rb_hash_aset(req, global_server_port, 
+          rb_str_substr(temp, colon - RSTRING(temp)->ptr+1, 
+            RSTRING(temp)->len));
+    } else {
+      rb_hash_aset(req, global_server_name, temp);
+      rb_hash_aset(req, global_server_port, global_port_80);
+    }
+  }
+
+  /* grab the initial body and stuff it into the hash */
+  if(length > 0) {
+    body = rb_hash_aref(req, global_http_body);
+    rb_io_write(body, rb_str_new(at, length));
+  }
+  
+  /* according to Rack specs, query string must be empty string if none */
+  if (rb_hash_aref(req, global_query_string) == Qnil) {
+    rb_hash_aset(req, global_query_string, global_empty);
+  }
+  
+  /* set some constants */
+  rb_hash_aset(req, global_server_protocol, global_server_protocol_value);
+  rb_hash_aset(req, global_url_scheme, global_url_scheme_value);
+  rb_hash_aset(req, global_script_name, global_empty);
+}
+
+
+void HttpParser_free(void *data) {
+  TRACE();
+
+  if(data) {
+    free(data);
+  }
+}
+
+
+VALUE HttpParser_alloc(VALUE klass)
+{
+  VALUE obj;
+  http_parser *hp = ALLOC_N(http_parser, 1);
+  TRACE();
+  hp->http_field = http_field;
+  hp->request_method = request_method;
+  hp->request_uri = request_uri;
+  hp->fragment = fragment;
+  hp->request_path = request_path;
+  hp->query_string = query_string;
+  hp->http_version = http_version;
+  hp->header_done = header_done;
+  http_parser_init(hp);
+
+  obj = Data_Wrap_Struct(klass, NULL, HttpParser_free, hp);
+
+  return obj;
+}
+
+
+/**
+ * call-seq:
+ *    parser.new -> parser
+ *
+ * Creates a new parser.
+ */
+VALUE HttpParser_init(VALUE self)
+{
+  http_parser *http = NULL;
+  DATA_GET(self, http_parser, http);
+  http_parser_init(http);
+
+  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.
+ */
+VALUE HttpParser_reset(VALUE self)
+{
+  http_parser *http = NULL;
+  DATA_GET(self, http_parser, http);
+  http_parser_init(http);
+
+  return Qnil;
+}
+
+
+/**
+ * call-seq:
+ *    parser.finish -> true/false
+ *
+ * Finishes a parser early which could put in a "good" or bad state.
+ * You should call reset after finish it or bad things will happen.
+ */
+VALUE HttpParser_finish(VALUE self)
+{
+  http_parser *http = NULL;
+  DATA_GET(self, http_parser, http);
+  http_parser_finish(http);
+
+  return http_parser_is_finished(http) ? Qtrue : Qfalse;
+}
+
+
+/**
+ * call-seq:
+ *    parser.execute(req_hash, data, start) -> Integer
+ *
+ * Takes a Hash and a String of data, parses the String of data filling in the Hash
+ * returning an Integer to indicate how much of the data has been read.  No matter
+ * what the return value, you should call HttpParser#finished? and HttpParser#error?
+ * to figure out if it's done parsing or there was an error.
+ * 
+ * This function now throws an exception when there is a parsing error.  This makes 
+ * the logic for working with the parser much easier.  You can still test for an 
+ * error, but now you need to wrap the parser with an exception handling block.
+ *
+ * The third argument allows for parsing a partial request and then continuing
+ * the parsing from that position.  It needs all of the original data as well 
+ * so you have to append to the data buffer as you read.
+ */
+VALUE HttpParser_execute(VALUE self, VALUE req_hash, VALUE data, VALUE start)
+{
+  http_parser *http = NULL;
+  int from = 0;
+  char *dptr = NULL;
+  long dlen = 0;
+
+  DATA_GET(self, http_parser, http);
+
+  from = FIX2INT(start);
+  dptr = RSTRING(data)->ptr;
+  dlen = RSTRING(data)->len;
+
+  if(from >= dlen) {
+    rb_raise(eHttpParserError, "Requested start is after data buffer end.");
+  } else {
+    http->data = (void *)req_hash;
+    http_parser_execute(http, dptr, dlen, from);
+
+    VALIDATE_MAX_LENGTH(http_parser_nread(http), HEADER);
+
+    if(http_parser_has_error(http)) {
+      rb_raise(eHttpParserError, "Invalid HTTP format, parsing fails.");
+    } else {
+      return INT2FIX(http_parser_nread(http));
+    }
+  }
+}
+
+
+
+/**
+ * call-seq:
+ *    parser.error? -> true/false
+ *
+ * Tells you whether the parser is in an error state.
+ */
+VALUE HttpParser_has_error(VALUE self)
+{
+  http_parser *http = NULL;
+  DATA_GET(self, http_parser, http);
+
+  return http_parser_has_error(http) ? Qtrue : Qfalse;
+}
+
+
+/**
+ * call-seq:
+ *    parser.finished? -> true/false
+ *
+ * Tells you whether the parser is finished or not and in a good state.
+ */
+VALUE HttpParser_is_finished(VALUE self)
+{
+  http_parser *http = NULL;
+  DATA_GET(self, http_parser, http);
+
+  return http_parser_is_finished(http) ? Qtrue : Qfalse;
+}
+
+
+/**
+ * call-seq:
+ *    parser.nread -> Integer
+ *
+ * Returns the amount of data processed so far during this processing cycle.  It is
+ * set to 0 on initialize or reset calls and is incremented each time execute is called.
+ */
+VALUE HttpParser_nread(VALUE self)
+{
+  http_parser *http = NULL;
+  DATA_GET(self, http_parser, http);
+
+  return INT2FIX(http->nread);
+}
+
+void Init_thin_parser()
+{
+
+  mThin = rb_define_module("Thin");
+
+  DEF_GLOBAL(empty, "");
+  DEF_GLOBAL(http_prefix, "HTTP_");
+  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(content_length, "CONTENT_LENGTH");
+  DEF_GLOBAL(http_content_length, "HTTP_CONTENT_LENGTH");
+  DEF_GLOBAL(content_type, "CONTENT_TYPE");
+  DEF_GLOBAL(http_content_type, "HTTP_CONTENT_TYPE");
+  DEF_GLOBAL(gateway_interface, "GATEWAY_INTERFACE");
+  DEF_GLOBAL(gateway_interface_value, "CGI/1.2");
+  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_host, "HTTP_HOST");
+  DEF_GLOBAL(port_80, "80");
+  DEF_GLOBAL(http_body, "rack.input");
+  DEF_GLOBAL(url_scheme, "rack.url_scheme");
+  DEF_GLOBAL(url_scheme_value, "http");
+  DEF_GLOBAL(script_name, "SCRIPT_NAME");
+  DEF_GLOBAL(path_info, "PATH_INFO");
+
+  eHttpParserError = rb_define_class_under(mThin, "InvalidRequest", rb_eIOError);
+
+  cHttpParser = rb_define_class_under(mThin, "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, "finish", HttpParser_finish,0);
+  rb_define_method(cHttpParser, "execute", HttpParser_execute,3);
+  rb_define_method(cHttpParser, "error?", HttpParser_has_error,0);
+  rb_define_method(cHttpParser, "finished?", HttpParser_is_finished,0);
+  rb_define_method(cHttpParser, "nread", HttpParser_nread,0);
+}