|
|
| version 1.276, 2013/07/23 12:10:56 | version 1.285, 2015/09/02 21:29:45 |
|---|---|
| Line 36 | Line 36 |
| #include "pcre.h" | #include "pcre.h" |
| #include "pa_request.h" | #include "pa_request.h" |
| #include "pa_idna.h" | |
| #include "pa_convert_utf.h" | |
| #ifdef _MSC_VER | #ifdef _MSC_VER |
| #include <windows.h> | #include <windows.h> |
| #include <direct.h> | #include <direct.h> |
| #endif | #endif |
| #ifdef _MSC_VER | |
| #define pa_mkdir(path, mode) _mkdir(path) | |
| #else | |
| #define pa_mkdir(path, mode) mkdir(path, mode) | |
| #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; | 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 |
| Line 72 const String file_status_name(FILE_STATU | Line 81 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) { | |
| //_asm int 3; | |
| const char* const eob=str+length; | |
| char* dest=str; | |
| // fix DOS: \r\n -> \n | |
| // fix Macintosh: \r -> \n | |
| char* bol=str; | |
| while(char* eol=(char*)memchr(bol, '\r', eob -bol)) { | |
| size_t len=eol-bol; | |
| if(dest!=bol) | |
| memmove(dest, bol, len); | |
| dest+=len; | |
| *dest++='\n'; | |
| if(&eol[1]<eob && eol[1]=='\n') { // \r, \n = DOS | |
| bol=eol+2; | |
| length--; | |
| } else // \r, not \n = Macintosh | |
| bol=eol+1; | |
| } | |
| // last piece without \r | |
| if(dest!=bol) | |
| memmove(dest, bol, eob-bol); | |
| str[length]=0; // terminating | |
| } | |
| char* file_read_text(Request_charsets& charsets, const String& file_spec, bool fail_on_read_problem, HashStringValue* params, bool transcode_result) { | char* file_read_text(Request_charsets& charsets, const String& file_spec, 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); | 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; |
| Line 331 void create_dir_for_file(const String& f | Line 289 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).taint_cstr(String::L_FILE_SPEC), 0775); | pa_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 478 static bool entry_readable(const String& | Line 436 static bool entry_readable(const String& |
| } | } |
| // 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=0) { | static void rmdir(const String& file_spec, size_t pos_after) { |
| char* dir_spec=file_spec.taint_cstrm(String::L_FILE_SPEC); | char* dir_spec=file_spec.taint_cstrm(String::L_FILE_SPEC); |
| size_t length=strlen(dir_spec); | size_t length=strlen(dir_spec); |
| while( (length=get_dir(dir_spec, length)) && (length > pos_after) ){ | while( (length=get_dir(dir_spec, length)) && (length > pos_after) ){ |
| Line 500 static void rmdir(const String& file_spe | Line 458 static void rmdir(const String& file_spe |
| bool file_delete(const String& file_spec, bool fail_on_problem, bool keep_empty_dirs) { | bool file_delete(const String& file_spec, bool fail_on_problem, bool keep_empty_dirs) { |
| const char* fname=file_spec.taint_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, |
| &file_spec, | &file_spec, |
| Line 508 bool file_delete(const String& file_spec | Line 466 bool file_delete(const String& file_spec |
| strerror(errno), errno, fname); | strerror(errno), errno, fname); |
| else | else |
| return false; | return false; |
| } | |
| if(!keep_empty_dirs) | if(!keep_empty_dirs) |
| rmdir(file_spec, 1); | rmdir(file_spec, 1); |
| Line 573 bool file_stat(const String& file_spec, | Line 532 bool file_stat(const String& file_spec, |
| bool fail_on_read_problem) { | bool fail_on_read_problem) { |
| const char* fname=file_spec.taint_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) |
| throw Exception("file.missing", | throw Exception("file.missing", |
| &file_spec, | &file_spec, |
| Line 581 bool file_stat(const String& file_spec, | Line 540 bool file_stat(const String& file_spec, |
| strerror(errno), errno, fname); | strerror(errno), errno, fname); |
| else | else |
| return false; | return false; |
| } | |
| rsize=finfo.st_size; | rsize=finfo.st_size; |
| ratime=finfo.st_atime; | ratime=finfo.st_atime; |
| rmtime=finfo.st_mtime; | rmtime=finfo.st_mtime; |
| Line 588 bool file_stat(const String& file_spec, | Line 548 bool file_stat(const String& file_spec, |
| return true; | return true; |
| } | } |
| /** | |
| String related 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) { | |
| //_asm int 3; | |
| const char* const eob=str+length; | |
| char* dest=str; | |
| // fix DOS: \r\n -> \n | |
| // fix Macintosh: \r -> \n | |
| char* bol=str; | |
| while(char* eol=(char*)memchr(bol, '\r', eob -bol)) { | |
| size_t len=eol-bol; | |
| if(dest!=bol) | |
| memmove(dest, bol, len); | |
| dest+=len; | |
| *dest++='\n'; | |
| if(&eol[1]<eob && eol[1]=='\n') { // \r, \n = DOS | |
| bol=eol+2; | |
| length--; | |
| } else // \r, not \n = Macintosh | |
| bol=eol+1; | |
| } | |
| // last piece without \r | |
| if(dest!=bol) | |
| memmove(dest, bol, eob-bol); | |
| str[length]=0; // terminating | |
| } | |
| /** | |
| scans for @a delim[default \n] in @a *row_ref, | |
| @return piece of line before it or end of string, if no @a delim found | |
| assigns @a *row_ref to point right after delimiter if there were one | |
| or to zero if no @a delim were found. | |
| */ | |
| char* getrow(char* *row_ref, char delim) { | char* getrow(char* *row_ref, char delim) { |
| char* result=*row_ref; | char* result=*row_ref; |
| if(result) { | if(result) { |
| Line 846 void back_slashes_to_slashes(char* s) { | Line 868 void back_slashes_to_slashes(char* s) { |
| } | } |
| #endif | #endif |
| bool StrStartFromNC(const char* str, const char* substr, bool equal){ | |
| while(true) { | |
| if(!(*substr)){ | |
| if(!(*str)) | |
| return true; | |
| else | |
| return !equal; | |
| } | |
| if(!(*str)) | |
| return false; | |
| if(isalpha((unsigned char)*str)) { | |
| if(tolower((unsigned char)*str)!=tolower((unsigned char)*substr)) | |
| return false; | |
| } else if((*str) != (*substr)) | |
| return false; | |
| str++; | |
| substr++; | |
| } | |
| } | |
| size_t strpos(const char *str, const char *substr) { | size_t strpos(const char *str, const char *substr) { |
| const char *p = strstr(str, substr); | const char *p = strstr(str, substr); |
| return (p==0)?STRING_NOT_FOUND:p-str; | return (p==0)?STRING_NOT_FOUND:p-str; |
| } | } |
| // content-type: xxx; charset=WE-NEED-THIS | |
| // content-type: xxx; charset="WE-NEED-THIS" | |
| // content-type: xxx; charset="WE-NEED-THIS"; | |
| Charset* detect_charset(const char* content_type){ | |
| if(content_type){ | |
| char* CONTENT_TYPE=pa_strdup(content_type); | |
| for(char *p=CONTENT_TYPE; *p; p++) | |
| *p=(char)toupper((unsigned char)*p); | |
| if(const char* begin=strstr(CONTENT_TYPE, "CHARSET=")){ | |
| begin+=8; // skip "CHARSET=" | |
| char* end=0; | |
| if(*begin && (*begin=='"' || *begin =='\'')){ | |
| char quote=*begin; | |
| begin++; | |
| end=(char*)strchr(begin, quote); | |
| } | |
| if(!end) | |
| end=(char*)strchr(begin, ';'); | |
| if(end) | |
| *end=0; // terminator | |
| return *begin?&charsets.get(begin):0; | |
| } | |
| } | |
| return 0; | |
| } | |
| static bool isLeap(int year) { | |
| return !( | |
| (year % 4) || ((year % 400) && !(year % 100)) | |
| ); | |
| } | |
| int getMonthDays(int year, int month) { | |
| static int monthDays[]={ | |
| 31, | |
| 28, | |
| 31, | |
| 30, | |
| 31, | |
| 30, | |
| 31, | |
| 31, | |
| 30, | |
| 31, | |
| 30, | |
| 31 | |
| }; | |
| return (month == 1 /* january -- 0 */ && isLeap(year)) ? 29 : monthDays[month]; | |
| } | |
| int remove_crlf(char* start, char* end) { | int remove_crlf(char* start, char* end) { |
| char* from=start; | char* from=start; |
| char* to=start; | char* to=start; |
| Line 953 int remove_crlf(char* start, char* end) | Line 900 int remove_crlf(char* start, char* end) |
| return to-start; | return to-start; |
| } | } |
| const char* hex_digits="0123456789ABCDEF"; | |
| const char* hex_string(unsigned char* bytes, size_t size, bool upcase) { | |
| char *bytes_hex=new(PointerFreeGC) char [size*2/*byte->hh*/+1/*for zero-teminator*/]; | |
| unsigned char *src=bytes; | |
| unsigned char *end=bytes+size; | |
| char *dest=bytes_hex; | |
| const char *hex=upcase? hex_digits : "0123456789abcdef"; | |
| for(; src<end; src++) { | |
| *dest++=hex[*src/0x10]; | |
| *dest++=hex[*src%0x10]; | |
| } | |
| *dest=0; | |
| return bytes_hex; | |
| } | |
| /// must be last in this file | /// must be last in this file |
| #undef vsnprintf | #undef vsnprintf |
| Line 961 int __vsnprintf(char* b, size_t s, const | Line 926 int __vsnprintf(char* b, size_t s, const |
| return 0; | return 0; |
| int r; | int r; |
| // note: on win32& maybe somewhere else | // note: on win32 & maybe somewhere else |
| // vsnprintf do not writes terminating 0 in 'buffer full' case, reducing | // vsnprintf do not writes terminating 0 in 'buffer full' case, reducing |
| // http://stackoverflow.com/questions/2915672/snprintf-and-visual-studio-2010 | |
| --s; | --s; |
| // clients do not check for negative 's', feature: ignore such prints | // clients do not check for negative 's', feature: ignore such prints |
| Line 970 int __vsnprintf(char* b, size_t s, const | Line 936 int __vsnprintf(char* b, size_t s, const |
| return 0; | return 0; |
| #ifdef _MSC_VER | #ifdef _MSC_VER |
| /* | // win32: if the number of bytes to write exceeds buffer, then count bytes are written and -1 is returned |
| win32: | |
| mk:@MSITStore:C:\Program%20Files\Microsoft%20Visual%20Studio\MSDN\2001APR\1033\vccore.chm::/html/_crt__vsnprintf.2c_._vsnwprintf.htm | |
| if the number of bytes to write exceeds buffer, then count bytes are written and Ö1 is returned | |
| */ | |
| r=_vsnprintf(b, s, f, l); | r=_vsnprintf(b, s, f, l); |
| if(r<0) | if(r<0) |
| r=s; | r=s; |
| #else | #else |
| r=vsnprintf(b, s, f, l); | r=vsnprintf(b, s, f, l); |
| /* | /* |
| solaris: | solaris: man vsnprintf |
| man vsnprintf | |
| The snprintf() function returns the number of characters | The snprintf() function returns the number of characters |
| formatted, that is, the number of characters that would have | formatted, that is, the number of characters that would have |
| been written to the buffer if it were large enough. If the | been written to the buffer if it were large enough. If the |
| value of n is 0 on a call to snprintf(), an unspecified | value of n is 0 on a call to snprintf(), an unspecified |
| Line 1278 char* pa_base64_encode(const char *in, s | Line 1238 char* pa_base64_encode(const char *in, s |
| return result; | return result; |
| } | } |
| struct File_base64_action_info { | |
| char* pa_base64_encode(const String& file_spec){ | unsigned char** base64; |
| unsigned char* base64=0; | }; |
| File_base64_action_info info={&base64}; | |
| file_read_action_under_lock(file_spec, | |
| "pa_base64_encode", file_base64_file_action, &info); | |
| return (char*)base64; | |
| } | |
| static void file_base64_file_action( | static void file_base64_file_action( |
| struct stat& finfo, | struct stat& finfo, |
| Line 1315 static void file_base64_file_action( | Line 1267 static void file_base64_file_action( |
| } | } |
| } | } |
| char* pa_base64_encode(const String& file_spec){ | |
| unsigned char* base64=0; | |
| File_base64_action_info info={&base64}; | |
| file_read_action_under_lock(file_spec, | |
| "pa_base64_encode", file_base64_file_action, &info); | |
| return (char*)base64; | |
| } | |
| void pa_base64_decode(const char *in, size_t in_size, char*& result, size_t& result_size, bool strict) { | void pa_base64_decode(const char *in, size_t in_size, char*& result, size_t& result_size, bool strict) { |
| // every 4 base64 bytes are converted into 3 normal bytes | // every 4 base64 bytes are converted into 3 normal bytes |
| // not full set (tail) of 4-bytes set is ignored | // not full set (tail) of 4-bytes set is ignored |
| Line 1343 int file_block_read(const int f, unsigne | Line 1305 int file_block_read(const int f, unsigne |
| return nCount; | return nCount; |
| } | } |
| static unsigned long crc32Table[256]; | |
| static void InitCrc32Table() | |
| { | |
| if(crc32Table[1] == 0){ | |
| // This is the official polynomial used by CRC32 in PKZip. | |
| // Often times the polynomial shown reversed as 0x04C11DB7. | |
| static const unsigned long dwPolynomial = 0xEDB88320; | |
| for(int i = 0; i < 256; i++) | |
| { | |
| unsigned long dwCrc = i; | |
| for(int j = 8; j > 0; j--) | |
| { | |
| if(dwCrc & 1) | |
| dwCrc = (dwCrc >> 1) ^ dwPolynomial; | |
| else | |
| dwCrc >>= 1; | |
| } | |
| crc32Table[i] = dwCrc; | |
| } | |
| } | |
| } | |
| inline void CalcCrc32(const unsigned char byte, unsigned long &crc32) | |
| { | |
| crc32 = ((crc32) >> 8) ^ crc32Table[(byte) ^ ((crc32) & 0x000000FF)]; | |
| } | |
| const unsigned long pa_crc32(const char *in, size_t in_size){ | const unsigned long pa_crc32(const char *in, size_t in_size){ |
| unsigned long crc32=0xFFFFFFFF; | unsigned long crc32=0xFFFFFFFF; |
| Line 1353 const unsigned long pa_crc32(const char | Line 1344 const unsigned long pa_crc32(const char |
| return ~crc32; | return ~crc32; |
| } | } |
| const unsigned long pa_crc32(const String& file_spec){ | |
| unsigned long crc32=0xFFFFFFFF; | |
| file_read_action_under_lock(file_spec, "crc32", file_crc32_file_action, &crc32); | |
| return ~crc32; | |
| } | |
| static void file_crc32_file_action( | static void file_crc32_file_action( |
| struct stat& finfo, | struct stat& finfo, |
| int f, | int f, |
| Line 1376 static void file_crc32_file_action( | Line 1361 static void file_crc32_file_action( |
| } | } |
| } | } |
| const unsigned long pa_crc32(const String& file_spec){ | |
| unsigned long crc32=0xFFFFFFFF; | |
| file_read_action_under_lock(file_spec, "crc32", file_crc32_file_action, &crc32); | |
| return ~crc32; | |
| } | |
| // content-type: xxx; charset=WE-NEED-THIS | |
| // content-type: xxx; charset="WE-NEED-THIS" | |
| // content-type: xxx; charset="WE-NEED-THIS"; | |
| Charset* detect_charset(const char* content_type){ | |
| if(content_type){ | |
| char* CONTENT_TYPE=pa_strdup(content_type); | |
| for(char *p=CONTENT_TYPE; *p; p++) | |
| *p=(char)toupper((unsigned char)*p); | |
| if(const char* begin=strstr(CONTENT_TYPE, "CHARSET=")){ | |
| begin+=8; // skip "CHARSET=" | |
| char* end=0; | |
| if(*begin && (*begin=='"' || *begin =='\'')){ | |
| char quote=*begin; | |
| begin++; | |
| end=(char*)strchr(begin, quote); | |
| } | |
| if(!end) | |
| end=(char*)strchr(begin, ';'); | |
| if(end) | |
| *end=0; // terminator | |
| return *begin?&charsets.get(begin):0; | |
| } | |
| } | |
| return 0; | |
| } | |
| static bool is_latin(const char *in){ | |
| for(; *in; in++){ | |
| if ((unsigned char)(*in) > 0x7F) | |
| return false; | |
| } | |
| return true; | |
| } | |
| #define MAX_IDNA_LENGTH 256 | |
| const char *pa_idna_encode(const char *in, Charset &source_charset){ | |
| if(!in || is_latin(in)) | |
| return in; | |
| uint32_t utf32[MAX_IDNA_LENGTH]; | |
| uint32_t *utf32_end=utf32; | |
| String::C sIn(in,strlen(in)); | |
| if(!source_charset.isUTF8()) | |
| sIn=Charset::transcode(sIn, source_charset, UTF8_charset); | |
| int status=pa_convertUTF8toUTF32((const UTF8**)&sIn.str, (const UTF8*)(sIn.str+sIn.length), &utf32_end, utf32+MAX_IDNA_LENGTH-1, strictConversion); | |
| if(status != conversionOK) | |
| throw Exception("idna encode", new String(in), "utf conversion failed (%d)", status); | |
| *utf32_end=0; | |
| char *result = (char *)pa_malloc(MAX_IDNA_LENGTH); | |
| status=pa_idna_to_ascii_4z(utf32, result, MAX_IDNA_LENGTH, 0); | |
| if(status != IDNA_SUCCESS) | |
| throw Exception("idna encode", new String(in), "encode failed: %s", pa_idna_strerror(status)); | |
| return result; | |
| } | |
| const char *pa_idna_decode(const char *in, Charset &asked_charset){ | |
| if(!in || !(*in)) | |
| return in; | |
| uint32_t utf32[MAX_IDNA_LENGTH]; | |
| const uint32_t *utf32_start=utf32; | |
| uint32_t *utf32_end; | |
| int status=pa_idna_to_unicode_4z(in, utf32, MAX_IDNA_LENGTH, 0); | |
| if(status != IDNA_SUCCESS) | |
| throw Exception("idna decode", new String(in), "decode failed: %s", pa_idna_strerror(status)); | |
| for(utf32_end=utf32; *utf32_end; utf32_end++); | |
| char *result = (char *)pa_malloc(MAX_IDNA_LENGTH); | |
| char *result_end = result; | |
| status=pa_convertUTF32toUTF8(&utf32_start, utf32_end, (UTF8**)&result_end, (UTF8*)(result+MAX_IDNA_LENGTH-1), strictConversion); | |
| if(status != conversionOK) | |
| throw Exception("idna decode", new String(in), "utf conversion failed (%d)", status); | |
| *result_end='\0'; | |
| if(!asked_charset.isUTF8()) | |
| result = (char *)Charset::transcode(result, UTF8_charset, asked_charset).cstr(); | |
| return result; | |
| } |