// -*- mode: cpp; mode: fold -*-
// Description /*{{{*/
-// $Id: strutl.cc,v 1.5 1998/10/20 02:39:30 jgg Exp $
+// $Id: strutl.cc,v 1.30 1999/10/17 07:30:23 jgg Exp $
/* ######################################################################
String Util - Some usefull string functions.
- strstrip - Remove whitespace from the front and end of a line.
+ 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
+ especially in APT methods.
This source is placed in the Public Domain, do with it what you will
- It was originally written by Jason Gunthorpe <jgg@gpu.srv.ualberta.ca>
+ It was originally written by Jason Gunthorpe <jgg@gpu.srv.ualberta.ca>
##################################################################### */
/*}}}*/
// Includes /*{{{*/
-#include <strutl.h>
+#ifdef __GNUG__
+#pragma implementation "apt-pkg/strutl.h"
+#endif
+
+#include <apt-pkg/strutl.h>
+#include <apt-pkg/fileutl.h>
+
#include <ctype.h>
#include <string.h>
#include <stdio.h>
+#include <unistd.h>
+#include <errno.h>
/*}}}*/
// strstrip - Remove white space from the front and back of a string /*{{{*/
return false;
// Jump to the next word
- for (;*C != 0 && *C != ' '; C++)
+ for (;*C != 0 && isspace(*C) == 0; C++)
{
if (*C == '"')
{
{
Tmp[0] = Start[1];
Tmp[1] = Start[2];
- Tmp[3] = 0;
+ Tmp[2] = 0;
*I = (char)strtol(Tmp,0,16);
Start += 3;
continue;
Res = Buffer;
// Skip ending white space
- for (;*C != 0 && *C == ' '; C++);
+ for (;*C != 0 && isspace(*C) != 0; C++);
String = C;
return true;
}
return Res;
}
/*}}}*/
+// DeQuoteString - Convert a string from quoted from /*{{{*/
+// ---------------------------------------------------------------------
+/* This undoes QuoteString */
+string DeQuoteString(string Str)
+{
+ string Res;
+ for (string::iterator I = Str.begin(); I != Str.end(); I++)
+ {
+ if (*I == '%' && I + 2 < Str.end())
+ {
+ char Tmp[3];
+ Tmp[0] = I[1];
+ Tmp[1] = I[2];
+ Tmp[2] = 0;
+ Res += (char)strtol(Tmp,0,16);
+ I += 2;
+ continue;
+ }
+ else
+ Res += *I;
+ }
+ return Res;
+}
+
+ /*}}}*/
// SizeToStr - Convert a long into a human readable size /*{{{*/
// ---------------------------------------------------------------------
-/* A max of 4 digits are shown before conversion to the next highest unit. The
- max length of the string will be 5 chars unless the size is > 10
+/* A max of 4 digits are shown before conversion to the next highest unit.
+ The max length of the string will be 5 chars unless the size is > 10
YottaBytes (E24) */
string SizeToStr(double Size)
{
/* bytes, KiloBytes, MegaBytes, GigaBytes, TeraBytes, PetaBytes,
ExaBytes, ZettaBytes, YottaBytes */
- char Ext[] = {'b','k','M','G','T','P','E','Z','Y'};
+ char Ext[] = {'\0','k','M','G','T','P','E','Z','Y'};
int I = 0;
while (I <= 8)
{
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 URI;
}
/*}}}*/
-// URIAccess - Return the access method for the URI /*{{{*/
-// ---------------------------------------------------------------------
-/* */
-string URIAccess(string URI)
-{
- string::size_type Pos = URI.find(':');
- if (Pos == string::npos)
- return string();
- return string(URI,0,Pos);
-}
- /*}}}*/
// Base64Encode - Base64 Encoding routine for short strings /*{{{*/
// ---------------------------------------------------------------------
/* This routine performs a base64 transformation on a string. It was ripped
if (strcasecmp(Text.c_str(),"no") == 0 ||
strcasecmp(Text.c_str(),"false") == 0 ||
strcasecmp(Text.c_str(),"without") == 0 ||
+ strcasecmp(Text.c_str(),"off") == 0 ||
strcasecmp(Text.c_str(),"disable") == 0)
return 0;
if (strcasecmp(Text.c_str(),"yes") == 0 ||
strcasecmp(Text.c_str(),"true") == 0 ||
strcasecmp(Text.c_str(),"with") == 0 ||
+ strcasecmp(Text.c_str(),"on") == 0 ||
strcasecmp(Text.c_str(),"enable") == 0)
return 1;
return Default;
}
/*}}}*/
+// TimeRFC1123 - Convert a time_t into RFC1123 format /*{{{*/
+// ---------------------------------------------------------------------
+/* This converts a time_t into a string time representation that is
+ year 2000 complient and timezone neutral */
+string TimeRFC1123(time_t Date)
+{
+ struct tm Conv = *gmtime(&Date);
+ char Buf[300];
+
+ const char *Day[] = {"Sun","Mon","Tue","Wed","Thu","Fri","Sat"};
+ const char *Month[] = {"Jan","Feb","Mar","Apr","May","Jun","Jul",
+ "Aug","Sep","Oct","Nov","Dec"};
+
+ sprintf(Buf,"%s, %02i %s %i %02i:%02i:%02i GMT",Day[Conv.tm_wday],
+ Conv.tm_mday,Month[Conv.tm_mon],Conv.tm_year+1900,Conv.tm_hour,
+ Conv.tm_min,Conv.tm_sec);
+ return Buf;
+}
+ /*}}}*/
+// ReadMessages - Read messages from the FD /*{{{*/
+// ---------------------------------------------------------------------
+/* This pulls full messages from the input FD into the message buffer.
+ It assumes that messages will not pause during transit so no
+ fancy buffering is used. */
+bool ReadMessages(int Fd, vector<string> &List)
+{
+ char Buffer[4000];
+ char *End = Buffer;
+
+ while (1)
+ {
+ int Res = read(Fd,End,sizeof(Buffer) - (End-Buffer));
+ if (Res < 0 && errno == EINTR)
+ continue;
+
+ // Process is dead, this is kind of bad..
+ if (Res == 0)
+ return false;
+
+ // No data
+ if (Res <= 0)
+ return true;
+
+ End += Res;
+
+ // Look for the end of the message
+ for (char *I = Buffer; I + 1 < End; I++)
+ {
+ if (I[0] != '\n' || I[1] != '\n')
+ continue;
+
+ // Pull the message out
+ string Message(Buffer,0,I-Buffer);
+
+ // Fix up the buffer
+ for (; I < End && *I == '\n'; I++);
+ End -= I-Buffer;
+ memmove(Buffer,I,End-Buffer);
+ I = Buffer;
+
+ List.push_back(Message);
+ }
+ if (End == Buffer)
+ return true;
+
+ if (WaitFd(Fd) == false)
+ return false;
+ }
+}
+ /*}}}*/
+// MonthConv - Converts a month string into a number /*{{{*/
+// ---------------------------------------------------------------------
+/* This was lifted from the boa webserver which lifted it from 'wn-v1.07'
+ Made it a bit more robust with a few touppers though. */
+static int MonthConv(char *Month)
+{
+ switch (toupper(*Month))
+ {
+ case 'A':
+ return toupper(Month[1]) == 'P'?3:7;
+ case 'D':
+ return 11;
+ case 'F':
+ return 1;
+ case 'J':
+ if (toupper(Month[1]) == 'A')
+ return 0;
+ return toupper(Month[2]) == 'N'?5:6;
+ case 'M':
+ return toupper(Month[2]) == 'R'?2:4;
+ case 'N':
+ return 10;
+ case 'O':
+ return 9;
+ case 'S':
+ return 8;
+
+ // Pretend it is January..
+ default:
+ return 0;
+ }
+}
+ /*}}}*/
+// timegm - Internal timegm function if gnu is not available /*{{{*/
+// ---------------------------------------------------------------------
+/* Ripped this evil little function from wget - I prefer the use of
+ GNU timegm if possible as this technique will have interesting problems
+ with leap seconds, timezones and other.
+
+ Converts struct tm to time_t, assuming the data in tm is UTC rather
+ than local timezone (mktime assumes the latter).
+
+ Contributed by Roger Beeman <beeman@cisco.com>, with the help of
+ Mark Baushke <mdb@cisco.com> and the rest of the Gurus at CISCO. */
+#ifndef __USE_MISC // glib sets this
+static time_t timegm(struct tm *t)
+{
+ time_t tl, tb;
+
+ tl = mktime (t);
+ if (tl == -1)
+ return -1;
+ tb = mktime (gmtime (&tl));
+ return (tl <= tb ? (tl + (tl - tb)) : (tl - (tb - tl)));
+}
+#endif
+ /*}}}*/
+// StrToTime - Converts a string into a time_t /*{{{*/
+// ---------------------------------------------------------------------
+/* 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 :<*/
+bool StrToTime(string Val,time_t &Result)
+{
+ struct tm Tm;
+ char Month[10];
+ const char *I = Val.c_str();
+
+ // Skip the day of the week
+ for (;*I != 0 && *I != ' '; I++);
+
+ // Handle RFC 1123 time
+ 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)
+ {
+ // Handle RFC 1036 time
+ 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)
+ Tm.tm_year += 1900;
+ else
+ {
+ // 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;
+ }
+ }
+
+ Tm.tm_isdst = 0;
+ Tm.tm_mon = MonthConv(Month);
+ Tm.tm_year -= 1900;
+
+ // Convert to local time and then to GMT
+ Result = timegm(&Tm);
+ 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;
+}
+ /*}}}*/
+
+// URI::CopyFrom - Copy from an object /*{{{*/
+// ---------------------------------------------------------------------
+/* This parses the URI into all of its components */
+void URI::CopyFrom(string U)
+{
+ string::const_iterator I = U.begin();
+
+ // Locate the first colon, this seperates the scheme
+ for (; I < U.end() && *I != ':' ; I++);
+ string::const_iterator FirstColon = I;
+
+ /* Determine if this is a host type URI with a leading double //
+ and then search for the first single / */
+ string::const_iterator SingleSlash = I;
+ if (I + 3 < U.end() && I[1] == '/' && I[2] == '/')
+ SingleSlash += 3;
+ for (; SingleSlash < U.end() && *SingleSlash != '/'; SingleSlash++);
+ if (SingleSlash > U.end())
+ SingleSlash = U.end();
+
+ // We can now write the access and path specifiers
+ Access = string(U,0,FirstColon - U.begin());
+ if (SingleSlash != U.end())
+ Path = string(U,SingleSlash - U.begin());
+ if (Path.empty() == true)
+ Path = "/";
+
+ // Now we attempt to locate a user:pass@host fragment
+ if (FirstColon[1] == '/' && FirstColon[2] == '/')
+ FirstColon += 3;
+ else
+ FirstColon += 1;
+ if (FirstColon >= U.end())
+ return;
+
+ if (FirstColon > SingleSlash)
+ FirstColon = SingleSlash;
+
+ // 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;
+
+ // Now write the host and user/pass
+ if (At == SingleSlash)
+ {
+ if (FirstColon < SingleSlash)
+ Host = string(U,FirstColon - U.begin(),SingleSlash - FirstColon);
+ }
+ else
+ {
+ Host = string(U,At - U.begin() + 1,SingleSlash - At - 1);
+ User = string(U,FirstColon - U.begin(),SecondColon - FirstColon);
+ if (SecondColon < At)
+ Password = string(U,SecondColon - U.begin() + 1,At - SecondColon - 1);
+ }
+
+ // Now we parse off a port number from the hostname
+ Port = 0;
+ string::size_type Pos = Host.rfind(':');
+ if (Pos == string::npos)
+ return;
+
+ Port = atoi(string(Host,Pos+1).c_str());
+ Host = string(Host,0,Pos);
+}
+ /*}}}*/
+// URI::operator string - Convert the URI to a string /*{{{*/
+// ---------------------------------------------------------------------
+/* */
+URI::operator string()
+{
+ string Res;
+
+ if (Access.empty() == false)
+ Res = Access + ':';
+
+ if (Host.empty() == false)
+ {
+ if (Access.empty() == false)
+ Res += "//";
+
+ if (User.empty() == false)
+ {
+ Res += User;
+ if (Password.empty() == false)
+ Res += ":" + Password;
+ Res += "@";
+ }
+
+ Res += Host;
+ if (Port != 0)
+ {
+ char S[30];
+ sprintf(S,":%u",Port);
+ Res += S;
+ }
+ }
+
+ if (Path.empty() == false)
+ {
+ if (Path[0] != '/')
+ Res += "/" + Path;
+ else
+ Res += Path;
+ }
+
+ return Res;
+}
+ /*}}}*/