|
|
| version 1.243, 2009/04/22 04:35:48 | version 1.273, 2013/07/22 15:32:23 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: commonly functions. | Parser: commonly functions. |
| Copyright(c) 2001-2009 ArtLebedev Group (http://www.artlebedev.com) | Copyright (c) 2000-2012 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) |
| * BASE64 part | * BASE64 part |
| Line 26 | Line 26 |
| * | * |
| */ | */ |
| static const char * const IDENT_COMMON_C="$Date$"; | |
| #include "pa_common.h" | #include "pa_common.h" |
| #include "pa_exception.h" | #include "pa_exception.h" |
| #include "pa_hash.h" | #include "pa_hash.h" |
| Line 38 static const char * const IDENT_COMMON_C | Line 36 static const char * const IDENT_COMMON_C |
| #include "pcre.h" | #include "pcre.h" |
| #include "pa_request.h" | #include "pa_request.h" |
| #ifdef _MSC_VER | |
| #include <direct.h> | |
| #endif | |
| volatile const char * IDENT_PA_COMMON_C="$Id$" IDENT_PA_COMMON_H IDENT_PA_HASH_H IDENT_PA_ARRAY_H IDENT_PA_STACK_H; | |
| // some maybe-undefined constants | // some maybe-undefined constants |
| #ifndef _O_TEXT | #ifndef _O_TEXT |
| Line 67 const String file_status_name(FILE_STATU | Line 71 const String file_status_name(FILE_STATU |
| // functions | // functions |
| bool capitalized(const char* s){ | |
| bool upper=true; | |
| for(const char* c=s; *c; c++){ | |
| if(*c != (upper ? toupper((unsigned char)*c) : tolower((unsigned char)*c))) | |
| return false; | |
| upper=strchr("-_ ", *c) != 0; | |
| } | |
| return true; | |
| } | |
| const char* capitalize(const char* s){ | |
| if(!s || capitalized(s)) | |
| return s; | |
| char* result=pa_strdup(s); | |
| if(result){ | |
| bool upper=true; | |
| for(char* c=result; *c; c++){ | |
| *c=upper ? (char)toupper((unsigned char)*c) : (char)tolower((unsigned char)*c); | |
| upper=strchr("-_ ", *c) != 0; | |
| } | |
| } | |
| return (const char*)result; | |
| } | |
| void fix_line_breaks(char *str, size_t& length) { | void fix_line_breaks(char *str, size_t& length) { |
| //_asm int 3; | //_asm int 3; |
| const char* const eob=str+length; | const char* const eob=str+length; |
| Line 77 void fix_line_breaks(char *str, size_t& | Line 106 void fix_line_breaks(char *str, size_t& |
| while(char* eol=(char*)memchr(bol, '\r', eob -bol)) { | while(char* eol=(char*)memchr(bol, '\r', eob -bol)) { |
| size_t len=eol-bol; | size_t len=eol-bol; |
| if(dest!=bol) | if(dest!=bol) |
| memcpy(dest, bol, len); | memmove(dest, bol, len); |
| dest+=len; | dest+=len; |
| *dest++='\n'; | *dest++='\n'; |
| Line 89 void fix_line_breaks(char *str, size_t& | Line 118 void fix_line_breaks(char *str, size_t& |
| } | } |
| // last piece without \r | // last piece without \r |
| if(dest!=bol) | if(dest!=bol) |
| memcpy(dest, bol, eob-bol); | memmove(dest, bol, eob-bol); |
| str[length]=0; // terminating | str[length]=0; // terminating |
| } | } |
| char* file_read_text(Request_charsets& charsets, | char* file_read_text(Request_charsets& charsets, const String& file_spec, bool fail_on_read_problem, HashStringValue* params, bool transcode_result) { |
| const String& file_spec, | File_read_result file=file_read(charsets, file_spec, true, params, fail_on_read_problem, 0, 0, 0, transcode_result); |
| bool fail_on_read_problem, | |
| HashStringValue* params, | |
| bool transcode_result) { | |
| File_read_result file= | |
| file_read(charsets, file_spec, true, params, fail_on_read_problem, 0, 0, 0, transcode_result); | |
| return file.success?file.str:0; | return file.success?file.str:0; |
| } | } |
| char* file_load_text(Request& r, | char* file_load_text(Request& r, const String& file_spec, bool fail_on_read_problem, HashStringValue* params, bool transcode_result) { |
| const String& file_spec, | File_read_result file=file_load(r, file_spec, true, params, fail_on_read_problem, 0, 0, 0, transcode_result); |
| bool fail_on_read_problem, | |
| HashStringValue* params, | |
| bool transcode_result) { | |
| File_read_result file= | |
| file_load(r, file_spec, true, params, fail_on_read_problem, 0, 0, 0, transcode_result); | |
| return file.success?file.str:0; | return file.success?file.str:0; |
| } | } |
| Line 135 struct File_read_action_info { | Line 154 struct File_read_action_info { |
| char* buf; size_t offset; size_t count; | char* buf; size_t offset; size_t count; |
| }; | }; |
| #endif | #endif |
| static void file_read_action( | |
| struct stat& finfo, | static void file_read_action(struct stat& finfo, int f, const String& file_spec, const char* /*fname*/, bool as_text, void *context) { |
| int f, | |
| const String& file_spec, const char* /*fname*/, bool as_text, | |
| void *context) { | |
| File_read_action_info& info=*static_cast<File_read_action_info *>(context); | File_read_action_info& info=*static_cast<File_read_action_info *>(context); |
| size_t to_read_size=info.count; | size_t to_read_size=info.count; |
| if(!to_read_size) | if(!to_read_size) |
| to_read_size=(size_t)finfo.st_size; | to_read_size=(size_t)finfo.st_size; |
| assert( !(info.buf && as_text) ); | assert( !(info.buf && as_text) ); |
| if(to_read_size) { | if(to_read_size) { |
| if(info.offset) | if(info.offset) |
| lseek(f, info.offset, SEEK_SET); | lseek(f, info.offset, SEEK_SET); |
| *info.data=info.buf | *info.data=info.buf ? info.buf : (char *)pa_malloc_atomic(to_read_size+1); |
| ? info.buf | ssize_t result=read(f, *info.data, to_read_size); |
| : new(PointerFreeGC) char[to_read_size+(as_text?1:0)]; | if(result<0) |
| *info.data_size=(size_t)read(f, *info.data, to_read_size); | throw Exception("file.read", &file_spec, "read failed: %s (%d)", strerror(errno), errno); |
| *info.data_size=result; | |
| if(ssize_t(*info.data_size)<0 || *info.data_size>to_read_size) | |
| throw Exception("file.read", | |
| &file_spec, | |
| "read failed: actually read %u bytes count not in [0..%u] valid range", | |
| *info.data_size, to_read_size); | |
| } else { // empty file | } else { // empty file |
| // for both, text and binary: for text we need that terminator, for binary we need nonzero pointer to be able to save such files | // for both, text and binary: for text we need that terminator, for binary we need nonzero pointer to be able to save such files |
| *info.data=new(PointerFreeGC) char[1]; | *info.data=(char *)pa_malloc_atomic(1); |
| *(char*)(*info.data)=0; | *(char*)(*info.data)=0; |
| *info.data_size=0; | *info.data_size=0; |
| return; | return; |
| Line 175 File_read_result file_read(Request_chars | Line 186 File_read_result file_read(Request_chars |
| if(params){ | if(params){ |
| int valid_options=pa_get_valid_file_options_count(*params); | int valid_options=pa_get_valid_file_options_count(*params); |
| if(valid_options!=params->count()) | if(valid_options!=params->count()) |
| throw Exception(PARSER_RUNTIME, | throw Exception(PARSER_RUNTIME, 0, CALLED_WITH_INVALID_OPTION); |
| 0, | |
| "invalid option passed"); | |
| } | } |
| File_read_action_info info={&result.str, &result.length, buf, offset, count}; | File_read_action_info info={&result.str, &result.length, buf, offset, count}; |
| Line 188 File_read_result file_read(Request_chars | Line 197 File_read_result file_read(Request_chars |
| if(as_text){ | if(as_text){ |
| if(result.success){ | if(result.success){ |
| Charset* asked_charset=0; | |
| if(result.length>=3 && strncmp(result.str, "\xEF\xBB\xBF", 3)==0){ | if(result.length>=3 && strncmp(result.str, "\xEF\xBB\xBF", 3)==0){ |
| // skip UTF-8 signature (BOM code) | // skip UTF-8 signature (BOM code) |
| result.str+=3; | result.str+=3; |
| result.length-=3; | result.length-=3; |
| asked_charset=&UTF8_charset; | |
| } | } |
| if(result.length && transcode_text_result && params){ // must be checked because transcode returns CONST string in case length==0, which contradicts hacking few lines below | if(params) |
| if(Value* vcharset_name=params->get(PA_CHARSET_NAME)){ | if(Value* vcharset_name=params->get(PA_CHARSET_NAME)) |
| Charset asked_charset=::charsets.get(vcharset_name->as_string(). | asked_charset=&::charsets.get(vcharset_name->as_string().change_case(charsets.source(), String::CC_UPPER)); |
| change_case(charsets.source(), String::CC_UPPER)); | |
| if(result.length && transcode_text_result && asked_charset){ // length must be checked because transcode returns CONST string in case length==0, which contradicts hacking few lines below | |
| String::C body=String::C(result.str, result.length); | |
| body=Charset::transcode(body, *asked_charset, charsets.source()); | |
| String::C body=String::C(result.str, result.length); | result.str=const_cast<char*>(body.str); // hacking a little |
| body=Charset::transcode(body, asked_charset, charsets.source()); | result.length=body.length; |
| result.str=const_cast<char*>(body.str); // hacking a little | |
| result.length=body.length; | |
| } | |
| } | } |
| } | } |
| if(result.length) | if(result.length) |
| Line 239 File_read_result file_load(Request& r, c | Line 249 File_read_result file_load(Request& r, c |
| return result; | return result; |
| } | } |
| #ifdef PA_SAFE_MODE | #ifdef PA_SAFE_MODE |
| void check_safe_mode(struct stat finfo, const String& file_spec, const char* fname) { | void check_safe_mode(struct stat finfo, const String& file_spec, const char* fname) { |
| if(finfo.st_uid/*foreign?*/!=geteuid() | if(finfo.st_uid/*foreign?*/!=geteuid() |
| && finfo.st_gid/*foreign?*/!=getegid()) | && finfo.st_gid/*foreign?*/!=getegid()) |
| throw Exception(PARSER_RUNTIME, | throw Exception(PARSER_RUNTIME, |
| Line 252 void check_safe_mode(struct stat finfo, | Line 263 void check_safe_mode(struct stat finfo, |
| "fuid(%d)!=euid(%d) or fgid(%d)!=egid(%d)", | "fuid(%d)!=euid(%d) or fgid(%d)!=egid(%d)", |
| fname, | fname, |
| finfo.st_uid, geteuid(), | finfo.st_uid, geteuid(), |
| finfo.st_gid, getegid()); | finfo.st_gid, getegid()); |
| } | } |
| #endif | #else |
| void check_safe_mode(struct stat, const String&, const char*) { | |
| } | |
| #endif | |
| bool file_read_action_under_lock(const String& file_spec, | bool file_read_action_under_lock(const String& file_spec, |
| const char* action_name, File_read_action action, void *context, | const char* action_name, File_read_action action, void *context, |
| bool as_text, | bool as_text, |
| bool fail_on_read_problem) { | bool fail_on_read_problem) { |
| const char* fname=file_spec.cstr(String::L_FILE_SPEC); | const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); |
| int f; | int f; |
| // first open, next stat: | // first open, next stat: |
| Line 281 bool file_read_action_under_lock(const S | Line 297 bool file_read_action_under_lock(const S |
| strerror(errno), errno, fname); | strerror(errno), errno, fname); |
| struct stat finfo; | struct stat finfo; |
| if(stat(fname, &finfo)!=0) | if(fstat(f, &finfo)!=0) |
| throw Exception("file.missing", // hardly possible: we just opened it OK | throw Exception("file.missing", // hardly possible: we just opened it OK |
| &file_spec, | &file_spec, |
| "stat failed: %s (%d), actual filename '%s'", | "stat failed: %s (%d), actual filename '%s'", |
| strerror(errno), errno, fname); | strerror(errno), errno, fname); |
| #ifdef PA_SAFE_MODE | |
| check_safe_mode(finfo, file_spec, fname); | check_safe_mode(finfo, file_spec, fname); |
| #endif | |
| action(finfo, f, file_spec, fname, as_text, context); | action(finfo, f, file_spec, fname, as_text, context); |
| } catch(...) { | } catch(...) { |
| Line 303 bool file_read_action_under_lock(const S | Line 317 bool file_read_action_under_lock(const S |
| return true; | return true; |
| } else { | } else { |
| if(fail_on_read_problem) | if(fail_on_read_problem) |
| throw Exception(errno==EACCES?"file.access":errno==ENOENT?"file.missing":0, | throw Exception(errno==EACCES?"file.access" |
| :(errno==ENOENT || errno==ENOTDIR || errno==ENODEV)?"file.missing":0, | |
| &file_spec, | &file_spec, |
| "%s failed: %s (%d), actual filename '%s'", | "%s failed: %s (%d), actual filename '%s'", |
| action_name, strerror(errno), errno, fname); | action_name, strerror(errno), errno, fname); |
| Line 315 void create_dir_for_file(const String& f | Line 330 void create_dir_for_file(const String& f |
| size_t pos_after=1; | size_t pos_after=1; |
| size_t pos_before; | size_t pos_before; |
| while((pos_before=file_spec.pos('/', pos_after))!=STRING_NOT_FOUND) { | while((pos_before=file_spec.pos('/', pos_after))!=STRING_NOT_FOUND) { |
| mkdir(file_spec.mid(0, pos_before).cstr(String::L_FILE_SPEC), 0775); | mkdir(file_spec.mid(0, pos_before).taint_cstr(String::L_FILE_SPEC), 0775); |
| pos_after=pos_before+1; | pos_after=pos_before+1; |
| } | } |
| } | } |
| Line 329 bool file_write_action_under_lock( | Line 344 bool file_write_action_under_lock( |
| bool do_append, | bool do_append, |
| bool do_block, | bool do_block, |
| bool fail_on_lock_problem) { | bool fail_on_lock_problem) { |
| const char* fname=file_spec.cstr(String::L_FILE_SPEC); | const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); |
| int f; | int f; |
| if(access(fname, W_OK)!=0) // no | if(access(fname, W_OK)!=0) // no |
| create_dir_for_file(file_spec); | create_dir_for_file(file_spec); |
| Line 350 bool file_write_action_under_lock( | Line 365 bool file_write_action_under_lock( |
| } | } |
| try { | try { |
| action(f, context); | #if (defined(HAVE_FCHMOD) && defined(PA_SAFE_MODE)) |
| struct stat finfo; | |
| if(fstat(f, &finfo)==0 && finfo.st_mode & 0111) | |
| fchmod(f, finfo.st_mode & 0666/*clear executable bits*/); // backward: ignore errors if any | |
| #endif | |
| action(f, context); | |
| } catch(...) { | } catch(...) { |
| #ifdef HAVE_FTRUNCATE | #ifdef HAVE_FTRUNCATE |
| if(!do_append) | if(!do_append) |
| Line 376 bool file_write_action_under_lock( | Line 396 bool file_write_action_under_lock( |
| #ifndef DOXYGEN | #ifndef DOXYGEN |
| struct File_write_action_info { | struct File_write_action_info { |
| const char* str; size_t length; | const char* str; |
| size_t length; | |
| }; | }; |
| #endif | #endif |
| static void file_write_action(int f, void *context) { | static void file_write_action(int f, void *context) { |
| File_write_action_info& info=*static_cast<File_write_action_info *>(context); | File_write_action_info& info=*static_cast<File_write_action_info *>(context); |
| if(info.length) { | if(info.length) { |
| int written=write(f, info.str, info.length); | ssize_t written=write(f, info.str, info.length); |
| if(written<0) | if(written<0) |
| throw Exception("file.access", | throw Exception("file.write", 0, "write failed: %s (%d)", strerror(errno), errno); |
| 0, | if(written!=info.length) |
| "write failed: %s (%d)", strerror(errno), errno); | throw Exception("file.write", 0, "write failed: %u of %u bytes written", written, info.length); |
| } | } |
| } | } |
| void file_write( | void file_write( |
| const String& file_spec, | Request_charsets& charsets, |
| const char* data, size_t size, | const String& file_spec, |
| const char* data, | |
| size_t size, | |
| bool as_text, | bool as_text, |
| bool do_append) { | bool do_append, |
| Charset* asked_charset) { | |
| if(as_text && asked_charset){ | |
| String::C body=String::C(data, size); | |
| body=Charset::transcode(body, charsets.source(), *asked_charset); | |
| data=body.str; | |
| size=body.length; | |
| }; | |
| File_write_action_info info={data, size}; | File_write_action_info info={data, size}; |
| file_write_action_under_lock( | file_write_action_under_lock( |
| Line 405 void file_write( | Line 439 void file_write( |
| do_append); | do_append); |
| } | } |
| static size_t get_dir(char* fname, size_t helper_length){ | |
| bool dir=false; | |
| size_t pos=0; | |
| for(pos=helper_length; pos; pos--){ | |
| char c=fname[pos-1]; | |
| if(c=='/' || c=='\\'){ | |
| fname[pos-1]=0; | |
| dir=true; | |
| } else if(dir) break; | |
| } | |
| return pos; | |
| } | |
| static bool entry_readable(char* fname, bool need_dir) { | |
| if(need_dir){ | |
| size_t size=strlen(fname); | |
| while(size) { | |
| char c=fname[size-1]; | |
| if(c=='/' || c=='\\') | |
| fname[--size]=0; | |
| else | |
| break; | |
| } | |
| } | |
| struct stat finfo; | |
| if(access(fname, R_OK)==0 && entry_exists(fname, &finfo)) { | |
| bool is_dir=(finfo.st_mode&S_IFDIR) != 0; | |
| return is_dir==need_dir; | |
| } | |
| return false; | |
| } | |
| static bool entry_readable(const String& file_spec, bool need_dir) { | |
| return entry_readable(file_spec.taint_cstrm(String::L_FILE_SPEC), need_dir); | |
| } | |
| // throws nothing! [this is required in file_move & file_delete] | // throws nothing! [this is required in file_move & file_delete] |
| static void rmdir(const String& file_spec, size_t pos_after) { | static void rmdir(const String& file_spec, size_t pos_after=0) { |
| size_t pos_before; | char* dir_spec=file_spec.taint_cstrm(String::L_FILE_SPEC); |
| if((pos_before=file_spec.pos('/', pos_after))!=STRING_NOT_FOUND) | size_t length=strlen(dir_spec); |
| rmdir(file_spec, pos_before+1); | while( (length=get_dir(dir_spec, length)) && (length > pos_after) ){ |
| #ifdef WIN32 | |
| rmdir(file_spec.mid(0, pos_after-1/* / */).cstr(String::L_FILE_SPEC)); | if(!entry_readable(dir_spec, true)) |
| break; | |
| DWORD attrs=GetFileAttributes(dir_spec); | |
| if( | |
| (attrs==INVALID_FILE_ATTRIBUTES) | |
| || !(attrs & FILE_ATTRIBUTE_DIRECTORY) | |
| || (attrs & FILE_ATTRIBUTE_REPARSE_POINT) | |
| ) | |
| break; | |
| #endif | |
| if( rmdir(dir_spec) ) | |
| break; | |
| }; | |
| } | } |
| bool file_delete(const String& file_spec, bool fail_on_problem) { | bool file_delete(const String& file_spec, bool fail_on_problem, bool keep_empty_dirs) { |
| const char* fname=file_spec.cstr(String::L_FILE_SPEC); | const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); |
| if(unlink(fname)!=0) | if(unlink(fname)!=0) |
| if(fail_on_problem) | if(fail_on_problem) |
| throw Exception(errno==EACCES?"file.access":errno==ENOENT?"file.missing":0, | throw Exception(errno==EACCES?"file.access":errno==ENOENT?"file.missing":0, |
| Line 425 bool file_delete(const String& file_spec | Line 508 bool file_delete(const String& file_spec |
| else | else |
| return false; | return false; |
| rmdir(file_spec, 1); | if(!keep_empty_dirs) |
| rmdir(file_spec, 1); | |
| return true; | return true; |
| } | } |
| void file_move(const String& old_spec, const String& new_spec) { | void file_move(const String& old_spec, const String& new_spec, bool keep_empty_dirs) { |
| const char* old_spec_cstr=old_spec.cstr(String::L_FILE_SPEC); | const char* old_spec_cstr=old_spec.taint_cstr(String::L_FILE_SPEC); |
| const char* new_spec_cstr=new_spec.cstr(String::L_FILE_SPEC); | const char* new_spec_cstr=new_spec.taint_cstr(String::L_FILE_SPEC); |
| create_dir_for_file(new_spec); | create_dir_for_file(new_spec); |
| Line 441 void file_move(const String& old_spec, c | Line 526 void file_move(const String& old_spec, c |
| "rename failed: %s (%d), actual filename '%s' to '%s'", | "rename failed: %s (%d), actual filename '%s' to '%s'", |
| strerror(errno), errno, old_spec_cstr, new_spec_cstr); | strerror(errno), errno, old_spec_cstr, new_spec_cstr); |
| rmdir(old_spec, 1); | if(!keep_empty_dirs) |
| rmdir(old_spec, 1); | |
| } | } |
| Line 454 bool entry_exists(const char* fname, str | Line 540 bool entry_exists(const char* fname, str |
| } | } |
| bool entry_exists(const String& file_spec) { | bool entry_exists(const String& file_spec) { |
| const char* fname=file_spec.cstr(String::L_FILE_SPEC); | const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); |
| return entry_exists(fname, 0); | return entry_exists(fname, 0); |
| } | } |
| static bool entry_readable(const String& file_spec, bool need_dir) { | |
| char* fname=file_spec.cstrm(String::L_FILE_SPEC); | |
| if(need_dir) { | |
| size_t size=strlen(fname); | |
| while(size) { | |
| char c=fname[size-1]; | |
| if(c=='/' || c=='\\') | |
| fname[--size]=0; | |
| else | |
| break; | |
| } | |
| } | |
| struct stat finfo; | |
| if(access(fname, R_OK)==0 && entry_exists(fname, &finfo)) { | |
| bool is_dir=(finfo.st_mode&S_IFDIR) != 0; | |
| return is_dir==need_dir; | |
| } | |
| return false; | |
| } | |
| bool file_exist(const String& file_spec) { | bool file_exist(const String& file_spec) { |
| return entry_readable(file_spec, false); | return entry_readable(file_spec, false); |
| } | } |
| Line 488 bool dir_exists(const String& file_spec) | Line 554 bool dir_exists(const String& file_spec) |
| const String* file_exist(const String& path, const String& name) { | const String* file_exist(const String& path, const String& name) { |
| String& result=*new String(path); | String& result=*new String(path); |
| result << "/"; | if(path.last_char() != '/') |
| result << "/"; | |
| result << name; | result << name; |
| return file_exist(result)?&result:0; | return file_exist(result)?&result:0; |
| } | } |
| bool file_executable(const String& file_spec) { | bool file_executable(const String& file_spec) { |
| return access(file_spec.cstr(String::L_FILE_SPEC), X_OK)==0; | return access(file_spec.taint_cstr(String::L_FILE_SPEC), X_OK)==0; |
| } | } |
| bool file_stat(const String& file_spec, | bool file_stat(const String& file_spec, |
| Line 503 bool file_stat(const String& file_spec, | Line 570 bool file_stat(const String& file_spec, |
| time_t& rmtime, | time_t& rmtime, |
| time_t& rctime, | time_t& rctime, |
| bool fail_on_read_problem) { | bool fail_on_read_problem) { |
| const char* fname=file_spec.cstr(String::L_FILE_SPEC); | const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); |
| struct stat finfo; | struct stat finfo; |
| if(stat(fname, &finfo)!=0) | if(stat(fname, &finfo)!=0) |
| if(fail_on_read_problem) | if(fail_on_read_problem) |
| Line 571 enum FormatType { | Line 638 enum FormatType { |
| FormatUInt, | FormatUInt, |
| FormatDouble | FormatDouble |
| }; | }; |
| FormatType format_type(char* fmt){ | FormatType format_type(const char* fmt){ |
| enum FormatState { | enum FormatState { |
| Percent, | Percent, |
| Flags, | Flags, |
| Line 582 FormatType format_type(char* fmt){ | Line 649 FormatType format_type(char* fmt){ |
| FormatType result=FormatInvalid; | FormatType result=FormatInvalid; |
| char* pos=fmt; | const char* pos=fmt; |
| while(char c=*(pos++)){ | while(char c=*(pos++)){ |
| switch(state){ | switch(state){ |
| case Percent: | case Percent: |
| Line 626 FormatType format_type(char* fmt){ | Line 693 FormatType format_type(char* fmt){ |
| } | } |
| const char* format(double value, char* fmt) { | const char* format(double value, const char* fmt) { |
| char local_buf[MAX_NUMBER]; | char local_buf[MAX_NUMBER]; |
| int size=-1; | int size=-1; |
| Line 683 enum EscapeState { | Line 750 enum EscapeState { |
| }; | }; |
| // @todo prescan for reduce required size (unescaped sting in 1 byte charset requires less memory usually) | // @todo prescan for reduce required size (unescaped sting in 1 byte charset requires less memory usually) |
| char* unescape_chars(const char* cp, int len, Charset* charset, bool ignore_plus){ | char* unescape_chars(const char* cp, int len, Charset* charset, bool js){ |
| char* s=new(PointerFreeGC) char[len+1]; // must be enough (%uXXXX==6 bytes, max utf-8 char length==6 bytes) | char* s=new(PointerFreeGC) char[len+1]; // must be enough (%uXXXX==6 bytes, max utf-8 char length==6 bytes) |
| char* dst=s; | char* dst=s; |
| EscapeState escapeState=EscapeRest; | EscapeState escapeState=EscapeRest; |
| Line 692 char* unescape_chars(const char* cp, int | Line 759 char* unescape_chars(const char* cp, int |
| short int jsCnt=0; | short int jsCnt=0; |
| while(srcPos<len){ | while(srcPos<len){ |
| uchar c=(uchar)cp[srcPos]; | uchar c=(uchar)cp[srcPos]; |
| if(c=='%'){ | if(c=='%' || (c=='\\' && js)){ |
| escapeState=EscapeFirst; | escapeState=EscapeFirst; |
| } else { | } else { |
| switch(escapeState) { | switch(escapeState) { |
| case EscapeRest: | case EscapeRest: |
| if(!ignore_plus && c=='+'){ | if(!js && c=='+'){ |
| *dst++=' '; | *dst++=' '; |
| } else { | } else { |
| *dst++=c; | *dst++=c; |
| Line 749 char* unescape_chars(const char* cp, int | Line 816 char* unescape_chars(const char* cp, int |
| return s; | return s; |
| } | } |
| char *search_stop(char*& current, char cstop_at) { | |
| // sanity check | |
| if(!current) | |
| return 0; | |
| // skip leading WS | |
| while(*current==' ' || *current=='\t') | |
| current++; | |
| if(!*current) | |
| return current=0; | |
| char *result=current; | |
| if(char *pstop_at=strchr(current, cstop_at)) { | |
| *pstop_at=0; | |
| current=pstop_at+1; | |
| } else | |
| current=0; | |
| return result; | |
| } | |
| #ifdef WIN32 | #ifdef WIN32 |
| void back_slashes_to_slashes(char* s) { | void back_slashes_to_slashes(char* s) { |
| if(s) | if(s) |
| Line 796 size_t strpos(const char *str, const cha | Line 883 size_t strpos(const char *str, const cha |
| // content-type: xxx; charset="WE-NEED-THIS"; | // content-type: xxx; charset="WE-NEED-THIS"; |
| Charset* detect_charset(const char* content_type){ | Charset* detect_charset(const char* content_type){ |
| if(content_type){ | if(content_type){ |
| size_t len=strlen(content_type); | char* CONTENT_TYPE=pa_strdup(content_type); |
| char* CONTENT_TYPE=new(PointerFreeGC) char[len+1]; | |
| memcpy(CONTENT_TYPE, content_type, len); | |
| for(char *p=CONTENT_TYPE; *p; p++) | for(char *p=CONTENT_TYPE; *p; p++) |
| *p=(char)toupper((unsigned char)*p); | *p=(char)toupper((unsigned char)*p); |
| Line 813 Charset* detect_charset(const char* cont | Line 899 Charset* detect_charset(const char* cont |
| if(!end) | if(!end) |
| end=(char*)strchr(begin, ';'); | end=(char*)strchr(begin, ';'); |
| size_t len; | if(end) |
| if(end){ | |
| *end=0; // terminator | *end=0; // terminator |
| len=end-begin; | |
| } else { | |
| len=strlen(begin); | |
| } | |
| String::Body NAME=String::Body(begin, len); | return *begin?&charsets.get(begin):0; |
| return &charsets.get(NAME); | |
| } | } |
| } | } |
| return 0; | return 0; |
| } | } |
| Charset* detect_charset(Charset& source_charset, const String& content_type){ | |
| const String& CONTENT_TYPE=content_type.change_case(source_charset, String::CC_UPPER); | |
| size_t begin=CONTENT_TYPE.pos("CHARSET="); | |
| if(begin!=STRING_NOT_FOUND) { | |
| begin+=8; // skip "CHARSET=" | |
| size_t end=STRING_NOT_FOUND; | |
| if(CONTENT_TYPE.pos('"', begin)==begin){ | |
| begin++; | |
| end=CONTENT_TYPE.pos('"', begin); | |
| } else if(CONTENT_TYPE.pos('\'', begin)==begin){ | |
| begin++; | |
| end=CONTENT_TYPE.pos('\'', begin); | |
| } | |
| if(end==STRING_NOT_FOUND) | |
| end=CONTENT_TYPE.pos(';', begin); | |
| if(end==STRING_NOT_FOUND) | |
| end=CONTENT_TYPE.length(); | |
| const String::Body NAME=CONTENT_TYPE.mid(begin, end); | |
| return &charsets.get(NAME); | |
| } | |
| return 0; | |
| } | |
| static bool isLeap(int year) { | static bool isLeap(int year) { |
| return !( | return !( |
| Line 923 int __vsnprintf(char* b, size_t s, const | Line 976 int __vsnprintf(char* b, size_t s, const |
| if((ssize_t)s<0) | if((ssize_t)s<0) |
| return 0; | return 0; |
| #if _MSC_VER | #ifdef _MSC_VER |
| /* | /* |
| win32: | win32: |
| mk:@MSITStore:C:\Program%20Files\Microsoft%20Visual%20Studio\MSDN\2001APR\1033\vccore.chm::/html/_crt__vsnprintf.2c_._vsnwprintf.htm | mk:@MSITStore:C:\Program%20Files\Microsoft%20Visual%20Studio\MSDN\2001APR\1033\vccore.chm::/html/_crt__vsnprintf.2c_._vsnwprintf.htm |
| Line 985 int __snprintf(char* b, size_t s, const | Line 1038 int __snprintf(char* b, size_t s, const |
| * Foundation, Inc., 59 Temple Street #330, Boston, MA 02111-1307, USA. | * Foundation, Inc., 59 Temple Street #330, Boston, MA 02111-1307, USA. |
| * | * |
| */ | */ |
| static char *base64_alphabet = | static const char *base64_alphabet = |
| "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; | "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; |
| /** | /** |
| Line 1003 static char *base64_alphabet = | Line 1056 static char *base64_alphabet = |
| * | * |
| * Returns the number of bytes encoded. | * Returns the number of bytes encoded. |
| **/ | **/ |
| #define BASE64_GROUPS_IN_LINE 19 | |
| static size_t | static size_t |
| g_mime_utils_base64_encode_step (const unsigned char *in, size_t inlen, unsigned char *out, int *state, int *save) | g_mime_utils_base64_encode_step (const unsigned char *in, size_t inlen, unsigned char *out, int *state, int *save) |
| { | { |
| Line 1040 g_mime_utils_base64_encode_step (const u | Line 1096 g_mime_utils_base64_encode_step (const u |
| *outptr++ = base64_alphabet [((c2 & 0x0f) << 2) | (c3 >> 6)]; | *outptr++ = base64_alphabet [((c2 & 0x0f) << 2) | (c3 >> 6)]; |
| *outptr++ = base64_alphabet [c3 & 0x3f]; | *outptr++ = base64_alphabet [c3 & 0x3f]; |
| /* this is a bit ugly ... */ | /* this is a bit ugly ... */ |
| if ((++already) >= 19) { | if ((++already) >= BASE64_GROUPS_IN_LINE) { |
| *outptr++ = '\n'; | *outptr++ = '\n'; |
| already = 0; | already = 0; |
| } | } |
| Line 1124 g_mime_utils_base64_encode_close (const | Line 1180 g_mime_utils_base64_encode_close (const |
| } | } |
| static unsigned char gmime_base64_rank[256] = { | static unsigned char gmime_base64_rank[256] = { |
| 255,255,255,255,255,255,255,255,255,254,254,255,255,254,255,255, | |
| 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, | 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, |
| 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, | 254,255,255,255,255,255,255,255,255,255,255, 62,255,255,255, 63, |
| 255,255,255,255,255,255,255,255,255,255,255, 62,255,255,255, 63, | |
| 52, 53, 54, 55, 56, 57, 58, 59, 60, 61,255,255,255, 0,255,255, | 52, 53, 54, 55, 56, 57, 58, 59, 60, 61,255,255,255, 0,255,255, |
| 255, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, | 255, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, |
| 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,255,255,255,255,255, | 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,255,255,255,255,255, |
| Line 1149 static unsigned char gmime_base64_rank[2 | Line 1205 static unsigned char gmime_base64_rank[2 |
| * @out: output stream | * @out: output stream |
| * @state: holds the number of bits that are stored in @save | * @state: holds the number of bits that are stored in @save |
| * @save: leftover bits that have not yet been decoded | * @save: leftover bits that have not yet been decoded |
| * @strict: only base64 and whitespace chars are allowed | |
| * | * |
| * Decodes a chunk of base64 encoded data. | * Decodes a chunk of base64 encoded data. |
| * | * |
| * Returns the number of bytes decoded (which have been dumped in @out). | * Returns the number of bytes decoded (which have been dumped in @out). |
| **/ | **/ |
| size_t | size_t |
| g_mime_utils_base64_decode_step (const unsigned char *in, size_t inlen, unsigned char *out, int *state, int *save) | g_mime_utils_base64_decode_step(const unsigned char *in, size_t inlen, unsigned char *out, int *state, int *save, bool strict=false) |
| { | { |
| const unsigned char *inptr; | const unsigned char *inptr; |
| unsigned char *outptr; | unsigned char *outptr; |
| Line 1173 g_mime_utils_base64_decode_step (const u | Line 1230 g_mime_utils_base64_decode_step (const u |
| inptr = in; | inptr = in; |
| while (inptr < inend) { | while (inptr < inend) { |
| c = gmime_base64_rank[*inptr++]; | c = gmime_base64_rank[*inptr++]; |
| if (c != 0xff) { | switch(c) { |
| saved = (saved << 6) | c; | case 0xff: // non-base64 and non-whitespace chars. not allowed in strict mode |
| i++; | if(strict) |
| if (i == 4) { | throw Exception(BASE64_FORMAT, 0, "Invalid base64 char on position %d is detected", inptr-in-1); |
| *outptr++ = (unsigned char)(saved >> 16); | case 0xfe: // whitespace chars 0x09, 0x0A, 0x0D, 0x20 are allowed in any mode |
| *outptr++ = (unsigned char)(saved >> 8); | break; |
| *outptr++ = (unsigned char)(saved); | default: |
| i = 0; | saved = (saved << 6) | c; |
| } | i++; |
| if (i == 4) { | |
| *outptr++ = (unsigned char)(saved >> 16); | |
| *outptr++ = (unsigned char)(saved >> 8); | |
| *outptr++ = (unsigned char)(saved); | |
| i = 0; | |
| } | |
| } | } |
| } | } |
| Line 1193 g_mime_utils_base64_decode_step (const u | Line 1256 g_mime_utils_base64_decode_step (const u |
| i = 2; | i = 2; |
| while (inptr > in && i) { | while (inptr > in && i) { |
| inptr--; | inptr--; |
| if (gmime_base64_rank[*inptr] != 0xff) { | if (gmime_base64_rank[*inptr] <= 0xfe) { |
| if (*inptr == '=' && outptr > out) | if (*inptr == '=' && outptr > out) |
| outptr--; | outptr--; |
| i--; | i--; |
| Line 1206 g_mime_utils_base64_decode_step (const u | Line 1269 g_mime_utils_base64_decode_step (const u |
| char* pa_base64_encode(const char *in, size_t in_size){ | char* pa_base64_encode(const char *in, size_t in_size){ |
| /* wont go to more than 2x size (overly conservative) */ | size_t new_size = ((in_size / 3 + 1) * 4); |
| char* result=new(PointerFreeGC) char[in_size * 2 + 6]; | new_size += new_size / (BASE64_GROUPS_IN_LINE * 4)/*new lines*/ + 1/*zero terminator*/; |
| char* result = new(PointerFreeGC) char[new_size]; | |
| int state=0; | int state=0; |
| int save=0; | int save=0; |
| #ifndef NDEBUG | #ifndef NDEBUG |
| size_t filled= | size_t filled= |
| #endif | #endif |
| g_mime_utils_base64_encode_close ((const unsigned char*)in, in_size, | g_mime_utils_base64_encode_close ((const unsigned char*)in, in_size, (unsigned char*)result, &state, &save); |
| (unsigned char*)result, &state, &save); | |
| assert(filled <= in_size * 2 + 6); | //throw Exception(PARSER_RUNTIME, 0, "%d %d %d", in_size, new_size, filled); |
| assert(filled <= new_size); | |
| return result; | return result; |
| } | } |
| Line 1257 static void file_base64_file_action( | Line 1322 static void file_base64_file_action( |
| } | } |
| } | } |
| void pa_base64_decode(const char *in, size_t in_size, char*& result, size_t& result_size) { | void pa_base64_decode(const char *in, size_t in_size, char*& result, size_t& result_size, bool strict) { |
| /* wont go to more than had (overly conservative) */ | // every 4 base64 bytes are converted into 3 normal bytes |
| result=new(PointerFreeGC) char[in_size+1/*terminator*/]; | // not full set (tail) of 4-bytes set is ignored |
| size_t new_size=in_size/4*3; | |
| result=new(PointerFreeGC) char[new_size+1/*terminator*/]; | |
| int state=0; | int state=0; |
| int save=0; | int save=0; |
| result_size= | result_size= |
| g_mime_utils_base64_decode_step ((const unsigned char*)in, in_size, | g_mime_utils_base64_decode_step ((const unsigned char*)in, in_size, |
| (unsigned char*)result, &state, &save); | (unsigned char*)result, &state, &save, strict); |
| assert(result_size <= in_size); | assert(result_size <= new_size); |
| result[result_size]=0; // for text files | result[result_size]=0; // for text files |
| if(strict && state!=0) | |
| throw Exception(BASE64_FORMAT, 0, "Unexpected end of chars"); | |
| } | } |