parser3/src/main/pa_http.C - diff

Return to pa_http.C CVS log

Up to [parser3project] / parser3 / src / main

Diff for /parser3/src/main/pa_http.C between versions 1.50 and 1.65

version 1.50, 2010/12/29 12:17:58	version 1.65, 2015/04/30 17:37:43
Line 1	Line 1
/** @file	/** @file
Parser: http support functions.	Parser: http support functions.

Copyright(c) 2001-2009 ArtLebedev Group (http://www.artlebedev.com)	Copyright (c) 2001-2012 Art. Lebedev Studio (http://www.artlebedev.com)
Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)	Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)
*/	*/

static const char * const IDENT_HTTP_C="$Date$";

#include "pa_http.h"	#include "pa_http.h"
#include "pa_common.h"	#include "pa_common.h"
#include "pa_charsets.h"	#include "pa_charsets.h"
Line 15 static const char * const IDENT_HTTP_C="	Line 13 static const char * const IDENT_HTTP_C="
#include "pa_vfile.h"	#include "pa_vfile.h"
#include "pa_random.h"	#include "pa_random.h"

	volatile const char * IDENT_PA_HTTP_C="$Id$" IDENT_PA_HTTP_H;

	#ifdef _MSC_VER
	#include <windows.h>
	#else
	#define closesocket close
	#endif

// defines	// defines

#define HTTP_METHOD_NAME "method"	#define HTTP_METHOD_NAME "method"
Line 22 static const char * const IDENT_HTTP_C="	Line 28 static const char * const IDENT_HTTP_C="
#define HTTP_BODY_NAME "body"	#define HTTP_BODY_NAME "body"
#define HTTP_TIMEOUT_NAME "timeout"	#define HTTP_TIMEOUT_NAME "timeout"
#define HTTP_HEADERS_NAME "headers"	#define HTTP_HEADERS_NAME "headers"
#define HTTP_COOKIES_NAME "cookies"
#define HTTP_FORM_ENCTYPE_NAME "enctype"	#define HTTP_FORM_ENCTYPE_NAME "enctype"
#define HTTP_ANY_STATUS_NAME "any-status"	#define HTTP_ANY_STATUS_NAME "any-status"
#define HTTP_OMIT_POST_CHARSET_NAME "omit-post-charset" // ^file::load[...;http://...;$.form[...]$.method[post]]	#define HTTP_OMIT_POST_CHARSET_NAME "omit-post-charset" // ^file::load[...;http://...;$.method[post]] by default adds charset to content-type
// by default add charset to content-type

#define HTTP_TABLES_NAME "tables"	#define HTTP_TABLES_NAME "tables"

Line 35 static const char * const IDENT_HTTP_C="	Line 39 static const char * const IDENT_HTTP_C="

#define DEFAULT_USER_AGENT "parser3"	#define DEFAULT_USER_AGENT "parser3"

# ifndef INADDR_NONE	#ifndef INADDR_NONE
# define INADDR_NONE ((ulong) -1)	#define INADDR_NONE ((ulong) -1)
# endif	#endif

#undef CRLF	#undef CRLF
#define CRLF "\r\n"	#define CRLF "\r\n"

	// helpers

	class Cookies_table_template_columns: public ArrayString {
	public:
	Cookies_table_template_columns() {
	*this+=new String("name");
	*this+=new String("value");
	*this+=new String("expires");
	*this+=new String("max-age");
	*this+=new String("domain");
	*this+=new String("path");
	*this+=new String("httponly");
	*this+=new String("secure");
	}
	};


static bool set_addr(struct sockaddr_in addr, const char host, const short port){	static bool set_addr(struct sockaddr_in addr, const char host, const short port){
memset(addr, 0, sizeof(*addr));	memset(addr, 0, sizeof(*addr));
addr->sin_family=AF_INET;	addr->sin_family=AF_INET;
addr->sin_port=htons(port);	addr->sin_port=htons(port);
if(host) {	if(host) {
ulong packed_ip=inet_addr(host);	struct hostent *hostIP=gethostbyname(host);
if(packed_ip!=INADDR_NONE)	if(hostIP && hostIP->h_addrtype == AF_INET){
memcpy(&addr->sin_addr, &packed_ip, sizeof(packed_ip));	memcpy(&addr->sin_addr, hostIP->h_addr, hostIP->h_length);
else {	return true;
struct hostent *hostIP=gethostbyname(host);	}
if(hostIP)	}
memcpy(&addr->sin_addr, hostIP->h_addr, hostIP->h_length);	return false;
else
return false;
}
} else
addr->sin_addr.s_addr=INADDR_ANY;
return true;
}	}

size_t guess_content_length(char* buf) {	size_t guess_content_length(char* buf) {
Line 299 struct Http_pass_header_info {	Line 314 struct Http_pass_header_info {
char pa_http_safe_header_name(const char name) {	char pa_http_safe_header_name(const char name) {
char *result=pa_strdup(name);	char *result=pa_strdup(name);
char *n=result;	char *n=result;
if(isdigit(*n))	if(!pa_isalpha((unsigned char)*n))
*n++ = '_';	*n++ = '_';
for(; *n; ++n) {	for(; *n; ++n) {
if (!isalnum(n) && n != '_')	if (!pa_isalnum((unsigned char)n) && n != '-' && *n != '_')
*n = '_';	*n = '_';
}	}
return result;	return result;
Line 326 static void http_pass_header(HashStringV	Line 341 static void http_pass_header(HashStringV
*info->user_agent_specified=true;	*info->user_agent_specified=true;
if(strcasecmp(name_cstr, HTTP_CONTENT_TYPE)==0){	if(strcasecmp(name_cstr, HTTP_CONTENT_TYPE)==0){
*info->content_type_specified=true;	*info->content_type_specified=true;
*info->content_type_url_encoded=StrStartFromNC(value.cstr(), HTTP_CONTENT_TYPE_FORM_URLENCODED);	*info->content_type_url_encoded=pa_strncasecmp(value.cstr(), HTTP_CONTENT_TYPE_FORM_URLENCODED)==0;
}	}
}	}

Line 389 static void form_value2string(	Line 404 static void form_value2string(
} else	} else
throw Exception(PARSER_RUNTIME,	throw Exception(PARSER_RUNTIME,
new String(key, String::L_TAINTED),	new String(key, String::L_TAINTED),
"is %s, "HTTP_FORM_NAME" option value can be string or table only (file is allowed for $."HTTP_METHOD_NAME"[POST] + $."HTTP_FORM_ENCTYPE_NAME"["HTTP_CONTENT_TYPE_MULTIPART_FORMDATA"])", value->type());	"is %s, " HTTP_FORM_NAME " option value can be string or table only (file is allowed for $." HTTP_METHOD_NAME "[POST] + $." HTTP_FORM_ENCTYPE_NAME "[" HTTP_CONTENT_TYPE_MULTIPART_FORMDATA "])", value->type());
}	}

const char* pa_form2string(HashStringValue& form, Request_charsets& charsets) {	const char* pa_form2string(HashStringValue& form, Request_charsets& charsets) {
Line 495 static void form_value2part(	Line 510 static void form_value2part(
} else	} else
throw Exception(PARSER_RUNTIME,	throw Exception(PARSER_RUNTIME,
new String(key, String::L_TAINTED),	new String(key, String::L_TAINTED),
"is %s, "HTTP_FORM_NAME" option value can be string, table or file only", value->type());	"is %s, " HTTP_FORM_NAME " option value can be string, table or file only", value->type());
}	}

const char* pa_form2string_multipart(HashStringValue& form, Request& r, const char* boundary, size_t& post_size){	const char* pa_form2string_multipart(HashStringValue& form, Request& r, const char* boundary, size_t& post_size){
Line 525 static void find_headers_end(char* p,	Line 540 static void find_headers_end(char* p,
headers_end_at=0;	headers_end_at=0;
}	}

	// Set-Cookie: name=value; Domain=docs.foo.com; Path=/accounts; Expires=Wed, 13-Jan-2021 22:23:01 GMT; Secure; HttpOnly
	static ArrayString* parse_cookie(Request& r, const String& cookie) {
	char *current=pa_strdup(cookie.cstr());

	const String* name=0;
	const String* value=&String::Empty;
	const String* expires=&String::Empty;
	const String* max_age=&String::Empty;
	const String* path=&String::Empty;
	const String* domain=&String::Empty;
	const String* httponly=&String::Empty;
	const String* secure=&String::Empty;

	bool first_pair=true;

	do {
	if(char *meaning=search_stop(current, ';'))
	if(char *attribute=search_stop(meaning, '=')) {
	const String* sname=new String(unescape_chars(attribute, strlen(attribute), &r.charsets.source(), true/don't convert '"' to space/), String::L_TAINTED);
	const String* smeaning=0;
	if(meaning)
	smeaning=new String(unescape_chars(meaning, strlen(meaning), &r.charsets.source(), true/don't convert '"' to space/), String::L_TAINTED);

	if(first_pair) {
	// name + value
	name=sname;
	value=smeaning;
	first_pair=false;
	} else {
	const String& slower=sname->change_case(r.charsets.source(), String::CC_LOWER);

	if(slower == "expires")
	expires=smeaning;
	else if(slower == "max-age")
	max_age=smeaning;
	else if(slower == "domain")
	domain=smeaning;
	else if(slower == "path")
	path=smeaning;
	else if(slower == "httponly")
	httponly=new String("1", String::L_CLEAN);
	else if(slower == "secure")
	secure=new String("1", String::L_CLEAN);
	else {
	// todo@ ?
	}
	}
	}
	} while(current);

	if(!name)
	return 0;

	ArrayString* result=new ArrayString(8);
	*result+=name;
	*result+=value;
	*result+=expires;
	*result+=max_age;
	*result+=domain;
	*result+=path;
	*result+=httponly;
	*result+=secure;

	return result;
	}

	Table* parse_cookies(Request& r, Table *cookies){
	Table& result=*new Table(new Cookies_table_template_columns);

	for(Array_iterator<Table::element_type> i(*cookies); i.has_next(); )
	if(ArrayString* row=parse_cookie(r, *i.next()->get(0)))
	result+=row;

	return &result;
	}

/// @todo build .cookies field. use ^file.tables.SET-COOKIES.menu{ for now	/// @todo build .cookies field. use ^file.tables.SET-COOKIES.menu{ for now
File_read_http_result pa_internal_file_read_http(Request& r,	File_read_http_result pa_internal_file_read_http(Request& r,
const String& file_spec,	const String& file_spec,
Line 545 File_read_http_result pa_internal_file_r	Line 636 File_read_http_result pa_internal_file_r
Value* vcookies=0;	Value* vcookies=0;
Value* vbody=0;	Value* vbody=0;
Charset *asked_remote_charset=0;	Charset *asked_remote_charset=0;
	Charset* real_remote_charset=0;
const char* user_cstr=0;	const char* user_cstr=0;
const char* password_cstr=0;	const char* password_cstr=0;
const char* encode=0;	const char* encode=0;
Line 588 File_read_http_result pa_internal_file_r	Line 680 File_read_http_result pa_internal_file_r
omit_post_charset=vomit_post_charset->as_bool();	omit_post_charset=vomit_post_charset->as_bool();
}	}
if(Value* vcharset_name=options->get(PA_CHARSET_NAME)) {	if(Value* vcharset_name=options->get(PA_CHARSET_NAME)) {
asked_remote_charset=&charsets.get(vcharset_name->as_string().	asked_remote_charset=&charsets.get(vcharset_name->as_string().change_case(r.charsets.source(), String::CC_UPPER));
change_case(r.charsets.source(), String::CC_UPPER));	}
	if(Value* vresponse_charset_name=options->get(PA_RESPONSE_CHARSET_NAME)) {
	valid_options++;
	real_remote_charset=&charsets.get(vresponse_charset_name->as_string().change_case(r.charsets.source(), String::CC_UPPER));
}	}
if(Value* vuser=options->get(HTTP_USER)) {	if(Value* vuser=options->get(HTTP_USER)) {
valid_options++;	valid_options++;
Line 610 File_read_http_result pa_internal_file_r	Line 705 File_read_http_result pa_internal_file_r
if(method_is_get)	if(method_is_get)
throw Exception(PARSER_RUNTIME,	throw Exception(PARSER_RUNTIME,
0,	0,
"you can not use $."HTTP_FORM_ENCTYPE_NAME" option with method GET");	"you can not use $." HTTP_FORM_ENCTYPE_NAME " option with method GET");

multipart=strcasecmp(encode, HTTP_CONTENT_TYPE_MULTIPART_FORMDATA)==0;	multipart=strcasecmp(encode, HTTP_CONTENT_TYPE_MULTIPART_FORMDATA)==0;

if(!multipart && strcasecmp(encode, HTTP_CONTENT_TYPE_FORM_URLENCODED)!=0)	if(!multipart && strcasecmp(encode, HTTP_CONTENT_TYPE_FORM_URLENCODED)!=0)
throw Exception(PARSER_RUNTIME,	throw Exception(PARSER_RUNTIME,
0,	0,
"$."HTTP_FORM_ENCTYPE_NAME" option value can be "HTTP_CONTENT_TYPE_FORM_URLENCODED" or "HTTP_CONTENT_TYPE_MULTIPART_FORMDATA" only");	"$." HTTP_FORM_ENCTYPE_NAME " option value can be " HTTP_CONTENT_TYPE_FORM_URLENCODED " or " HTTP_CONTENT_TYPE_MULTIPART_FORMDATA " only");
}	}

if(vbody){	if(vbody){
if(method_is_get)	if(method_is_get)
throw Exception(PARSER_RUNTIME,	throw Exception(PARSER_RUNTIME,
0,	0,
"you can not use $."HTTP_BODY_NAME" option with method GET");	"you can not use $." HTTP_BODY_NAME " option with method GET");

if(form)	if(form)
throw Exception(PARSER_RUNTIME,	throw Exception(PARSER_RUNTIME,
0,	0,
"you can not use options $."HTTP_BODY_NAME" and $."HTTP_FORM_NAME" together");	"you can not use options $." HTTP_BODY_NAME " and $." HTTP_FORM_NAME " together");
}	}

//preparing request	//preparing request
Line 789 File_read_http_result pa_internal_file_r	Line 884 File_read_http_result pa_internal_file_r

// sending request	// sending request
int status_code=http_request(response, response_size,	int status_code=http_request(response, response_size,
host, port, request, request_size,	pa_idna_encode(host, r.charsets.source()), port, request, request_size,
timeout_secs, fail_on_status_ne_200);	timeout_secs, fail_on_status_ne_200);

// processing results	// processing results
Line 803 File_read_http_result pa_internal_file_r	Line 898 File_read_http_result pa_internal_file_r
result.headers=new HashStringValue;	result.headers=new HashStringValue;
VHash* vtables=new VHash;	VHash* vtables=new VHash;
result.headers->put(HTTP_TABLES_NAME, vtables);	result.headers->put(HTTP_TABLES_NAME, vtables);
Charset* real_remote_charset=0; // undetected, yet

if(headers_end_at) {	if(headers_end_at) {
*headers_end_at=0;	*headers_end_at=0;
Line 826 File_read_http_result pa_internal_file_r	Line 920 File_read_http_result pa_internal_file_r
"bad response from host - bad header \"%s\"", line.cstr());	"bad response from host - bad header \"%s\"", line.cstr());
const String::Body HEADER_NAME=line.mid(0, pos).change_case(r.charsets.source(), String::CC_UPPER);	const String::Body HEADER_NAME=line.mid(0, pos).change_case(r.charsets.source(), String::CC_UPPER);
const String& HEADER_VALUE=line.mid(pos+1, line.length()).trim(String::TRIM_BOTH, " \t\r");	const String& HEADER_VALUE=line.mid(pos+1, line.length()).trim(String::TRIM_BOTH, " \t\r");
if(as_text && HEADER_NAME==HTTP_CONTENT_TYPE_UPPER)	if(as_text && HEADER_NAME==HTTP_CONTENT_TYPE_UPPER && !real_remote_charset)
real_remote_charset=detect_charset(HEADER_VALUE.cstr());	real_remote_charset=detect_charset(HEADER_VALUE.cstr());

// tables	// tables
Line 854 File_read_http_result pa_internal_file_r	Line 948 File_read_http_result pa_internal_file_r

result.headers->put(HEADER_NAME, new VString(HEADER_VALUE));	result.headers->put(HEADER_NAME, new VString(HEADER_VALUE));
}	}

	// filling $.cookies
	if(Value vcookies=(Value )tables.get("SET-COOKIE"))
	result.headers->put(HTTP_COOKIES_NAME, new VTable(parse_cookies(r, vcookies->get_table())));
}	}

if(as_text && raw_body_size>=3 && strncmp(raw_body, "\xEF\xBB\xBF", 3)==0){	if(as_text && raw_body_size>=3 && strncmp(raw_body, "\xEF\xBB\xBF", 3)==0){

E-mail:

Removed from v.1.50
changed lines
	Added in v.1.65