// -*- mode: cpp; mode: fold -*-
// Description /*{{{*/
-// $Id: strutl.cc,v 1.23 1999/03/16 07:27:17 jgg Exp $
+// $Id: strutl.cc,v 1.37 2001/02/23 06:08:57 jgg Exp $
/* ######################################################################
- String Util - Some usefull string functions.
+ String Util - Some useful string functions.
- These have been collected from here and there to do all sorts of usefull
- things to strings. They are usefull in file parsers, URI handlers and
+ These have been collected from here and there to do all sorts of useful
+ things to strings. They are useful in file parsers, URI handlers and
especially in APT methods.
This source is placed in the Public Domain, do with it what you will
#include <apt-pkg/strutl.h>
#include <apt-pkg/fileutl.h>
+#include <apt-pkg/error.h>
+#include <apti18n.h>
+
#include <ctype.h>
#include <string.h>
#include <stdio.h>
#include <unistd.h>
+#include <regex.h>
#include <errno.h>
+#include <stdarg.h>
/*}}}*/
// strstrip - Remove white space from the front and back of a string /*{{{*/
// ---------------------------------------------------------------------
/* This grabs a single word, converts any % escaped characters to their
proper values and advances the pointer. Double quotes are understood
- and striped out as well. This is for URI/URL parsing. */
+ and striped out as well. This is for URI/URL parsing. It also can
+ understand [] brackets.*/
bool ParseQuoteWord(const char *&String,string &Res)
{
// Skip leading whitespace
return false;
// Jump to the next word
- for (;*C != 0 && *C != ' '; C++)
+ for (;*C != 0 && isspace(*C) == 0; C++)
{
if (*C == '"')
{
- for (C++;*C != 0 && *C != '"'; C++);
+ for (C++; *C != 0 && *C != '"'; C++);
+ if (*C == 0)
+ return false;
+ }
+ if (*C == '[')
+ {
+ for (C++; *C != 0 && *C != ']'; C++);
if (*C == 0)
return false;
}
Res = Buffer;
// Skip ending white space
- for (;*C != 0 && *C == ' '; C++);
+ for (;*C != 0 && isspace(*C) != 0; C++);
String = C;
return true;
}
/*}}}*/
// ParseCWord - Parses a string like a C "" expression /*{{{*/
// ---------------------------------------------------------------------
-/* This expects a series of space seperated strings enclosed in ""'s.
+/* This expects a series of space separated strings enclosed in ""'s.
It concatenates the ""'s into a single string. */
-bool ParseCWord(const char *String,string &Res)
+bool ParseCWord(const char *&String,string &Res)
{
// Skip leading whitespace
const char *C = String;
if (isspace(*C) == 0)
return false;
*Buf++ = ' ';
- }
+ }
*Buf = 0;
Res = Buffer;
+ String = C;
return true;
}
/*}}}*/
return Temp + string(Str,OldPos);
}
+
+string SubstVar(string Str,const struct SubstVar *Vars)
+{
+ for (; Vars->Subst != 0; Vars++)
+ Str = SubstVar(Str,Vars->Subst,*Vars->Contents);
+ return Str;
+}
/*}}}*/
// URItoFileName - Convert the uri into a unique file name /*{{{*/
// ---------------------------------------------------------------------
file name should be unique and never occur again for a different file */
string URItoFileName(string URI)
{
- string::const_iterator I = URI.begin() + URI.find(':') + 1;
- for (; I < URI.end() && *I == '/'; I++);
-
+ // Nuke 'sensitive' items
+ ::URI U(URI);
+ U.User = string();
+ U.Password = string();
+ U.Access = "";
+
// "\x00-\x20{}|\\\\^\\[\\]<>\"\x7F-\xFF";
- URI = QuoteString(string(I,URI.end() - I),"\\|{}[]<>\"^~_=!@#$%^&*");
+ URI = QuoteString(U,"\\|{}[]<>\"^~_=!@#$%^&*");
string::iterator J = URI.begin();
for (; J != URI.end(); J++)
if (*J == '/')
return false;
// No data
- if (Res <= 0)
+ if (Res < 0 && errno == EAGAIN)
return true;
-
+ if (Res < 0)
+ return false;
+
End += Res;
// Look for the end of the message
/* This handles all 3 populare time formats including RFC 1123, RFC 1036
and the C library asctime format. It requires the GNU library function
'timegm' to convert a struct tm in UTC to a time_t. For some bizzar
- reason the C library does not provide any such function :<*/
+ reason the C library does not provide any such function :< This also
+ handles the weird, but unambiguous FTP time format*/
bool StrToTime(string Val,time_t &Result)
{
struct tm Tm;
for (;*I != 0 && *I != ' '; I++);
// Handle RFC 1123 time
+ Month[0] = 0;
if (sscanf(I," %d %3s %d %d:%d:%d GMT",&Tm.tm_mday,Month,&Tm.tm_year,
&Tm.tm_hour,&Tm.tm_min,&Tm.tm_sec) != 6)
{
// asctime format
if (sscanf(I," %3s %d %d:%d:%d %d",Month,&Tm.tm_mday,
&Tm.tm_hour,&Tm.tm_min,&Tm.tm_sec,&Tm.tm_year) != 6)
- return false;
+ {
+ // 'ftp' time
+ if (sscanf(Val.c_str(),"%4d%2d%2d%2d%2d%2d",&Tm.tm_year,&Tm.tm_mon,
+ &Tm.tm_mday,&Tm.tm_hour,&Tm.tm_min,&Tm.tm_sec) != 6)
+ return false;
+ Tm.tm_mon--;
+ }
}
}
Tm.tm_isdst = 0;
- Tm.tm_mon = MonthConv(Month);
+ if (Month[0] != 0)
+ Tm.tm_mon = MonthConv(Month);
Tm.tm_year -= 1900;
// Convert to local time and then to GMT
return true;
}
/*}}}*/
+// StrToNum - Convert a fixed length string to a number /*{{{*/
+// ---------------------------------------------------------------------
+/* This is used in decoding the crazy fixed length string headers in
+ tar and ar files. */
+bool StrToNum(const char *Str,unsigned long &Res,unsigned Len,unsigned Base)
+{
+ char S[30];
+ if (Len >= sizeof(S))
+ return false;
+ memcpy(S,Str,Len);
+ S[Len] = 0;
+
+ // All spaces is a zero
+ Res = 0;
+ unsigned I;
+ for (I = 0; S[I] == ' '; I++);
+ if (S[I] == 0)
+ return true;
+
+ char *End;
+ Res = strtoul(S,&End,Base);
+ if (End == S)
+ return false;
+
+ return true;
+}
+ /*}}}*/
+// HexDigit - Convert a hex character into an integer /*{{{*/
+// ---------------------------------------------------------------------
+/* Helper for Hex2Num */
+static int HexDigit(int c)
+{
+ if (c >= '0' && c <= '9')
+ return c - '0';
+ if (c >= 'a' && c <= 'f')
+ return c - 'a' + 10;
+ if (c >= 'A' && c <= 'F')
+ return c - 'A' + 10;
+ return 0;
+}
+ /*}}}*/
+// Hex2Num - Convert a long hex number into a buffer /*{{{*/
+// ---------------------------------------------------------------------
+/* The length of the buffer must be exactly 1/2 the length of the string. */
+bool Hex2Num(const char *Start,const char *End,unsigned char *Num,
+ unsigned int Length)
+{
+ if (End - Start != (signed)(Length*2))
+ return false;
+
+ // Convert each digit. We store it in the same order as the string
+ int J = 0;
+ for (const char *I = Start; I < End;J++, I += 2)
+ {
+ if (isxdigit(*I) == 0 || isxdigit(I[1]) == 0)
+ return false;
+
+ Num[J] = HexDigit(I[0]) << 4;
+ Num[J] += HexDigit(I[1]);
+ }
+
+ return true;
+}
+ /*}}}*/
+// TokSplitString - Split a string up by a given token /*{{{*/
+// ---------------------------------------------------------------------
+/* This is intended to be a faster splitter, it does not use dynamic
+ memories. Input is changed to insert nulls at each token location. */
+bool TokSplitString(char Tok,char *Input,char **List,
+ unsigned long ListMax)
+{
+ // Strip any leading spaces
+ char *Start = Input;
+ char *Stop = Start + strlen(Start);
+ for (; *Start != 0 && isspace(*Start) != 0; Start++);
+
+ unsigned long Count = 0;
+ char *Pos = Start;
+ while (Pos != Stop)
+ {
+ // Skip to the next Token
+ for (; Pos != Stop && *Pos != Tok; Pos++);
+
+ // Back remove spaces
+ char *End = Pos;
+ for (; End > Start && (End[-1] == Tok || isspace(End[-1]) != 0); End--);
+ *End = 0;
+
+ List[Count++] = Start;
+ if (Count >= ListMax)
+ {
+ List[Count-1] = 0;
+ return false;
+ }
+
+ // Advance pos
+ for (; Pos != Stop && (*Pos == Tok || isspace(*Pos) != 0 || *Pos == 0); Pos++);
+ Start = Pos;
+ }
+
+ List[Count] = 0;
+ return true;
+}
+ /*}}}*/
+// RegexChoice - Simple regex list/list matcher /*{{{*/
+// ---------------------------------------------------------------------
+/* */
+unsigned long RegexChoice(RxChoiceList *Rxs,const char **ListBegin,
+ const char **ListEnd)
+{
+ for (RxChoiceList *R = Rxs; R->Str != 0; R++)
+ R->Hit = false;
+
+ unsigned long Hits = 0;
+ for (; ListBegin != ListEnd; ListBegin++)
+ {
+ // Check if the name is a regex
+ const char *I;
+ bool Regex = true;
+ for (I = *ListBegin; *I != 0; I++)
+ if (*I == '.' || *I == '?' || *I == '*' || *I == '|')
+ break;
+ if (*I == 0)
+ Regex = false;
+
+ // Compile the regex pattern
+ regex_t Pattern;
+ if (Regex == true)
+ if (regcomp(&Pattern,*ListBegin,REG_EXTENDED | REG_ICASE |
+ REG_NOSUB) != 0)
+ Regex = false;
+
+ // Search the list
+ bool Done = false;
+ for (RxChoiceList *R = Rxs; R->Str != 0; R++)
+ {
+ if (R->Str[0] == 0)
+ continue;
+
+ if (strcasecmp(R->Str,*ListBegin) != 0)
+ {
+ if (Regex == false)
+ continue;
+ if (regexec(&Pattern,R->Str,0,0,0) != 0)
+ continue;
+ }
+ Done = true;
+
+ if (R->Hit == false)
+ Hits++;
+
+ R->Hit = true;
+ }
+
+ if (Regex == true)
+ regfree(&Pattern);
+
+ if (Done == false)
+ _error->Warning(_("Selection %s not found"),*ListBegin);
+ }
+
+ return Hits;
+}
+ /*}}}*/
+// ioprintf - C format string outputter to C++ iostreams /*{{{*/
+// ---------------------------------------------------------------------
+/* This is used to make the internationalization strinc easier to translate
+ and to allow reordering of parameters */
+void ioprintf(ostream &out,const char *format,...)
+{
+ va_list args;
+ va_start(args,format);
+
+ // sprintf the description
+ char S[400];
+ vsnprintf(S,sizeof(S),format,args);
+ out << S;
+}
+ /*}}}*/
+
+// CheckDomainList - See if Host is in a , seperate list /*{{{*/
+// ---------------------------------------------------------------------
+/* The domain list is a comma seperate list of domains that are suffix
+ matched against the argument */
+bool CheckDomainList(string Host,string List)
+{
+ const char *Start = List.begin();
+ for (const char *Cur = List.begin(); Cur <= List.end() ; Cur++)
+ {
+ if (Cur < List.end() && *Cur != ',')
+ continue;
+
+ // Match the end of the string..
+ if ((Host.size() >= (unsigned)(Cur - List.begin())) &&
+ Cur - Start != 0 &&
+ stringcasecmp(Host.end() - (Cur - Start),Host.end(),Start,Cur) == 0)
+ return true;
+
+ Start = Cur + 1;
+ }
+ return false;
+}
+ /*}}}*/
// URI::CopyFrom - Copy from an object /*{{{*/
// ---------------------------------------------------------------------
{
string::const_iterator I = U.begin();
- // Locate the first colon, this seperates the scheme
+ // Locate the first colon, this separates the scheme
for (; I < U.end() && *I != ':' ; I++);
string::const_iterator FirstColon = I;
string::const_iterator SingleSlash = I;
if (I + 3 < U.end() && I[1] == '/' && I[2] == '/')
SingleSlash += 3;
- for (; SingleSlash < U.end() && *SingleSlash != '/'; SingleSlash++);
+
+ /* Find the / indicating the end of the hostname, ignoring /'s in the
+ square brackets */
+ bool InBracket = false;
+ for (; SingleSlash < U.end() && (*SingleSlash != '/' || InBracket == true); SingleSlash++)
+ {
+ if (*SingleSlash == '[')
+ InBracket = true;
+ if (InBracket == true && *SingleSlash == ']')
+ InBracket = false;
+ }
+
if (SingleSlash > U.end())
SingleSlash = U.end();
if (FirstColon > SingleSlash)
FirstColon = SingleSlash;
- // Search for the @
- I = FirstColon;
+ // Find the colon...
+ I = FirstColon + 1;
+ if (I > SingleSlash)
+ I = SingleSlash;
+ for (; I < SingleSlash && *I != ':'; I++);
+ string::const_iterator SecondColon = I;
+
+ // Search for the @ after the colon
for (; I < SingleSlash && *I != '@'; I++);
string::const_iterator At = I;
- // Colon in the @ section
- I = FirstColon + 1;
- for (; I < At && *I != ':'; I++);
- string::const_iterator SecondColon = I;
-
// Now write the host and user/pass
if (At == SingleSlash)
{
Password = string(U,SecondColon - U.begin() + 1,At - SecondColon - 1);
}
- // Now we parse off a pot number from the hostname
+ // Now we parse the RFC 2732 [] hostnames.
+ unsigned long PortEnd = 0;
+ InBracket = false;
+ for (unsigned I = 0; I != Host.length();)
+ {
+ if (Host[I] == '[')
+ {
+ InBracket = true;
+ Host.erase(I,1);
+ continue;
+ }
+
+ if (InBracket == true && Host[I] == ']')
+ {
+ InBracket = false;
+ Host.erase(I,1);
+ PortEnd = I;
+ continue;
+ }
+ I++;
+ }
+
+ // Tsk, weird.
+ if (InBracket == true)
+ {
+ Host = string();
+ return;
+ }
+
+ // Now we parse off a port number from the hostname
Port = 0;
string::size_type Pos = Host.rfind(':');
- if (Pos == string::npos)
+ if (Pos == string::npos || Pos < PortEnd)
return;
Port = atoi(string(Host,Pos+1).c_str());
/* */
URI::operator string()
{
- string Res = Access + ':';
+ string Res;
+
+ if (Access.empty() == false)
+ Res = Access + ':';
+
if (Host.empty() == false)
- {
- Res += "//";
+ {
+ if (Access.empty() == false)
+ Res += "//";
+
if (User.empty() == false)
{
- Res += "//" + User;
+ Res += User;
if (Password.empty() == false)
Res += ":" + Password;
Res += "@";
}
- Res += Host;
+
+ // Add RFC 2732 escaping characters
+ if (Access.empty() == false &&
+ (Host.find('/') != string::npos || Host.find(':') != string::npos))
+ Res += '[' + Host + ']';
+ else
+ Res += Host;
+
if (Port != 0)
{
char S[30];
return Res;
}
/*}}}*/
+// URI::SiteOnly - Return the schema and site for the URI /*{{{*/
+// ---------------------------------------------------------------------
+/* */
+string URI::SiteOnly(string URI)
+{
+ ::URI U(URI);
+ U.User = string();
+ U.Password = string();
+ U.Path = string();
+ U.Port = 0;
+ return U;
+}
+ /*}}}*/
projects / apt.git / blobdiff