// -*- mode: cpp; mode: fold -*-
// Description /*{{{*/
-// $Id: strutl.cc,v 1.2 1998/07/16 06:08:41 jgg Exp $
+// $Id: strutl.cc,v 1.9 1998/10/30 07:53:45 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>
+#include <apt-pkg/fileutl.h>
+
#include <ctype.h>
#include <string.h>
#include <stdio.h>
+#include <time.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. */
+ and striped out as well. This is for URI/URL parsing. */
bool ParseQuoteWord(const char *&String,string &Res)
{
// Skip leading whitespace
return true;
}
/*}}}*/
+// ParseCWord - Parses a string like a C "" expression /*{{{*/
+// ---------------------------------------------------------------------
+/* This expects a series of space seperated strings enclosed in ""'s.
+ It concatenates the ""'s into a single string. */
+bool ParseCWord(const char *String,string &Res)
+{
+ // Skip leading whitespace
+ const char *C = String;
+ for (;*C != 0 && *C == ' '; C++);
+ if (*C == 0)
+ return false;
+
+ char Buffer[1024];
+ char *Buf = Buffer;
+ if (strlen(String) >= sizeof(Buffer))
+ return false;
+
+ for (; *C != 0; C++)
+ {
+ if (*C == '"')
+ {
+ for (C++; *C != 0 && *C != '"'; C++)
+ *Buf++ = *C;
+
+ if (*C == 0)
+ return false;
+
+ continue;
+ }
+
+ if (C != String && isspace(*C) != 0 && isspace(C[-1]) != 0)
+ continue;
+ if (isspace(*C) == 0)
+ return false;
+ *Buf++ = ' ';
+ }
+ *Buf = 0;
+ Res = Buffer;
+ return true;
+}
+ /*}}}*/
// QuoteString - Convert a string into quoted from /*{{{*/
// ---------------------------------------------------------------------
/* */
/*}}}*/
// 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)
{
return Temp + string(Str,OldPos);
}
/*}}}*/
+// URItoFileName - Convert the uri into a unique file name /*{{{*/
+// ---------------------------------------------------------------------
+/* This converts a URI into a safe filename. It quotes all unsafe characters
+ and converts / to _ and removes the scheme identifier. The resulting
+ 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++);
+
+ // "\x00-\x20{}|\\\\^\\[\\]<>\"\x7F-\xFF";
+ URI = QuoteString(string(I,URI.end() - I),"\\|{}[]<>\"^~_=!@#$%^&*");
+ string::iterator J = URI.begin();
+ for (; J != URI.end(); J++)
+ if (*J == '/')
+ *J = '_';
+ return URI;
+}
+ /*}}}*/
// Base64Encode - Base64 Encoding routine for short strings /*{{{*/
// ---------------------------------------------------------------------
/* This routine performs a base64 transformation on a string. It was ripped
for (; A != AEnd && B != BEnd; A++, B++)
if (toupper(*A) != toupper(*B))
break;
-
+
if (A == AEnd && B == BEnd)
return 0;
if (A == AEnd)
return 1;
}
/*}}}*/
+// LookupTag - Lookup the value of a tag in a taged string /*{{{*/
+// ---------------------------------------------------------------------
+/* The format is like those used in package files and the method
+ communication system */
+string LookupTag(string Message,const char *Tag,const char *Default)
+{
+ // Look for a matching tag.
+ int Length = strlen(Tag);
+ for (string::iterator I = Message.begin(); I + Length < Message.end(); I++)
+ {
+ // Found the tag
+ if (I[Length] == ':' && stringcasecmp(I,I+Length,Tag) == 0)
+ {
+ // Find the end of line and strip the leading/trailing spaces
+ string::iterator J;
+ I += Length + 1;
+ for (; isspace(*I) != 0 && I < Message.end(); I++);
+ for (J = I; *J != '\n' && J < Message.end(); J++);
+ for (; J > I && isspace(J[-1]) != 0; J--);
+
+ return string(I,J-I);
+ }
+
+ for (; *I != '\n' && I < Message.end(); I++);
+ }
+
+ // Failed to find a match
+ if (Default == 0)
+ return string();
+ return Default;
+}
+ /*}}}*/
+// StringToBool - Converts a string into a boolean /*{{{*/
+// ---------------------------------------------------------------------
+/* This inspects the string to see if it is true or if it is false and
+ then returns the result. Several varients on true/false are checked. */
+int StringToBool(string Text,int Default = -1)
+{
+ char *End;
+ int Res = strtol(Text.c_str(),&End,0);
+ if (End != Text.c_str() && Res >= 0 && Res <= 1)
+ return Res;
+
+ // Check for positives
+ if (strcasecmp(Text.c_str(),"no") == 0 ||
+ strcasecmp(Text.c_str(),"false") == 0 ||
+ strcasecmp(Text.c_str(),"without") == 0 ||
+ strcasecmp(Text.c_str(),"disable") == 0)
+ return 0;
+
+ // Check for negatives
+ if (strcasecmp(Text.c_str(),"yes") == 0 ||
+ strcasecmp(Text.c_str(),"true") == 0 ||
+ strcasecmp(Text.c_str(),"with") == 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));
+
+ // 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;
+ }
+}
+ /*}}}*/
+// 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;
+}
+ /*}}}*/
+
+// URI::URI - Constructor /*{{{*/
+// ---------------------------------------------------------------------
+/* This parses the URI into all of its components */
+URI::URI(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 //
+ string::const_iterator SingleSlash = I;
+ if (I + 3 < U.end() && I[1] == '/' && I[2] == '/')
+ {
+ // Locate the single / that starts the path
+ for (; I < U.end(); I++)
+ {
+ if (*I == '/' && I[1] == '/')
+ I += 2;
+ else
+ if (*I == '/')
+ break;
+ }
+ if (I > U.end())
+ I = U.end();
+ SingleSlash = I;
+ }
+
+ // 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() + 1);
+
+ // Now we attempt to locate a user:pass@host fragment
+ FirstColon += 3;
+ if (FirstColon >= U.end())
+ return;
+
+ if (FirstColon > SingleSlash)
+ FirstColon = SingleSlash;
+
+ // Search for the @
+ I = FirstColon;
+ 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)
+ {
+ 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 pot 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 = Access + ':';
+ if (Host.empty() == false)
+ {
+ if (User.empty() == false)
+ {
+ Res += "//" + User;
+ if (Password.empty() == false)
+ Res += ":" + Password;
+ Res += "@";
+ }
+ Res += Host;
+ }
+
+ if (Path.empty() == false)
+ Res += "/" + Path;
+
+ return Res;
+}
+ /*}}}*/
projects / apt.git / blobdiff