|
|
| version 1.257, 2010/04/19 19:35:55 | version 1.285, 2015/09/02 21:29:45 |
|---|---|
| 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" |
| #include "pa_idna.h" | |
| #include "pa_convert_utf.h" | |
| #ifdef _MSC_VER | |
| #include <windows.h> | |
| #include <direct.h> | |
| #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; | |
| // some maybe-undefined constants | // some maybe-undefined constants |
| #ifndef _O_TEXT | #ifndef _O_TEXT |
| Line 67 const String file_status_name(FILE_STATU | Line 81 const String file_status_name(FILE_STATU |
| // functions | // functions |
| bool capitalized(const char* s){ | char* file_read_text(Request_charsets& charsets, const String& file_spec, bool fail_on_read_problem, HashStringValue* params, bool transcode_result) { |
| bool upper=true; | File_read_result file=file_read(charsets, file_spec, true, params, fail_on_read_problem, 0, 0, 0, transcode_result); |
| 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) | |
| memcpy(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) | |
| memcpy(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) { | |
| 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 160 struct File_read_action_info { | Line 113 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); |
| : (char *)pa_malloc_atomic(to_read_size+1); | 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=(char *)pa_malloc_atomic(1); | *info.data=(char *)pa_malloc_atomic(1); |
| Line 200 File_read_result file_read(Request_chars | Line 145 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 213 File_read_result file_read(Request_chars | Line 156 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); | String::C body=String::C(result.str, result.length); |
| body=Charset::transcode(body, asked_charset, charsets.source()); | body=Charset::transcode(body, *asked_charset, charsets.source()); |
| result.str=const_cast<char*>(body.str); // hacking a little | result.str=const_cast<char*>(body.str); // hacking a little |
| result.length=body.length; | result.length=body.length; |
| } | |
| } | } |
| } | } |
| if(result.length) | if(result.length) |
| Line 265 File_read_result file_load(Request& r, c | Line 209 File_read_result file_load(Request& r, c |
| } | } |
| void check_safe_mode(struct stat finfo, const String& file_spec, const char* fname) { | |
| #ifdef PA_SAFE_MODE | #ifdef PA_SAFE_MODE |
| 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 278 void check_safe_mode(struct stat finfo, | Line 222 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()); |
| } | |
| #else | |
| void check_safe_mode(struct stat, const String&, const char*) { | |
| } | |
| #endif | #endif |
| } | |
| bool file_read_action_under_lock(const String& file_spec, | bool file_read_action_under_lock(const String& file_spec, |
| Line 341 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 411 struct File_write_action_info { | Line 359 struct File_write_action_info { |
| size_t length; | 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((size_t)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( |
| Request_charsets& charsets, | Request_charsets& charsets, |
| const String& file_spec, | const String& file_spec, |
| Line 448 void file_write( | Line 398 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) { |
| 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 _MSC_VER | |
| rmdir(file_spec.mid(0, pos_after-1/* / */).taint_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.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 467 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) | |
| rmdir(file_spec, 1); | |
| 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.taint_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.taint_cstr(String::L_FILE_SPEC); | const char* new_spec_cstr=new_spec.taint_cstr(String::L_FILE_SPEC); |
| Line 484 void file_move(const String& old_spec, c | Line 486 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 501 bool entry_exists(const String& file_spe | Line 504 bool entry_exists(const String& file_spe |
| 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.taint_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 531 bool dir_exists(const String& file_spec) | Line 514 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; |
| } | } |
| Line 548 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 556 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 563 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 614 enum FormatType { | Line 661 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 625 FormatType format_type(char* fmt){ | Line 672 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 669 FormatType format_type(char* fmt){ | Line 716 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 726 enum EscapeState { | Line 773 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 735 char* unescape_chars(const char* cp, int | Line 782 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 792 char* unescape_chars(const char* cp, int | Line 839 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 799 void back_slashes_to_slashes(char* s) { | Line 866 void back_slashes_to_slashes(char* s) { |
| if(*s=='\\') | if(*s=='\\') |
| *s='/'; | *s='/'; |
| } | } |
| /* | |
| void slashes_to_back_slashes(char* s) { | |
| if(s) | |
| for(; *s; s++) | |
| if(*s=='/') | |
| *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 916 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 924 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 |
| if((ssize_t)s<0) | if((ssize_t)s<0) |
| return 0; | return 0; |
| #if _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 994 int __snprintf(char* b, size_t s, const | Line 991 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 1136 g_mime_utils_base64_encode_close (const | Line 1133 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 1161 static unsigned char gmime_base64_rank[2 | Line 1158 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 1185 g_mime_utils_base64_decode_step (const u | Line 1183 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 1205 g_mime_utils_base64_decode_step (const u | Line 1209 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 1234 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 1271 static void file_base64_file_action( | Line 1267 static void file_base64_file_action( |
| } | } |
| } | } |
| void pa_base64_decode(const char *in, size_t in_size, char*& result, size_t& result_size) { | char* pa_base64_encode(const String& file_spec){ |
| /* wont go to more than had (overly conservative) */ | unsigned char* base64=0; |
| result=new(PointerFreeGC) char[in_size+1/*terminator*/]; | 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) { | |
| // every 4 base64 bytes are converted into 3 normal bytes | |
| // 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"); | |
| } | } |
| Line 1293 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 1303 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 1326 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; | |
| } |