|
|
| version 1.23, 2009/01/25 06:12:04 | version 1.31, 2009/07/15 13:00:24 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: http support functions. | Parser: http support functions. |
| Copyright(c) 2001-2005 ArtLebedev Group (http://www.artlebedev.com) | Copyright(c) 2001-2009 ArtLebedev Group (http://www.artlebedev.com) |
| Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru) | Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru) |
| */ | */ |
| Line 41 static const char * const IDENT_HTTP_C=" | Line 41 static const char * const IDENT_HTTP_C=" |
| #undef CRLF | #undef CRLF |
| #define CRLF "\r\n" | #define CRLF "\r\n" |
| #define DCRLF "\r\n\r\n" | |
| 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)); |
| Line 340 static void http_pass_cookie(HashStringV | Line 339 static void http_pass_cookie(HashStringV |
| Http_pass_header_info *info) { | Http_pass_header_info *info) { |
| *info->request << String(name, String::L_HTTP_COOKIE) << "=" | *info->request << String(name, String::L_HTTP_COOKIE) << "=" |
| << attributed_meaning_to_string(*value, String::L_HTTP_COOKIE, false) | << attributed_meaning_to_string(*value, String::L_HTTP_COOKIE, true) |
| << "; "; | << "; "; |
| } | } |
| Line 366 static void form_string_value2string( | Line 365 static void form_string_value2string( |
| const String& value, | const String& value, |
| String& result) | String& result) |
| { | { |
| result << String(key, String::L_URI) << "="; | result << String(key, String::L_URI) << "=" << String(value, String::L_URI) << "&"; |
| result.append(value, String::L_URI, true); | |
| result << "&"; | |
| } | } |
| #ifndef DOXYGEN | #ifndef DOXYGEN |
| Line 402 static void form_value2string( | Line 399 static void form_value2string( |
| const char* pa_form2string(HashStringValue& form, Request_charsets& charsets) { | const char* pa_form2string(HashStringValue& form, Request_charsets& charsets) { |
| String string; | String string; |
| form.for_each<String*>(form_value2string, &string); | form.for_each<String*>(form_value2string, &string); |
| return string.cstr(String::L_UNSPECIFIED, 0, &charsets); | return string.untaint_cstr(String::L_AS_IS, 0, &charsets); |
| } | } |
| struct FormPart { | struct FormPart { |
| Line 412 struct FormPart { | Line 409 struct FormPart { |
| Form_table_value2string_info* info; | Form_table_value2string_info* info; |
| }; | }; |
| static void form_part_boundary_header(FormPart& part, String name, const char* file_name=0){ | static void form_part_boundary_header(FormPart& part, String::Body name, const char* file_name=0){ |
| part.string << "--" << part.boundary; | part.string << "--" << part.boundary |
| part.string << CRLF HTTP_CONTENT_DISPOSITION ": form-data; name=\"" << name << "\""; | << CRLF HTTP_CONTENT_DISPOSITION ": form-data; name=\"" |
| << Charset::transcode(name, part.r->charsets.source(), part.r->charsets.client()) | |
| << "\""; | |
| if(file_name){ | if(file_name){ |
| if(strcmp(file_name, NONAME_DAT)!=0) | if(strcmp(file_name, NONAME_DAT)!=0) |
| part.string << "; filename=\"" << file_name << "\""; | part.string << "; filename=\"" << file_name << "\""; |
| part.string << CRLF HTTP_CONTENT_TYPE ": " << part.r->mime_type_of(file_name); | part.string << CRLF HTTP_CONTENT_TYPE ": " << part.r->mime_type_of(file_name); |
| } | } |
| part.string << DCRLF; | part.string << CRLF CRLF; |
| } | } |
| static void form_string_value2part( | static void form_string_value2part( |
| HashStringValue::key_type key, | HashStringValue::key_type key, |
| const String& value, | const String& value, |
| FormPart& part) | FormPart& part) |
| { | { |
| form_part_boundary_header(part, String(key, String::L_URI)); | form_part_boundary_header(part, key); |
| part.string.append(value, String::L_AS_IS, true); | part.string << Charset::transcode(value, part.r->charsets.source(), part.r->charsets.client()) << CRLF; |
| part.string << CRLF; | |
| } | } |
| static void form_file_value2part( | static void form_file_value2part( |
| HashStringValue::key_type key, | HashStringValue::key_type key, |
| VFile& vfile, | VFile& vfile, |
| FormPart& part) | FormPart& part) |
| { | { |
| form_part_boundary_header(part, String(key, String::L_URI), vfile.fields().get(name_name)->as_string().cstr()); | form_part_boundary_header(part, key, vfile.fields().get(name_name)->as_string().cstr()); |
| part.string.append_know_length(vfile.value_ptr(), vfile.value_size(), String::L_FILE_POST); | part.string.append_know_length(vfile.value_ptr(), vfile.value_size(), String::L_FILE_POST); |
| part.string << CRLF; | part.string << CRLF; |
| } | } |
| Line 448 static void form_table_value2part(Table: | Line 446 static void form_table_value2part(Table: |
| } | } |
| static void form_value2part( | static void form_value2part( |
| HashStringValue::key_type key, | HashStringValue::key_type key, |
| HashStringValue::value_type value, | HashStringValue::value_type value, |
| FormPart& part) | FormPart& part) |
| { | { |
| if(const String* svalue=value->get_string()) | if(const String* svalue=value->get_string()) |
| form_string_value2part(key, *svalue, part); | form_string_value2part(key, *svalue, part); |
| Line 472 const char* pa_form2string_multipart(Has | Line 470 const char* pa_form2string_multipart(Has |
| formpart.boundary=boundary; | formpart.boundary=boundary; |
| formpart.info=NULL; | formpart.info=NULL; |
| form.for_each<FormPart&>(form_value2part, formpart); | form.for_each<FormPart&>(form_value2part, formpart); |
| formpart.string << CRLF << "--" << boundary << "--"; | formpart.string << "--" << boundary << "--"; |
| post_size=formpart.string.length(); | post_size=formpart.string.length(); // very surprizing, but it calculates correct post_size even with binary files! |
| return formpart.string.cstr(String::L_UNSPECIFIED, 0, &(r.charsets)); | return formpart.string.untaint_cstr(String::L_AS_IS); // without transcoding |
| } | } |
| static void find_headers_end(char* p, | static void find_headers_end(char* p, |
| Line 607 File_read_http_result pa_internal_file_r | Line 605 File_read_http_result pa_internal_file_r |
| } | } |
| //preparing request | //preparing request |
| String& connect_string=*new String; | String& connect_string=*new String(file_spec); |
| // not in ^sql{... L_SQL ...} spirit, but closer to ^file::load one | |
| connect_string.append(file_spec, String::L_URI); // tainted pieces -> URI pieces | |
| String request_head_and_body; | String request_head_and_body; |
| { | { |
| // influence URLencoding of tainted pieces to String::L_URI lang | // influence URLencoding of tainted pieces to String::L_URI lang |
| Temp_client_charset temp(r.charsets, *asked_remote_charset); | Temp_client_charset temp(r.charsets, *asked_remote_charset); |
| const char* connect_string_cstr=connect_string.cstr(String::L_UNSPECIFIED, 0, &(r.charsets)); | const char* connect_string_cstr=connect_string.untaint_cstr(String::L_URI, 0, &(r.charsets)); |
| const char* current=connect_string_cstr; | const char* current=connect_string_cstr; |
| if(strncmp(current, "http://", 7)!=0) | if(strncmp(current, "http://", 7)!=0) |
| Line 632 File_read_http_result pa_internal_file_r | Line 628 File_read_http_result pa_internal_file_r |
| char* error_pos=0; | char* error_pos=0; |
| port=port_cstr?(short)strtol(port_cstr, &error_pos, 0):80; | port=port_cstr?(short)strtol(port_cstr, &error_pos, 0):80; |
| bool uri_has_query_string=strchr(uri, '?')!=0; | |
| // making request head | // making request head |
| String head; | String head; |
| head << method << " " << uri; | head << method << " " << uri; |
| if(form && method_is_get) | if(method_is_get && form) |
| head << (uri_has_query_string?"&":"?") << pa_form2string(*form, r.charsets); | head << (strchr(uri, '?')!=0?"&":"?") << pa_form2string(*form, r.charsets); |
| head <<" HTTP/1.0" CRLF "host: "<< host << CRLF; | head <<" HTTP/1.0" CRLF "host: "<< host << CRLF; |
| char* boundary; | char* boundary=0; |
| if(multipart){ | if(multipart){ |
| uuid uuid=get_uuid(); | uuid uuid=get_uuid(); |
| Line 658 File_read_http_result pa_internal_file_r | Line 652 File_read_http_result pa_internal_file_r |
| size_t post_size=0; | size_t post_size=0; |
| if(form && !method_is_get) { | if(form && !method_is_get) { |
| head << HTTP_CONTENT_TYPE ": "; | head << HTTP_CONTENT_TYPE ": " << (multipart ? HTTP_CONTENT_TYPE_MULTIPART_FORMDATA : HTTP_CONTENT_TYPE_FORM_URLENCODED); |
| if(!omit_post_charset) | |
| head << "; charset=" << asked_remote_charset->NAME_CSTR(); | |
| if(multipart) { | if(multipart) { |
| head << HTTP_CONTENT_TYPE_MULTIPART_FORMDATA "; boundary=" << boundary << CRLF; | head << "; boundary=" << boundary; |
| // !!! charset? | body_cstr=pa_form2string_multipart(*form, r/*charsets & mime_type needed*/, boundary, post_size/*correct post_size returned here*/); |
| body_cstr=pa_form2string_multipart(*form, r, boundary, post_size); | |
| } else { | } else { |
| head << HTTP_CONTENT_TYPE_FORM_URLENCODED; | |
| if(!omit_post_charset) | |
| head << "; charset=" << asked_remote_charset->NAME_CSTR() << ";"; | |
| head << CRLF; | |
| body_cstr=pa_form2string(*form, r.charsets); | body_cstr=pa_form2string(*form, r.charsets); |
| post_size=strlen(body_cstr); | post_size=strlen(body_cstr); |
| } | } |
| head << CRLF; | |
| } else if (vbody) { | } else if (vbody) { |
| body_cstr=vbody->as_string().cstr(String::L_UNSPECIFIED, 0, &(r.charsets)); | // transcode tainted pieces and then URI-encode them |
| // needed for transcoded $.body[] first of all | body_cstr=vbody->as_string().untaint_cstr(String::L_AS_IS, 0, &(r.charsets)); |
| // now transcode is needed only if own content-type was specified _and_ clean chars with code>127 are in the body | |
| // @todo: I don't like the current behaviour | |
| body_cstr=Charset::transcode( | body_cstr=Charset::transcode( |
| String::C(body_cstr, strlen(body_cstr)), | String::C(body_cstr, strlen(body_cstr)), |
| r.charsets.source(), | r.charsets.source(), |
| *asked_remote_charset | *asked_remote_charset |
| ); | ); |
| post_size=strlen(body_cstr); | |
| } | } |
| // http://www.ietf.org/rfc/rfc2617.txt | // http://www.ietf.org/rfc/rfc2617.txt |
| Line 718 File_read_http_result pa_internal_file_r | Line 716 File_read_http_result pa_internal_file_r |
| "cookies param must be hash"); | "cookies param must be hash"); |
| } | } |
| if(body_cstr) { | if(body_cstr) |
| head << "content-length: " << format(post_size, "%u") << CRLF; | head << "content-length: " << format(post_size, "%u") << CRLF; |
| } | |
| const char* head_cstr=head.cstr(String::L_UNSPECIFIED, 0, &(r.charsets)); | |
| // head + end of header | // head + end of header |
| request_head_and_body << head_cstr << CRLF; | request_head_and_body << head.untaint_cstr(String::L_AS_IS, 0, &(r.charsets)) << CRLF; |
| // body | // body |
| if(body_cstr) | if(body_cstr) |
| request_head_and_body << body_cstr; | request_head_and_body << body_cstr; |
| } | } |
| //sending request | const char* request_cstr=request_head_and_body.cstr(); |
| char* response; | size_t request_size=strlen(request_cstr); |
| size_t response_size; | |
| const char* request=request_head_and_body.cstr(); | |
| size_t request_size=strlen(request); | |
| if(multipart) | if(multipart) |
| request_size=file_untaint(request, request_size); | request_size=file_untaint(request_cstr, request_size); |
| char* response; | |
| size_t response_size; | |
| // sending request | |
| int status_code=http_request(response, response_size, | int status_code=http_request(response, response_size, |
| host, port, request, request_size, | host, port, request_cstr, request_size, |
| timeout_secs, fail_on_status_ne_200); | timeout_secs, fail_on_status_ne_200); |
| //processing results | // processing results |
| char* raw_body; size_t raw_body_size; | char* raw_body; size_t raw_body_size; |
| char* headers_end_at; | char* headers_end_at; |
| find_headers_end(response, | find_headers_end(response, |
| Line 761 File_read_http_result pa_internal_file_r | Line 756 File_read_http_result pa_internal_file_r |
| if(headers_end_at) { | if(headers_end_at) { |
| *headers_end_at=0; | *headers_end_at=0; |
| const String header_block(String::C(response, headers_end_at-response), true); | const String header_block(String::C(response, headers_end_at-response), String::L_TAINTED); |
| ArrayString aheaders; | ArrayString aheaders; |
| HashStringValue& tables=vtables->hash(); | HashStringValue& tables=vtables->hash(); |
| Line 769 File_read_http_result pa_internal_file_r | Line 764 File_read_http_result pa_internal_file_r |
| size_t pos_after=0; | size_t pos_after=0; |
| header_block.split(aheaders, pos_after, "\n"); | header_block.split(aheaders, pos_after, "\n"); |
| //processing headers | // processing headers |
| size_t aheaders_count=aheaders.count(); | size_t aheaders_count=aheaders.count(); |
| for(size_t i=1; i<aheaders_count; i++) { | for(size_t i=1; i<aheaders_count; i++) { |
| const String& line=*aheaders.get(i); | const String& line=*aheaders.get(i); |
| Line 781 File_read_http_result pa_internal_file_r | Line 776 File_read_http_result pa_internal_file_r |
| 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=detect_charset(r.charsets.source(), HEADER_VALUE); | real_remote_charset=detect_charset(HEADER_VALUE.cstr(), true/*already uppercased*/); |
| // tables | // tables |
| { | { |