/*********************************************************************
*
* Helper Functions for Microchip TCP/IP Stack
*
*********************************************************************
* FileName: Helpers.C
* Dependencies: None
* Processor: PIC18, PIC24F, PIC24H, dsPIC30F, dsPIC33F, PIC32
* Processor: PIC18, PIC24F, PIC24H, dsPIC30F, dsPIC33F, PIC32
* Compiler: Microchip C32 v1.05 or higher
* Microchip C30 v3.12 or higher
* Microchip C18 v3.30 or higher
* HI-TECH PICC-18 PRO 9.63PL2 or higher
* Company: Microchip Technology, Inc.
*
* Software License Agreement
*
* Copyright (C) 2002-2009 Microchip Technology Inc. All rights
* reserved.
*
* Microchip licenses to you the right to use, modify, copy, and
* distribute:
* (i) the Software when embedded on a Microchip microcontroller or
* digital signal controller product ("Device") which is
* integrated into Licensee's product; or
* (ii) ONLY the Software driver source files ENC28J60.c, ENC28J60.h,
* ENCX24J600.c and ENCX24J600.h ported to a non-Microchip device
* used in conjunction with a Microchip ethernet controller for
* the sole purpose of interfacing with the ethernet controller.
*
* You should refer to the license agreement accompanying this
* Software for additional information regarding your rights and
* obligations.
*
* THE SOFTWARE AND DOCUMENTATION ARE PROVIDED "AS IS" WITHOUT
* WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT
* LIMITATION, ANY WARRANTY OF MERCHANTABILITY, FITNESS FOR A
* PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT SHALL
* MICROCHIP BE LIABLE FOR ANY INCIDENTAL, SPECIAL, INDIRECT OR
* CONSEQUENTIAL DAMAGES, LOST PROFITS OR LOST DATA, COST OF
* PROCUREMENT OF SUBSTITUTE GOODS, TECHNOLOGY OR SERVICES, ANY CLAIMS
* BY THIRD PARTIES (INCLUDING BUT NOT LIMITED TO ANY DEFENSE
* THEREOF), ANY CLAIMS FOR INDEMNITY OR CONTRIBUTION, OR OTHER
* SIMILAR COSTS, WHETHER ASSERTED ON THE BASIS OF CONTRACT, TORT
* (INCLUDING NEGLIGENCE), BREACH OF WARRANTY, OR OTHERWISE.
*
*
* Author Date Comment
*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
* Nilesh Rajbharti 5/17/01 Original (Rev 1.0)
* Nilesh Rajbharti 2/9/02 Cleanup
* Nilesh Rajbharti 6/25/02 Rewritten CalcIPChecksum() to avoid
* multi-byte shift operation.
* Howard Schlunder 2/9/05 Added hexatob(), btohexa_high(), and
* btohexa_low()
* Howard Schlunder 10/10/06 Optimized swapl()
* Elliott Wood 11/20/07 Added leftRotateDWORD()
********************************************************************/
#define __HELPERS_C
#include "TCPIP Stack/TCPIP.h"
/*****************************************************************************
Function:
DWORD GenerateRandomDWORD(void)
Summary:
Generates a random DWORD.
Description:
This function generates a random 32-bit integer. It collects
randomness by comparing the A/D converter's internal R/C oscillator
clock with our main system clock. By passing collected entropy to the
C rand()/srand() functions, the output is normalized to meet statistical
randomness tests.
Precondition:
None
Parameters:
None
Returns:
Random 32-bit number.
Side Effects:
This function uses the A/D converter (and so you must disable
interrupts if you use the A/D converted in your ISR). The C rand()
function will be reseeded, and Timer0 (PIC18) and Timer1 (PIC24,
dsPIC, and PIC32) will be used. TMR#H:TMR#L will have a new value.
Note that this is the same timer used by the Tick module.
Remarks:
This function times out after 1 second of attempting to generate the
random DWORD. In such a case, the output may not be truly random.
Typically, this function executes in around 500,000 instruction cycles.
The intent of this function is to produce statistically random and
cryptographically secure random number. Whether or not this is true on
all (or any) devices/voltages/temperatures is not tested.
***************************************************************************/
DWORD GenerateRandomDWORD(void)
{
BYTE vBitCount;
WORD w, wTime, wLastValue;
DWORD dwTotalTime;
DWORD dwRandomResult;
#if defined __18CXX
{
BYTE ADCON0Save, ADCON2Save;
BYTE T0CONSave, TMR0HSave, TMR0LSave;
// Save hardware SFRs
ADCON0Save = ADCON0;
ADCON2Save = ADCON2;
T0CONSave = T0CON;
TMR0LSave = TMR0L;
TMR0HSave = TMR0H;
// Set up Timer and A/D converter module
ADCON0 = 0x01; // Turn on the A/D module
ADCON2 = 0x3F; // 20 Tad acquisition, Frc A/D clock used for conversion
T0CON = 0x88; // TMR0ON = 1, no prescalar
vBitCount = 0;
dwTotalTime = 0;
wLastValue = 0;
dwRandomResult = rand();
while(1)
{
// Time the duration of an A/D acquisition and conversion
TMR0H = 0x00;
TMR0L = 0x00;
ADCON0bits.GO = 1;
ClrWdt();
while(ADCON0bits.GO);
((BYTE*)&wTime)[0] = TMR0L;
((BYTE*)&wTime)[1] = TMR0H;
w = rand();
// Wait no longer than 1 second obtaining entropy
dwTotalTime += wTime;
if(dwTotalTime >= GetInstructionClock())
{
dwRandomResult ^= rand() | (((DWORD)rand())<<15ul) | (((DWORD)rand())<<30ul);
break;
}
// Keep sampling if minimal entropy was likely obtained this round
if(wLastValue == wTime)
continue;
// Add this entropy into the pseudo random number generator by reseeding
srand(w + (wLastValue - wTime));
wLastValue = wTime;
// Accumulate at least 32 bits of randomness over time
dwRandomResult <<= 1;
if(rand() >= 16384)
dwRandomResult |= 0x1;
// See if we've collected a fair amount of entropy and can quit early
if(++vBitCount == 0u)
break;
}
// Restore hardware SFRs
ADCON0 = ADCON0Save;
ADCON2 = ADCON2Save;
TMR0H = TMR0HSave;
TMR0L = TMR0LSave;
T0CON = T0CONSave;
}
#else
{
WORD AD1CON1Save, AD1CON2Save, AD1CON3Save;
WORD T1CONSave, PR1Save;
// Save hardware SFRs
AD1CON1Save = AD1CON1;
AD1CON2Save = AD1CON2;
AD1CON3Save = AD1CON3;
T1CONSave = T1CON;
PR1Save = PR1;
// Set up Timer and A/D converter module
AD1CON1 = 0x80E4; // Turn on the A/D module, auto-convert
AD1CON2 = 0x003F; // Interrupt after every 16th sample/convert
AD1CON3 = 0x9F00; // Frc A/D clock, 31 Tad acquisition
T1CON = 0x8000; // TON = 1, no prescalar
PR1 = 0xFFFF; // Don't clear timer early
vBitCount = 0;
dwTotalTime = 0;
wLastValue = 0;
dwRandomResult = rand();
while(1)
{
ClrWdt();
#if defined(__C30__)
while(!IFS0bits.AD1IF);
#else
while(!IFS1bits.AD1IF);
#endif
wTime = TMR1;
TMR1 = 0x0000;
#if defined(__C30__)
IFS0bits.AD1IF = 0;
#else
IFS1bits.AD1IF = 0;
#endif
w = rand();
// Wait no longer than 1 second obtaining entropy
dwTotalTime += wTime;
if(dwTotalTime >= GetInstructionClock())
{
dwRandomResult ^= rand() | (((DWORD)rand())<<15) | (((DWORD)rand())<<30);
break;
}
// Keep sampling if minimal entropy was likely obtained this round
if(wLastValue == wTime)
continue;
// Add this entropy into the pseudo random number generator by reseeding
srand(w + (wLastValue - wTime));
wLastValue = wTime;
// Accumulate at least 32 bits of randomness over time
dwRandomResult <<= 1;
if(rand() >= 16384)
dwRandomResult |= 0x1;
// See if we've collected a fair amount of entropy and can quit early
if(++vBitCount == 0u)
break;
}
// Restore hardware SFRs
AD1CON1 = AD1CON1Save;
AD1CON2 = AD1CON2Save;
AD1CON3 = AD1CON3Save;
T1CON = T1CONSave;
PR1 = PR1Save;
}
#endif
return dwRandomResult;
}
#if defined(STACK_USE_HTTP_SERVER)
/*****************************************************************************
Function:
void UnencodeURL(BYTE* URL)
Summary:
Decodes a URL-encoded string.
Description:
This function is deprecated except for use with HTTP Classic. It
attempts to decode a URL encoded string, converting all hex escape
sequences into a literal byte. However, it is inefficient over long
strings and does not handle URL-encoded data strings ('&' and '=').
Precondition:
None
Parameters:
URL - the null-terminated string to decode
Returns:
None
***************************************************************************/
void UnencodeURL(BYTE* URL)
{
BYTE *Right, *Copy;
WORD_VAL Number;
while((Right = (BYTE*)strchr((char*)URL, '%')))
{
// Make sure the string is long enough
if(Right[1] == '\0')
break;
if(Right[2] == '\0')
break;
// Update the string in place
Number.v[0] = Right[2];
Number.v[1] = Right[1];
*Right++ = hexatob(Number);
URL = Right;
// Remove two blank spots by shifting all remaining characters right two
Copy = Right + 2;
while((*Right++ = *Copy++));
}
}
#endif
/*****************************************************************************
Function:
BOOL StringToIPAddress(BYTE* str, IP_ADDR* IPAddress)
Summary:
Converts a string to an IP address
Description:
This function parses a dotted-quad decimal IP address string into an
IP_ADDR struct. The output result is big-endian.
Precondition:
None
Parameters:
str - Pointer to a dotted-quad IP address string
IPAddress - Pointer to IP_ADDR in which to store the result
Return Values:
TRUE - an IP address was successfully decoded
FALSE - no IP address could be found, or the format was incorrect
***************************************************************************/
BOOL StringToIPAddress(BYTE* str, IP_ADDR* IPAddress)
{
DWORD_VAL dwVal;
BYTE i, charLen, currentOctet;
charLen = 0;
currentOctet = 0;
dwVal.Val = 0;
while((i = *str++))
{
if(currentOctet > 3u)
break;
i -= '0';
// Validate the character is a numerical digit or dot, depending on location
if(charLen == 0u)
{
if(i > 9u)
return FALSE;
}
else if(charLen == 3u)
{
if(i != (BYTE)('.' - '0'))
return FALSE;
if(dwVal.Val > 0x00020505ul)
return FALSE;
IPAddress->v[currentOctet++] = dwVal.v[2]*((BYTE)100) + dwVal.v[1]*((BYTE)10) + dwVal.v[0];
charLen = 0;
dwVal.Val = 0;
continue;
}
else
{
if(i == (BYTE)('.' - '0'))
{
if(dwVal.Val > 0x00020505ul)
return FALSE;
IPAddress->v[currentOctet++] = dwVal.v[2]*((BYTE)100) + dwVal.v[1]*((BYTE)10) + dwVal.v[0];
charLen = 0;
dwVal.Val = 0;
continue;
}
if(i > 9u)
return FALSE;
}
charLen++;
dwVal.Val <<= 8;
dwVal.v[0] = i;
}
// Make sure the very last character is a valid termination character
// (i.e., not more hostname, which could be legal and not an IP
// address as in "10.5.13.233.picsaregood.com"
if(i != 0u && i != '/' && i != '\r' && i != '\n' && i != ' ' && i != '\t')
return FALSE;
// Verify and convert the last octet and return the result
if(dwVal.Val > 0x00020505ul)
return FALSE;
IPAddress->v[3] = dwVal.v[2]*((BYTE)100) + dwVal.v[1]*((BYTE)10) + dwVal.v[0];
return TRUE;
}
/*****************************************************************************
Function:
BOOL ROMStringToIPAddress(ROM BYTE* str, IP_ADDR* IPAddress)
Summary:
Converts a string to an IP address
Description:
This function parses a dotted-quad decimal IP address string into an
IP_ADDR struct. The output result is big-endian.
Precondition:
None
Parameters:
str - Pointer to a dotted-quad IP address string
IPAddress - Pointer to IP_ADDR in which to store the result
Return Values:
TRUE - an IP address was successfully decoded
FALSE - no IP address could be found, or the format was incorrect
Remarks:
This function is aliased to StringToIPAddress on non-PIC18 platforms.
***************************************************************************/
#if defined(__18CXX)
BOOL ROMStringToIPAddress(ROM BYTE* str, IP_ADDR* IPAddress)
{
DWORD_VAL dwVal;
BYTE i, charLen, currentOctet;
charLen = 0;
currentOctet = 0;
dwVal.Val = 0;
while(i = *str++)
{
if(currentOctet > 3u)
break;
i -= '0';
// Validate the character is a numerical digit or dot, depending on location
if(charLen == 0u)
{
if(i > 9u)
return FALSE;
}
else if(charLen == 3u)
{
if(i != (BYTE)('.' - '0'))
return FALSE;
if(dwVal.Val > 0x00020505ul)
return FALSE;
IPAddress->v[currentOctet++] = dwVal.v[2]*((BYTE)100) + dwVal.v[1]*((BYTE)10) + dwVal.v[0];
charLen = 0;
dwVal.Val = 0;
continue;
}
else
{
if(i == (BYTE)('.' - '0'))
{
if(dwVal.Val > 0x00020505ul)
return FALSE;
IPAddress->v[currentOctet++] = dwVal.v[2]*((BYTE)100) + dwVal.v[1]*((BYTE)10) + dwVal.v[0];
charLen = 0;
dwVal.Val = 0;
continue;
}
if(i > 9u)
return FALSE;
}
charLen++;
dwVal.Val <<= 8;
dwVal.v[0] = i;
}
// Make sure the very last character is a valid termination character
// (i.e., not more hostname, which could be legal and not an IP
// address as in "10.5.13.233.picsaregood.com"
if(i != 0u && i != '/' && i != '\r' && i != '\n' && i != ' ' && i != '\t')
return FALSE;
// Verify and convert the last octet and return the result
if(dwVal.Val > 0x00020505ul)
return FALSE;
IPAddress->v[3] = dwVal.v[2]*((BYTE)100) + dwVal.v[1]*((BYTE)10) + dwVal.v[0];
return TRUE;
}
#endif
/*****************************************************************************
Function:
WORD Base64Decode(BYTE* cSourceData, WORD wSourceLen,
BYTE* cDestData, WORD wDestLen)
Description:
Decodes a Base-64 array to its literal representation.
Precondition:
None
Parameters:
cSourceData - Pointer to a string of Base-64 encoded data
wSourceLen - Length of the Base-64 source data
cDestData - Pointer to write the decoded data
wSourceLen - Maximum length that can be written to cDestData
Returns:
Number of decoded bytes written to cDestData.
Remarks:
This function is binary safe and will ignore invalid characters (CR, LF,
etc). If cSourceData is equal to cDestData, the data will be converted
in-place. If cSourceData is not equal to cDestData, but the regions
overlap, the behavior is undefined.
Decoded data is always at least 1/4 smaller than the source data.
***************************************************************************/
#if defined(STACK_USE_BASE64_DECODE)
WORD Base64Decode(BYTE* cSourceData, WORD wSourceLen, BYTE* cDestData, WORD wDestLen)
{
BYTE i;
BYTE vByteNumber;
BOOL bPad;
WORD wBytesOutput;
vByteNumber = 0;
wBytesOutput = 0;
// Loop over all provided bytes
while(wSourceLen--)
{
// Fetch a Base64 byte and decode it to the original 6 bits
i = *cSourceData++;
bPad = (i == '=');
if(i >= 'A' && i <= 'Z') // Regular data
i -= 'A' - 0;
else if(i >= 'a' && i <= 'z')
i -= 'a' - 26;
else if(i >= '0' && i <= '9')
i -= '0' - 52;
else if(i == '+' || i == '-')
i = 62;
else if(i == '/' || i == '_')
i = 63;
else // Skip all padding (=) and non-Base64 characters
continue;
// Write the 6 bits to the correct destination location(s)
if(vByteNumber == 0u)
{
if(bPad) // Padding here would be illegal, treat it as a non-Base64 chacter and just skip over it
continue;
vByteNumber++;
if(wBytesOutput >= wDestLen)
break;
wBytesOutput++;
*cDestData = i << 2;
}
else if(vByteNumber == 1u)
{
vByteNumber++;
*cDestData++ |= i >> 4;
if(wBytesOutput >= wDestLen)
break;
if(bPad)
continue;
wBytesOutput++;
*cDestData = i << 4;
}
else if(vByteNumber == 2u)
{
vByteNumber++;
*cDestData++ |= i >> 2;
if(wBytesOutput >= wDestLen)
break;
if(bPad)
continue;
wBytesOutput++;
*cDestData = i << 6;
}
else if(vByteNumber == 3u)
{
vByteNumber = 0;
*cDestData++ |= i;
}
}
return wBytesOutput;
}
#endif // #if defined(STACK_USE_BASE64_DECODE)
/*****************************************************************************
Function:
WORD Base64Encode(BYTE* cSourceData, WORD wSourceLen,
BYTE* cDestData, WORD wDestLen)
Description:
Encodes a binary array to Base-64.
Precondition:
None
Parameters:
cSourceData - Pointer to a string of binary data
wSourceLen - Length of the binary source data
cDestData - Pointer to write the Base-64 encoded data
wSourceLen - Maximum length that can be written to cDestData
Returns:
Number of encoded bytes written to cDestData. This will always be
a multiple of 4.
Remarks:
Encoding cannot be performed in-place. If cSourceData overlaps with
cDestData, the behavior is undefined.
Encoded data is always at least 1/3 larger than the source data. It may
be 1 or 2 bytes larger than that.
***************************************************************************/
#if defined(STACK_USE_BASE64_ENCODE) || defined(STACK_USE_SMTP_CLIENT) || defined(STACK_USE_DYNAMICDNS_CLIENT)
WORD Base64Encode(BYTE* cSourceData, WORD wSourceLen, BYTE* cDestData, WORD wDestLen)
{
BYTE i, j;
BYTE vOutput[4];
WORD wOutputLen;
wOutputLen = 0;
while(wDestLen >= 4u)
{
// Start out treating the output as all padding
vOutput[0] = 0xFF;
vOutput[1] = 0xFF;
vOutput[2] = 0xFF;
vOutput[3] = 0xFF;
// Get 3 input octets and split them into 4 output hextets (6-bits each)
if(wSourceLen == 0u)
break;
i = *cSourceData++;
wSourceLen--;
vOutput[0] = (i & 0xFC)>>2;
vOutput[1] = (i & 0x03)<<4;
if(wSourceLen)
{
i = *cSourceData++;
wSourceLen--;
vOutput[1] |= (i & 0xF0)>>4;
vOutput[2] = (i & 0x0F)<<2;
if(wSourceLen)
{
i = *cSourceData++;
wSourceLen--;
vOutput[2] |= (i & 0xC0)>>6;
vOutput[3] = i & 0x3F;
}
}
// Convert hextets into Base 64 alphabet and store result
for(i = 0; i < 4u; i++)
{
j = vOutput[i];
if(j <= 25u)
j += 'A' - 0;
else if(j <= 51u)
j += 'a' - 26;
else if(j <= 61u)
j += '0' - 52;
else if(j == 62u)
j = '+';
else if(j == 63u)
j = '/';
else // Padding
j = '=';
*cDestData++ = j;
}
// Update counters
wDestLen -= 4;
wOutputLen += 4;
}
return wOutputLen;
}
#endif // #if defined(STACK_USE_BASE64_ENCODE) || defined(STACK_USE_SMTP) || defined(STACK_USE_DYNAMICDNS_CLIENT)
/*****************************************************************************
Function:
void uitoa(WORD Value, BYTE* Buffer)
Summary:
Converts an unsigned integer to a decimal string.
Description:
Converts a 16-bit unsigned integer to a null-terminated decimal string.
Precondition:
None
Parameters:
Value - The number to be converted
Buffer - Pointer in which to store the converted string
Returns:
None
***************************************************************************/
void uitoa(WORD Value, BYTE* Buffer)
{
BYTE i;
WORD Digit;
WORD Divisor;
BOOL Printed = FALSE;
if(Value)
{
for(i = 0, Divisor = 10000; i < 5u; i++)
{
Digit = Value/Divisor;
if(Digit || Printed)
{
*Buffer++ = '0' + Digit;
Value -= Digit*Divisor;
Printed = TRUE;
}
Divisor /= 10;
}
}
else
{
*Buffer++ = '0';
}
*Buffer = '\0';
}
/*****************************************************************************
Function:
void ultoa(DWORD Value, BYTE* Buffer)
Summary:
Converts an unsigned integer to a decimal string.
Description:
Converts a 32-bit unsigned integer to a null-terminated decimal string.
Precondition:
None
Parameters:
Value - The number to be converted
Buffer - Pointer in which to store the converted string
Returns:
None
***************************************************************************/
// HI-TECH PICC-18 PRO 9.63 already has a ultoa() library function
#if !defined(__18CXX) && !defined(HI_TECH_C)
void ultoa(DWORD Value, BYTE* Buffer)
{
BYTE i;
DWORD Digit;
DWORD Divisor;
BOOL Printed = FALSE;
if(Value)
{
for(i = 0, Divisor = 1000000000; i < 10; i++)
{
Digit = Value/Divisor;
if(Digit || Printed)
{
*Buffer++ = '0' + Digit;
Value -= Digit*Divisor;
Printed = TRUE;
}
Divisor /= 10;
}
}
else
{
*Buffer++ = '0';
}
*Buffer = '\0';
}
#endif
/*****************************************************************************
Function:
BYTE hexatob(WORD_VAL AsciiChars)
Summary:
Converts a hex string to a single byte.
Description:
Converts a two-character ASCII hex string to a single packed byte.
Precondition:
None
Parameters:
AsciiChars - WORD_VAL where .v[0] is the ASCII value for the lower nibble
and .v[1] is the ASCII value for the upper nibble. Each
must range from '0'-'9', 'A'-'F', or 'a'-'f'.
Returns:
Resulting packed byte 0x00 - 0xFF.
***************************************************************************/
BYTE hexatob(WORD_VAL AsciiChars)
{
// Convert lowercase to uppercase
if(AsciiChars.v[1] > 'F')
AsciiChars.v[1] -= 'a'-'A';
if(AsciiChars.v[0] > 'F')
AsciiChars.v[0] -= 'a'-'A';
// Convert 0-9, A-F to 0x0-0xF
if(AsciiChars.v[1] > '9')
AsciiChars.v[1] -= 'A' - 10;
else
AsciiChars.v[1] -= '0';
if(AsciiChars.v[0] > '9')
AsciiChars.v[0] -= 'A' - 10;
else
AsciiChars.v[0] -= '0';
// Concatenate
return (AsciiChars.v[1]<<4) | AsciiChars.v[0];
}
/*****************************************************************************
Function:
BYTE btohexa_high(BYTE b)
Summary:
Converts the upper nibble of a binary value to a hexadecimal ASCII byte.
Description:
Converts the upper nibble of a binary value to a hexadecimal ASCII byte.
For example, btohexa_high(0xAE) will return 'A'.
Precondition:
None
Parameters:
b - the byte to convert
Returns:
The upper hexadecimal ASCII byte '0'-'9' or 'A'-'F'.
***************************************************************************/
BYTE btohexa_high(BYTE b)
{
b >>= 4;
return (b>0x9u) ? b+'A'-10:b+'0';
}
/*****************************************************************************
Function:
BYTE btohexa_high(BYTE b)
Summary:
Converts the lower nibble of a binary value to a hexadecimal ASCII byte.
Description:
Converts the lower nibble of a binary value to a hexadecimal ASCII byte.
For example, btohexa_high(0xAE) will return 'E'.
Precondition:
None
Parameters:
b - the byte to convert
Returns:
The lower hexadecimal ASCII byte '0'-'9' or 'A'-'F'.
***************************************************************************/
BYTE btohexa_low(BYTE b)
{
b &= 0x0F;
return (b>9u) ? b+'A'-10:b+'0';
}
/*****************************************************************************
Function:
signed char stricmppgm2ram(BYTE* a, ROM BYTE* b)
Summary:
Case-insensitive comparison of a string in RAM to a string in ROM.
Description:
Performs a case-insensitive comparison of a string in RAM to a string
in ROM. This function performs identically to strcmppgm2ram, except that
the comparison is not case-sensitive.
Precondition:
None
Parameters:
a - Pinter to tring in RAM
b - Pointer to string in ROM
Return Values:
\-1 - a < b
0 - a = b
1 - a > b
***************************************************************************/
signed char stricmppgm2ram(BYTE* a, ROM BYTE* b)
{
BYTE cA, cB;
// Load first two characters
cA = *a;
cB = *b;
// Loop until one string terminates
while(cA != '\0' && cB != '\0')
{
// Shift case if necessary
if(cA >= 'a' && cA <= 'z')
cA -= 'a' - 'A';
if(cB >= 'a' && cB <= 'z')
cB -= 'a' - 'A';
// Compare
if(cA > cB)
return 1;
if(cA < cB)
return -1;
// Characters matched, so continue
a++;
b++;
cA = *a;
cB = *b;
}
// See if one string terminated first
if(cA > cB)
return 1;
if(cA < cB)
return -1;
// Strings match
return 0;
}
/*****************************************************************************
Function:
WORD swaps(WORD v)
Description:
Swaps the endian-ness of a WORD.
Precondition:
None
Parameters:
v - the WORD to swap
Returns:
The swapped version of v.
***************************************************************************/
WORD swaps(WORD v)
{
WORD_VAL t;
BYTE b;
t.Val = v;
b = t.v[1];
t.v[1] = t.v[0];
t.v[0] = b;
return t.Val;
}
/*****************************************************************************
Function:
DWORD swapl(DWORD v)
Description:
Swaps the endian-ness of a DWORD.
Precondition:
None
Parameters:
v - the DWORD to swap
Returns:
The swapped version of v.
***************************************************************************/
DWORD swapl(DWORD v)
{
// Swap bytes 0 and 3
((DWORD_VAL*)&v)->v[0] ^= ((DWORD_VAL*)&v)->v[3];
((DWORD_VAL*)&v)->v[3] ^= ((DWORD_VAL*)&v)->v[0];
((DWORD_VAL*)&v)->v[0] ^= ((DWORD_VAL*)&v)->v[3];
// Swap bytes 1 and 2
((DWORD_VAL*)&v)->v[1] ^= ((DWORD_VAL*)&v)->v[2];
((DWORD_VAL*)&v)->v[2] ^= ((DWORD_VAL*)&v)->v[1];
((DWORD_VAL*)&v)->v[1] ^= ((DWORD_VAL*)&v)->v[2];
return v;
}
/*****************************************************************************
Function:
WORD CalcIPChecksum(BYTE* buffer, WORD count)
Summary:
Calculates an IP checksum value.
Description:
This function calculates an IP checksum over an array of input data. The
checksum is the 16-bit one's complement of one's complement sum of all
words in the data (with zero-padding if an odd number of bytes are
summed). This checksum is defined in RFC 793.
Precondition:
buffer is WORD aligned (even memory address) on 16- and 32-bit PICs.
Parameters:
buffer - pointer to the data to be checksummed
count - number of bytes to be checksummed
Returns:
The calculated checksum.
Internal:
This function could be improved to do 32-bit sums on PIC32 platforms.
***************************************************************************/
WORD CalcIPChecksum(BYTE* buffer, WORD count)
{
WORD i;
WORD *val;
union
{
WORD w[2];
DWORD dw;
} sum;
i = count >> 1;
val = (WORD*)buffer;
// Calculate the sum of all words
sum.dw = 0x00000000ul;
while(i--)
sum.dw += (DWORD)*val++;
// Add in the sum of the remaining byte, if present
if(count & 0x1)
sum.dw += (DWORD)*(BYTE*)val;
// Do an end-around carry (one's complement arrithmatic)
sum.dw = sum.w[0] + sum.w[1];
// Do another end-around carry in case if the prior add
// caused a carry out
sum.w[0] += sum.w[1];
// Return the resulting checksum
return ~sum.w[0];
}
/*****************************************************************************
Function:
WORD CalcIPBufferChecksum(WORD len)
Summary:
Calculates an IP checksum in the MAC buffer itself.
Description:
This function calculates an IP checksum over an array of input data
existing in the MAC buffer. The checksum is the 16-bit one's complement
of one's complement sum of all words in the data (with zero-padding if
an odd number of bytes are summed). This checksum is defined in RFC 793.
Precondition:
TCP is initialized and the MAC buffer pointer is set to the start of
the buffer.
Parameters:
len - number of bytes to be checksummed
Returns:
The calculated checksum.
Remarks:
All Microchip MACs should perform this function in hardware.
***************************************************************************/
#if defined(NON_MCHP_MAC)
WORD CalcIPBufferChecksum(WORD len)
{
DWORD_VAL Checksum = {0x00000000ul};
WORD ChunkLen;
BYTE DataBuffer[20]; // Must be an even size
WORD *DataPtr;
while(len)
{
// Obtain a chunk of data (less SPI overhead compared
// to requesting one byte at a time)
ChunkLen = len > sizeof(DataBuffer) ? sizeof(DataBuffer) : len;
MACGetArray(DataBuffer, ChunkLen);
len -= ChunkLen;
// Take care of a last odd numbered data byte
if(((WORD_VAL*)&ChunkLen)->bits.b0)
{
DataBuffer[ChunkLen] = 0x00;
ChunkLen++;
}
// Calculate the checksum over this chunk
DataPtr = (WORD*)&DataBuffer[0];
while(ChunkLen)
{
Checksum.Val += *DataPtr++;
ChunkLen -= 2;
}
}
// Do an end-around carry (one's complement arrithmatic)
Checksum.Val = (DWORD)Checksum.w[0] + (DWORD)Checksum.w[1];
// Do another end-around carry in case if the prior add
// caused a carry out
Checksum.w[0] += Checksum.w[1];
// Return the resulting checksum
return ~Checksum.w[0];
}
#endif
/*****************************************************************************
Function:
char* strupr(char* s)
Summary:
Converts a string to uppercase.
Description:
This function converts strings to uppercase on platforms that do not
already have this function defined. All lower-case characters are
converted, an characters not included in 'a'-'z' are left as-is.
Precondition:
None
Parameters:
s - the null-terminated string to be converted.
Returns:
Pointer to the initial string.
***************************************************************************/
#if !defined(__18CXX) || defined(HI_TECH_C)
char* strupr(char* s)
{
char c;
char *t;
t = s;
while( (c = *t) )
{
if(c >= 'a' && c <= 'z')
{
*t -= ('a' - 'A');
}
t++;
}
return s;
}
#endif
#if defined(__18CXX)
// Make this variable global for the following function.
// Hi-Tech PICC18 cannot access local function variables from inline asm.
DWORD_VAL toRotate;
#endif
/*****************************************************************************
Function:
DWORD leftRotateDWORD(DWORD val, BYTE bits)
Summary:
Left-rotates a DWORD.
Description:
This function rotates the bits in a 32-bit DWORD left by a specific
number of bits.
Precondition:
None
Parameters:
val - the DWORD to be rotated
bits - the number of bits by which to shift
Returns:
Rotated DWORD value.
Remarks:
This function is only implemented on 8-bit platforms for now. The
8-bit compilers generate excessive code for this function, while C30
and C32 already generate compact code. Those compilers are served
by a macro defined in Helpers.h.
***************************************************************************/
#if defined(__18CXX)
DWORD leftRotateDWORD(DWORD val, BYTE bits)
{
BYTE i, t;
//DWORD_VAL toRotate;
toRotate.Val = val;
for(i = bits; i >= 8u; i -= 8)
{
t = toRotate.v[3];
toRotate.v[3] = toRotate.v[2];
toRotate.v[2] = toRotate.v[1];
toRotate.v[1] = toRotate.v[0];
toRotate.v[0] = t;
}
#if defined(HI_TECH_C)
for(; i != 0; i--)
{
asm("movlb (_toRotate)>>8");
//asm("bcf _STATUS,0,C");
asm("bcf 0xFD8,0,C"); // HI-TECH PICC-18 PRO 9.63PL1 doesn't define _STATUS
asm("btfsc (_toRotate)&0ffh+3,7,B");
//asm("bsf _STATUS,0,C");
asm("bsf 0xFD8,0,C"); // HI-TECH PICC-18 PRO 9.63PL1 doesn't define _STATUS
asm("rlcf (_toRotate)&0ffh+0,F,B");
asm("rlcf (_toRotate)&0ffh+1,F,B");
asm("rlcf (_toRotate)&0ffh+2,F,B");
asm("rlcf (_toRotate)&0ffh+3,F,B");
}
#else
for(; i != 0u; i--)
{
_asm
movlb toRotate
bcf STATUS,0,0
btfsc toRotate+3,7,1
bsf STATUS,0,0
rlcf toRotate+0,1,1
rlcf toRotate+1,1,1
rlcf toRotate+2,1,1
rlcf toRotate+3,1,1
_endasm
}
#endif
return toRotate.Val;
}
#endif
/*****************************************************************************
Function:
void FormatNetBIOSName(BYTE Name[])
Summary:
Formats a string to a valid NetBIOS name.
Description:
This function formats a string to a valid NetBIOS name. Names will be
exactly 16 characters, as defined by the NetBIOS spec. The 16th
character will be a 0x00 byte, while the other 15 will be the
provided string, padded with spaces as necessary.
Precondition:
None
Parameters:
Name - the string to format as a NetBIOS name. This parameter must have
at least 16 bytes allocated.
Returns:
None
***************************************************************************/
void FormatNetBIOSName(BYTE Name[])
{
BYTE i;
Name[15] = '\0';
strupr((char*)Name);
i = 0;
while(i < 15u)
{
if(Name[i] == '\0')
{
while(i < 15u)
{
Name[i++] = ' ';
}
break;
}
i++;
}
}
/*****************************************************************************
Function:
char * strnchr(const char *searchString, size_t count, char c)
Summary:
Searches a string up to a specified number of characters for a specific
character.
Description:
Searches a string up to a specified number of characters for a specific
character. The string is searched forward and the first occurance
location is returned. If the search character is not present in the
string, or if the maximum character count is reached first, then a NULL
pointer is returned.
Precondition:
None
Parameters:
searchString - Pointer to a null terminated string to search. If count is
less than the string size, then the string need not be null terminated.
count - Maximum number of characters to search before aborting.
c - Character to search for
Returns:
Pointer to the first occurance of the character c in the string
searchString. If the character is not found or the maximum count is
reached, a NULL pointer is returned.
***************************************************************************/
char * strnchr(const char *searchString, size_t count, char c)
{
char c2;
while(count--)
{
c2 = *searchString++;
if(c2 == 0u)
return NULL;
if(c2 == c)
return (char*)--searchString;
}
return NULL;
}
/*****************************************************************************
Function:
BYTE ExtractURLFields(BYTE *vURL,
PROTOCOLS *protocol,
BYTE *vUsername, WORD *wUsernameLen,
BYTE *vPassword, WORD *wPasswordLen,
BYTE *vHostname, WORD *wHostnameLen,
WORD *wPort,
BYTE *vFilePath, WORD *wFilePathLen)
Summary:
Extracts all parameters from an URL string (ex:
"http://admin:passwd@www.microchip.com:8080/myfile.gif" is split into
{PROTOCOL_HTTP, "admin", "passwd", "www.microchip.com", 8080, "/myfile.gif"}.
Description:
Extracts all parameters from an URL string (ex:
"http://admin:passwd@www.microchip.com:8080/myfile.gif" is split into
{PROTOCOL_HTTP, "admin", "passwd", "www.microchip.com", 8080, "/myfile.gif"}.
The URL string can be null terminated, or alternatively could be terminated
by a carriage return or line feed.
If the protocol is unrecognized or the protocol is recognized but the URL
is malformed, than an error is safely returned. For more information on
URL/URI interpretation see RFC 2396.
Precondition:
This function is commented out by default to save code space because
it is not used by any current stack features. However, if you want to use
it, go ahead and uncomment it. It has been tested, so it (should) work
correctly.
Parameters:
vURL - Pointer to null terminated URL to decode and extract from. This
parameter is required and needs to have the minimum RFC 1738 components
in it (protocol and hostname).
protocol - Optional pointer to a PROTOCOLS enum to retrieve the decoded
protocol type. If this parameter is unneeded, specify a NULL pointer.
The protocol is a required part of the URL, so it must always be
present. The protocol also determines what scheme all other parameters
are decoded using, so the function will fail if an unrecognized
protocol is provided. The PROTOCOLS enum members show all of the
currently supported protocols for this function.
<p>For the example URL provided in the function description,
PROTOCOL_HTTP would be returned for this field.
vUsername - Optional pointer to a buffer to write the decoded username
portion of the URL. If the URL does not contain a username or a NULL
pointer is supplied, then this field is ignored.
<p>For the example URL provided in the function description, "admin"
would be returned for this field.
wUsernameLen -
On call\: Optional pointer to a WORD specifying the maximum length of
the vUsername buffer, including the null terminator character.
<p>Upon return\: If wUsernameLen and vUsername are non-NULL, the
*wUsernameLen WORD is updated with the actual number of characters
written to the vUsername buffer, including the null terminator
character. If vUsername is NULL but wUsernameLen is non-NULL, then no
characters are copied, but *wUsernameLen will return the number of
characters required to fit the full username string. If wUsernameLen
is NULL, then the username field in the URL, if present, is ignored and
the vUsername pointer is not used.
<p>If zero characters were written, this indicates that the URL did not
contain a username field. If one character was written, this indicates
that a username field was present, but was a zero character string
(ex\: "").
<p>For the example URL provided in the function description, 6 (0x0006)
would be returned for this field.
vPassword - Optional pointer to a buffer to write the decoded password
portion of the URL. If the URL does not contain a password or a NULL
pointer is supplied, then this field is ignored.
<p>For the example URL provided in the function description, "passwd"
would be returned for this field.
wPasswordLen -
On call\: Optional pointer to a WORD specifying the maximum length of
the vPassword buffer, including the null terminator character.
<p>Upon return\: If wPasswordLen and vPassword are non-NULL, the
*wPasswordLen WORD is updated with the actual number of characters
written to the vPassword buffer, including the null terminator
character. If vPassword is NULL but wPasswordLen is non-NULL, then no
characters are copied, but *wPasswordLen will return the number of
characters required to fit the full password string. If wPasswordLen
is NULL, then the password field in the URL, if present, is ignored and
the vPassword pointer is not used.
<p>If zero characters were written, this indicates that the URL did not
contain a password field. If one character was written, this indicates
that a password field was present, but was a zero character string
(ex\: "").
<p>For the example URL provided in the function description, 7 (0x0007)
would be returned for this field.
vHostname - Optional pointer to a buffer to write the decoded hostname
portion of the URL. All Internet URLs must contain a hostname or IP
address, however, if a NULL pointer is supplied, then this field is
ignored.
<p>For the example URL provided in the function description,
"www.microchip.com" would be returned for this field. If the URL was
"http://192.168.0.1", then this field would be returned as
"192.168.0.1". The IP address would not be decoded to a DWORD (use the
StringToIPAddress() helper function to do this).
wHostnameLen -
On call\: Optional pointer to a WORD specifying the maximum length of
the vHostname buffer, including the null terminator character.
<p>Upon return\: If wHostnameLen and vHostname are non-NULL, the
*wHostnameLen WORD is updated with the actual number of characters
written to the vHostname buffer, including the null terminator
character. If vHostname is NULL but wHostnameLen is non-NULL, then no
characters are copied, but *wHostnameLen will return the number of
characters required to fit the full hostname string. If wHostnameLen
is NULL, then the hostname field in the URL, is ignored and the
vHostname pointer is not used.
<p>For the example URL provided in the function description,
18 (0x0012) would be returned for this field. If the URL was
"http://192.168.0.1", then this field would be returned as 12 (0x000C).
wPort - Optional pointer to a WORD specifying the TCP or UDP port that the
server is listening on. If the port field is absent from the URL, then
this parameter will specify the default port for the protocol. For
example, "http://www.microchip.com" would result in 80 being return as
the specified port.
<p>If the wPort pointer is NULL, then the port field in the URL
is ignored, if present.
vFilePath - Optional pointer to a buffer to write the decoded file path
portion of the URL. If a NULL pointer is supplied, then this field is
ignored. If a file path is not present in the URL, then "/" will be
returned in this field.
<p>For the example URL provided in the function description,
"/myfile.gif" would be returned for this field.
wFilePathLen -
On call\: Optional pointer to a WORD specifying the maximum length of
the vFilePath buffer, including the null terminator character.
<p>Upon return\: If wFilePathLen and vFilePath are non-NULL, the
*wFilePathLen WORD is updated with the actual number of characters
written to the vFilePath buffer, including the null terminator
character. If vFilePath is NULL but wFilePathLen is non-NULL, then no
characters are copied, but *wFilePathLen will return the number of
characters required to fit the full file path string. If wFilePathLen
is NULL, then the file path field in the URL, if present, is ignored and
the vFilePath pointer is not used.
<p>This function always returns "/" if no file path is present, so
*wFilePathLen will also be at least 2 characters ('/' and null
terminator) if the pointer is non-NULL.
<p>For the example URL provided in the function description, 12 (0x000C)
would be returned for this field.
Returns:
Zero on success. Nonzero indicates an error code. If a nonzero error code
is returned, none of the returned buffers or pointer values should be
treated as valid, but some of them may have been written to. The following
are all possible return values.
<table>
0 No error
1 Protocol unknown (additional code needs to be added to
ExtractURLFields() and the PROTOCOLS enum needs to be updated if
you want to decode URLs of this protocol type.
2 URL malformed. Illegal or unknown URL format encountered.
3 Buffer too small. One of the input buffer sizes is too small to
contain the URL parameter.
</table>
***************************************************************************/
#if 0
BYTE ExtractURLFields(BYTE *vURL, PROTOCOLS *protocol, BYTE *vUsername, WORD *wUsernameLen, BYTE *vPassword, WORD *wPasswordLen, BYTE *vHostname, WORD *wHostnameLen, WORD *wPort, BYTE *vFilePath, WORD *wFilePathLen)
{
// These two arrays must exactly match up each other and the PROTOCOLS enum
// elements. The protocol name strings must also be specified in all
// lowercase.
static ROM char * ROM vProtocolNames[] = {"http", "https", "mms", "rtsp"};
static ROM WORD wProtocolPorts[] = { 80, 443, 1755, 554};
WORD w, w2;
BYTE i, j;
PROTOCOLS prot;
BYTE *temp, *temp2;
WORD wURLLen;
WORD wLocalPort;
// Calculate how long this URL is
wURLLen = strlen((char*)vURL);
temp = (BYTE*)strnchr((char*)vURL, wURLLen, '\r');
if(temp)
wURLLen = temp - vURL;
temp = (BYTE*)strnchr((char*)vURL, wURLLen, '\n');
if(temp)
wURLLen = temp - vURL;
// Parse starting protocol field
// Find out how long the protocol name field is
temp = (BYTE*)strnchr((char*)vURL, wURLLen, ':');
if(temp == NULL)
return 2;
// Search protocol list to see if this is a recognized protocol
for(prot = 0; (BYTE)prot < sizeof(wProtocolPorts)/sizeof(wProtocolPorts[0]); prot++)
{
w = strlenpgm(vProtocolNames[prot]);
if((WORD)(temp - vURL) == w)
{
w2 = 0;
temp2 = vURL;
while(w)
{
i = *temp2++;
if((i >= 'A') && (i <= 'Z'))
i += 'a' - 'A';
if(i != (BYTE)vProtocolNames[prot][w2++])
break;
w--;
}
if(w == 0u)
{
if(protocol)
*protocol = prot;
break;
}
}
}
// If we've search the whole list and didn't find a match, then
// this protocol is unknown and this URL cannot be parsed.
if((BYTE)prot >= sizeof(wProtocolPorts)/sizeof(wProtocolPorts[0]))
return 1;
w = temp - vURL + 1;
vURL += w;
wURLLen -= w;
// Protocols using the authority field all must have a double
// slash "//" prefix
if(wURLLen < 2u)
return 2;
for(j = 0; j < 2u; j++)
{
i = *vURL++;
if(i != '/')
return 2;
}
wURLLen -= 2;
// Parse username and password fields
// See if there is a @ sign, indicating that there is at
// least a username and possibly a password in this URL
temp = (BYTE*)strnchr((char*)vURL, wURLLen, '@');
if(temp == NULL)
{
if(wUsernameLen)
*wUsernameLen = 0;
if(wPasswordLen)
*wPasswordLen = 0;
}
else
{
// If we get down here, there is a user name present, let's
// see if a password is also present by searching for a
// colon between the current string position and the @
// symbol.
temp2 = (BYTE*)strnchr((char*)vURL, temp - vURL, ':');
// Calculate username length and password length, including
// null terminator (if the field exists)
if(temp2 == NULL)
{
w = temp - vURL + 1; // Username
w2 = 0; // Password
}
else
{
w = temp2 - vURL + 1; // Username
w2 = temp - temp2; // Password
}
if(wUsernameLen)
{
if(vUsername)
{
if(*wUsernameLen < w)
return 3;
memcpy((void*)vUsername, (void*)vURL, w - 1);
vUsername[w-1] = 0;
}
*wUsernameLen = w;
}
if(wPasswordLen)
{
if(vPassword)
{
if(*wPasswordLen < w2)
return 3;
if(w2)
{
memcpy((void*)vPassword, (void*)temp2+1, w2 - 1);
vPassword[w2-1] = 0;
}
}
*wPasswordLen = w2;
}
vURL += w;
wURLLen -= w;
if(w2)
{
vURL += w2;
wURLLen -= w2;
}
}
// Parse hostname field
// Find the length of the hostname, including NULL
// terminator
temp = (BYTE*)strnchr((char*)vURL, wURLLen, ':');
temp2 = (BYTE*)strnchr((char*)vURL, wURLLen, '/');
if(temp && temp2)
{
if(temp > temp2)
temp = NULL;
}
if(temp == NULL)
{
temp = temp2;
if(temp2 == NULL)
temp = vURL + wURLLen;
}
w = temp - vURL + 1;
if(wHostnameLen)
{
if(vHostname)
{
if(*wHostnameLen < w)
return 3;
memcpy((void*)vHostname, (void*)vURL, w - 1);
vHostname[w-1] = 0;
}
*wHostnameLen = w;
}
vURL += w - 1;
wURLLen -= w - 1;
// Parse port field
if(*vURL == ':')
{
vURL++;
wURLLen--;
wLocalPort = 0;
w = wURLLen;
temp = (BYTE*)strnchr((char*)vURL, wURLLen, '/');
if(temp != NULL)
w = temp - vURL;
w2 = w;
if(wPort)
{
while(w--)
{
wLocalPort *= 10;
wLocalPort += *vURL++ - '0';
}
*wPort = wLocalPort;
}
else
vURL += w2;
wURLLen -= w2;
}
else if(wPort)
*wPort = wProtocolPorts[prot];
// Parse file path field
if(wFilePathLen)
{
w = ++wURLLen;
if(wURLLen == 1u)
w = 2;
if(vFilePath)
{
if(*wFilePathLen < w)
return 3;
if(wURLLen == 1u)
vFilePath[0] = '/';
else
memcpy((void*)vFilePath, (void*)vURL, wURLLen - 1);
vFilePath[w - 1] = 0;
*wFilePathLen = w;
return 0;
}
*wFilePathLen = w;
}
return 0;
}
#endif
/*****************************************************************************
Function:
SHORT Replace(BYTE *vExpression, ROM BYTE *vFind, ROM BYTE *vReplacement,
WORD wMaxLen, BOOL bSearchCaseInsensitive)
Summary:
Replaces all instances of a particular substring with a new string
Description:
Searches a string (vExpression) and replaces all instances of a particular
substring (vFind) with a new string (vReplacement). The start offset to
being searching and a maximum number of replacements can be specified. The
search can be performed in a case sensitive or case insensitive manner.
Precondition:
This function is commented out by default to save code space because
it is not used by any current stack features. However, if you want to use
it, go ahead and uncomment it. It has been tested, so it (should) work
correctly.
Parameters:
vExpression - Null terminated string to search and make replacements within.
vFind - Null terminated string to search for.
vReplacement - Null terminated string to replace all instances of vFind with.
wMaxLen - Maximum length of the output vExpression string if string
expansion is going to occur (replacement length is longer than find
length). If the replacements will cause this maximum string length to
be exceeded, then no replacements will be made and a negative result
will be returned, indicating failure. If the replacement length is
shorter or equal to the search length, then this parameter is ignored.
bSearchCaseInsensitive - Boolean indicating if the search should be
performed in a case insensitive manner. Specify TRUE for case
insensitive searches (slower) or FALSE for case sensitive
searching (faster).
Remarks:
If the replacement string length is shorter than or equal to the search
string length and the search string occurs in multiple overlapping
locations (ex\: expression is "aaa", find is "aa", and replacement is "bb")
then the first find match occuring when searching from left to right will
be replaced. (ex\: output expression will be "bba").
However, if the replacement string length is longer than the search string
length, the search will occur starting from the end of the string and
proceed to the beginning (right to left searching). In this case if the
expression was "aaa", find was "aa", and replacement was "bbb", then the
final output expression will be "abbb".
Returns:
If zero or greater, indicates the count of how many replacements were made.
If less than zero (negative result), indicates that wMaxLen was too small
to make the necessary replacements. In this case, no replacements were
made.
***************************************************************************/
#if 0
SHORT Replace(BYTE *vExpression, ROM BYTE *vFind, ROM BYTE *vReplacement, WORD wMaxLen, BOOL bSearchCaseInsensitive)
{
WORD wExpressionLen, wFindLen, wFindLenMinusOne, wReplacementLen;
WORD wFindCount, wReplacementsLeft;
BYTE i, j;
BYTE vFirstFindChar;
WORD wBytesLeft;
BYTE *vDest;
BYTE *vExpressionCompare;
ROM BYTE *vFindCompare;
WORD w;
wFindLen = strlenpgm((ROM char*)vFind);
if(wFindLen == 0u)
return 0;
wExpressionLen = strlen((char*)vExpression);
wReplacementLen = strlenpgm((ROM char*)vReplacement);
wFindCount = 0;
wFindLenMinusOne = wFindLen - 1;
vFirstFindChar = *vFind++;
if(bSearchCaseInsensitive) // Convert to all lowercase if needed
if((vFirstFindChar >= (BYTE)'A') && (vFirstFindChar <= (BYTE)'Z'))
vFirstFindChar += 'a' - 'A';
// If the replacement string is the same length as the search string, then
// we can immediately do the needed replacements inline and return.
if(wFindLen == wReplacementLen)
{
for(wBytesLeft = wExpressionLen; wBytesLeft; wBytesLeft--)
{
i = *vExpression++;
if(bSearchCaseInsensitive)
{
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if(i != vFirstFindChar)
continue;
vExpressionCompare = vExpression;
vFindCompare = vFind;
w = wFindLenMinusOne;
while(w)
{
i = *vExpressionCompare++;
j = *vFindCompare++;
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if((j >= (BYTE)'A') && (j <= (BYTE)'Z'))
j += 'a' - 'A';
if(i != j)
break;
w--;
}
if(w)
continue;
}
else
{
if(i != vFirstFindChar)
continue;
if(memcmppgm2ram((void*)vExpression, (ROM void*)vFind, wFindLenMinusOne))
continue;
}
memcpypgm2ram((void*)vExpression-1, (ROM void*)vReplacement, wReplacementLen);
wFindCount++;
vExpression += wFindLenMinusOne;
wBytesLeft -= wFindLenMinusOne;
}
return wFindCount;
}
// If the replacement string is shorter than the search string, then we can
// search from left to right and move the string over as we find occurrences.
if(wFindLen > wReplacementLen)
{
vDest = vExpression;
for(wBytesLeft = wExpressionLen; wBytesLeft; wBytesLeft--)
{
i = *vExpression++;
*vDest++ = i;
if(bSearchCaseInsensitive)
{
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if(i != vFirstFindChar)
continue;
vExpressionCompare = vExpression;
vFindCompare = vFind;
w = wFindLenMinusOne;
while(w)
{
i = *vExpressionCompare++;
j = *vFindCompare++;
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if((j >= (BYTE)'A') && (j <= (BYTE)'Z'))
j += 'a' - 'A';
if(i != j)
break;
w--;
}
if(w)
continue;
}
else
{
if(i != vFirstFindChar)
continue;
if(memcmppgm2ram((void*)vExpression, (ROM void*)vFind, wFindLenMinusOne))
continue;
}
memcpypgm2ram((void*)vDest-1, (ROM void*)vReplacement, wReplacementLen);
vDest += wReplacementLen-1;
wFindCount++;
vExpression += wFindLenMinusOne;
wBytesLeft -= wFindLenMinusOne;
}
*vDest = 0x00; // Write new null terminator since the string may have shrunk
return wFindCount;
}
// If the replacement string is longer than the search string, then we will
// take a two pass approach. On the first pass, we will merely count how
// many replacements to make. With this we can calculate how long the
// final string is going to be. On the second pass, we will search from
// right to left and expand the string as needed.
// Pass 1: count how many occurrences of vFind are in vExpression
for(wBytesLeft = wExpressionLen; wBytesLeft; wBytesLeft--)
{
i = *vExpression++;
if(bSearchCaseInsensitive)
{
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if(i != vFirstFindChar)
continue;
vExpressionCompare = vExpression;
vFindCompare = vFind;
w = wFindLenMinusOne;
while(w)
{
i = *vExpressionCompare++;
j = *vFindCompare++;
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if((j >= (BYTE)'A') && (j <= (BYTE)'Z'))
j += 'a' - 'A';
if(i != j)
break;
w--;
}
if(w)
continue;
}
else
{
if(i != vFirstFindChar)
continue;
if(memcmppgm2ram((void*)vExpression, (ROM void*)vFind, wFindLenMinusOne))
continue;
}
wFindCount++;
vExpression += wFindLenMinusOne;
wBytesLeft -= wFindLenMinusOne;
}
// Return immediately if no replacements are needed
if(wFindCount == 0u)
return 0;
// Pass 2: make replacements and move string over
vDest = vExpression + wFindCount * (wReplacementLen - wFindLen);
if(vDest > vExpression - wExpressionLen + wMaxLen)
return -1;
*vDest-- = 0x00; // Write new null terminator
vExpression -= 1;
vFind -= 1;
vFirstFindChar = vFind[wFindLenMinusOne];
if(bSearchCaseInsensitive) // Convert to all lowercase if needed
if((vFirstFindChar >= (BYTE)'A') && (vFirstFindChar <= (BYTE)'Z'))
vFirstFindChar += 'a' - 'A';
wReplacementsLeft = wFindCount;
while(wReplacementsLeft)
{
i = *vExpression--;
*vDest-- = i;
if(bSearchCaseInsensitive)
{
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if(i != vFirstFindChar)
continue;
vExpressionCompare = vExpression;
vFindCompare = &vFind[wFindLenMinusOne-1];
w = wFindLenMinusOne;
while(w)
{
i = *vExpressionCompare--;
j = *vFindCompare--;
if((i >= (BYTE)'A') && (i <= (BYTE)'Z'))
i += 'a' - 'A';
if((j >= (BYTE)'A') && (j <= (BYTE)'Z'))
j += 'a' - 'A';
if(i != j)
break;
w--;
}
if(w)
continue;
}
else
{
if(i != vFirstFindChar)
continue;
if(memcmppgm2ram((void*)vExpression-wFindLenMinusOne, (ROM void*)vFind, wFindLenMinusOne))
continue;
}
memcpypgm2ram((void*)vDest-wReplacementLen+2, (ROM void*)vReplacement, wReplacementLen);
vDest -= wReplacementLen-1;
vExpression -= wFindLenMinusOne;
wBytesLeft -= wFindLenMinusOne;
wReplacementsLeft--;
}
return wFindCount;
}
#endif
|