moab
|
Parse a file as space-separated tokens. More...
#include <FileTokenizer.hpp>
Public Member Functions | |
FileTokenizer (std::FILE *file_ptr, ReadUtilIface *read_util_ptr) | |
constructor | |
~FileTokenizer () | |
destructor : closes file. | |
const char * | get_string () |
get next token | |
bool | get_newline () |
check for newline | |
bool | get_doubles (size_t count, double *array) |
Parse a sequence of double values. | |
bool | get_floats (size_t count, float *array) |
Parse a sequence of float values. | |
bool | get_integers (size_t count, int *array) |
Parse a sequence of integer values. | |
bool | get_long_ints (size_t count, long *array) |
Parse a sequence of integer values. | |
bool | get_short_ints (size_t count, short *array) |
Parse a sequence of integer values. | |
bool | get_bytes (size_t count, unsigned char *array) |
Parse a sequence of integer values. | |
bool | get_binary (size_t bytes, void *mem) |
Read binary data (interleaved with ASCII) | |
bool | get_booleans (size_t count, bool *array) |
Parse a sequence of bit or boolean values. | |
bool | eof () const |
int | line_number () const |
void | unget_token () |
bool | match_token (const char *string, bool print_error=true) |
int | match_token (const char *const *string_list, bool print_error=true) |
Private Member Functions | |
bool | get_double_internal (double &result) |
bool | get_long_int_internal (long &result) |
bool | get_boolean_internal (bool &result) |
bool | get_float_internal (float &result) |
bool | get_integer_internal (int &result) |
bool | get_short_int_internal (short &result) |
bool | get_byte_internal (unsigned char &result) |
Private Attributes | |
std::FILE * | filePtr |
ReadUtilIface * | readUtilPtr |
char | buffer [512] |
char * | nextToken |
char * | bufferEnd |
int | lineNumber |
char | lastChar |
Parse a file as space-separated tokens.
Read a file, separating it into space-separated tokens. This is provided in place of using the standard C or C++ file parsing routines because it counts lines, which is useful for error reporting. Also provides some useful utility methods for parsing VTK files (which is the intended use of this implementation.)
Uses raw reads/writes, implementing internal buffering. Token size may not exceed buffer size.
Definition at line 43 of file FileTokenizer.hpp.
moab::FileTokenizer::FileTokenizer | ( | std::FILE * | file_ptr, |
ReadUtilIface * | read_util_ptr | ||
) |
constructor
file_ptr | The file to read from. |
read_util_ptr | Pointer to ReadUtilIface to use for reporting errors. |
Definition at line 27 of file FileTokenizer.cpp.
: filePtr( file_ptr ), readUtilPtr( rif_ptr ), nextToken( buffer ), bufferEnd( buffer ), lineNumber( 1 ), lastChar( '\0' ) {}
destructor : closes file.
The destructor closes the passed file handle. This is done as a convenience feature. If the caller creates an instance of this object on the stack, the file will automatically be closed when the caller returns.
Definition at line 36 of file FileTokenizer.cpp.
{ fclose( filePtr ); }
bool moab::FileTokenizer::eof | ( | ) | const |
Check for end-of-file condition.
Definition at line 39 of file FileTokenizer.cpp.
bool moab::FileTokenizer::get_binary | ( | size_t | bytes, |
void * | mem | ||
) |
Read binary data (interleaved with ASCII)
Read a block of binary data.
bytes | Number of bytes to read |
mem | Memory address at which to store data. |
Definition at line 467 of file FileTokenizer.cpp.
{ // if data in buffer if (nextToken != bufferEnd) { // if requested size is less than buffer contents, // just pass back part of the buffer if (bufferEnd - nextToken <= (int)size) { memcpy( mem, nextToken, size ); nextToken += size; return true; } // copy buffer contents into memory and clear buffer memcpy( mem, nextToken, bufferEnd - nextToken ); size -= bufferEnd - nextToken; mem = reinterpret_cast<char*>(mem) + (bufferEnd - nextToken); nextToken = bufferEnd; } // read any additional data from file return size == fread( mem, 1, size, filePtr ); }
bool moab::FileTokenizer::get_boolean_internal | ( | bool & | result | ) | [private] |
Internal implementation of get_Booleans
Definition at line 248 of file FileTokenizer.cpp.
{ // Get a token const char *token = get_string( ); if (!token) return false; if (token[1] || (token[0] != '0' && token[0] != '1')) { readUtilPtr->report_error( "Syntax error at line %d: expected 0 or 1, got \"%s\"", line_number(), token ); return false; } result = token[0] == '1'; return true; }
bool moab::FileTokenizer::get_booleans | ( | size_t | count, |
bool * | array | ||
) |
Parse a sequence of bit or boolean values.
Read the specified number of space-delimited values.
count | The number of values to read. |
array | The memory at which to store the values. |
Definition at line 334 of file FileTokenizer.cpp.
{ for (size_t i = 0; i < count; ++i) { if (!get_boolean_internal( *array )) return false; ++array; } return true; }
bool moab::FileTokenizer::get_byte_internal | ( | unsigned char & | result | ) | [private] |
Internal implementation of get_bytes
Definition at line 200 of file FileTokenizer.cpp.
{ long i; if (!get_long_int_internal( i )) return false; result = (unsigned char)i; if (i != (long)result) { readUtilPtr->report_error( "Numberic overflow at line %d.", line_number() ); return false; } return true; }
bool moab::FileTokenizer::get_bytes | ( | size_t | count, |
unsigned char * | array | ||
) |
Parse a sequence of integer values.
Read the specified number of space-delimited ints.
count | The number of values to read. |
array | The memory at which to store the values. |
Definition at line 289 of file FileTokenizer.cpp.
{ for (size_t i = 0; i < count; ++i) { if (!get_byte_internal( *array )) return false; ++array; } return true; }
bool moab::FileTokenizer::get_double_internal | ( | double & | result | ) | [private] |
Internal implementation of get_doubles
Definition at line 130 of file FileTokenizer.cpp.
{ // Get a token const char *token_end, *token = get_string( ); if (!token) return false; // Check for hex value -- on some platforms (e.g. Linux), strtod // will accept hex values, on others (e.g. Sun) it wil not. Force // failure on hex numbers for consistancy. if (token[0] && token[1] && token[0] == '0' && toupper(token[1]) == 'X') { readUtilPtr->report_error( "Syntax error at line %d: expected number, got \"%s\"", line_number(), token ); return false; } // Parse token as double result = strtod( token, (char**)&token_end ); // If the one past the last char read by strtod is // not the NULL character terminating the string, // then parse failed. if (*token_end) { readUtilPtr->report_error( "Syntax error at line %d: expected number, got \"%s\"", line_number(), token ); return false; } return true; }
bool moab::FileTokenizer::get_doubles | ( | size_t | count, |
double * | array | ||
) |
Parse a sequence of double values.
Read the specified number of space-delimited doubles.
count | The number of values to read. |
array | The memory at which to store the values. |
Definition at line 278 of file FileTokenizer.cpp.
{ for (size_t i = 0; i < count; ++i) { if (!get_double_internal( *array )) return false; ++array; } return true; }
bool moab::FileTokenizer::get_float_internal | ( | float & | result | ) | [private] |
Internal implementation of get_floats
Definition at line 166 of file FileTokenizer.cpp.
{ double d; if (!get_double_internal( d )) return false; result = (float)d; return true; }
bool moab::FileTokenizer::get_floats | ( | size_t | count, |
float * | array | ||
) |
Parse a sequence of float values.
Read the specified number of space-delimited doubles.
count | The number of values to read. |
array | The memory at which to store the values. |
Definition at line 267 of file FileTokenizer.cpp.
{ for (size_t i = 0; i < count; ++i) { if (!get_float_internal( *array )) return false; ++array; } return true; }
bool moab::FileTokenizer::get_integer_internal | ( | int & | result | ) | [private] |
Internal implementation of get_integers
Definition at line 232 of file FileTokenizer.cpp.
{ long i; if (!get_long_int_internal( i )) return false; result = (int)i; if (i != (long)result) { readUtilPtr->report_error( "Numberic overflow at line %d.", line_number() ); return false; } return true; }
bool moab::FileTokenizer::get_integers | ( | size_t | count, |
int * | array | ||
) |
Parse a sequence of integer values.
Read the specified number of space-delimited ints.
count | The number of values to read. |
array | The memory at which to store the values. |
Definition at line 312 of file FileTokenizer.cpp.
{ for (size_t i = 0; i < count; ++i) { if (!get_integer_internal( *array )) return false; ++array; } return true; }
bool moab::FileTokenizer::get_long_int_internal | ( | long & | result | ) | [private] |
Internal implementation of get_long_ints
Definition at line 176 of file FileTokenizer.cpp.
{ // Get a token const char *token_end, *token = get_string( ); if (!token) return false; // Parse token as long result = strtol( token, (char**)&token_end, 0 ); // If the one past the last char read by strtol is // not the NULL character terminating the string, // then parse failed. if (*token_end) { readUtilPtr->report_error( "Syntax error at line %d: expected integer, got \"%s\"", line_number(), token ); return false; } return true; }
bool moab::FileTokenizer::get_long_ints | ( | size_t | count, |
long * | array | ||
) |
Parse a sequence of integer values.
Read the specified number of space-delimited ints.
count | The number of values to read. |
array | The memory at which to store the values. |
Definition at line 323 of file FileTokenizer.cpp.
{ for (size_t i = 0; i < count; ++i) { if (!get_long_int_internal( *array )) return false; ++array; } return true; }
bool moab::FileTokenizer::get_newline | ( | ) |
check for newline
Consume whitespace up to and including the next newline. If a non-space character is found before a newline, the function will stop, set the error message, and return false.
Definition at line 415 of file FileTokenizer.cpp.
{ if (lastChar == '\n') { lastChar = ' '; ++lineNumber; return true; } // Loop until either we a) find a newline, b) find a non-whitespace // character or c) reach the end of the file. for (;;) { // If the buffer is empty, read more. if (nextToken == bufferEnd) { size_t count = fread( buffer, 1, sizeof(buffer), filePtr ); if (!count) { if (eof()) readUtilPtr->report_error( "File truncated at line %d.", line_number() ); else readUtilPtr->report_error( "I/O Error" ); break; } nextToken = buffer; bufferEnd = buffer + count; } // If the current character is not a space, the we've failed. if (!isspace(*nextToken)) { readUtilPtr->report_error( "Expected newline at line %d.", line_number() ); break; } // If the current space character is a newline, // increment the line number count. if (*nextToken == '\n') { ++lineNumber; ++nextToken; lastChar = ' '; return true; } ++nextToken; } return false; }
bool moab::FileTokenizer::get_short_int_internal | ( | short & | result | ) | [private] |
Internal implementation of get_short_ints
Definition at line 216 of file FileTokenizer.cpp.
{ long i; if (!get_long_int_internal( i )) return false; result = (short)i; if (i != (long)result) { readUtilPtr->report_error( "Numberic overflow at line %d.", line_number() ); return false; } return true; }
bool moab::FileTokenizer::get_short_ints | ( | size_t | count, |
short * | array | ||
) |
Parse a sequence of integer values.
Read the specified number of space-delimited ints.
count | The number of values to read. |
array | The memory at which to store the values. |
Definition at line 300 of file FileTokenizer.cpp.
{ for (size_t i = 0; i < count; ++i) { if (!get_short_int_internal( *array )) return false; ++array; } return true; }
const char * moab::FileTokenizer::get_string | ( | ) |
get next token
Get the next whitespace-delimited token from the file. NOTE: The returned string is only valid until the next call to any of the functions in this class that read from the file.
Definition at line 42 of file FileTokenizer.cpp.
{ // If the whitepsace character marking the end of the // last token was a newline, increment the line count. if (lastChar == '\n') ++lineNumber; // Loop until either found the start of a token to return or have // reached the end of the file. for (;;) { // If the buffer is empty, read more. if (nextToken == bufferEnd) { size_t count = fread( buffer, 1, sizeof(buffer) - 1, filePtr ); if (!count) { if (feof(filePtr)) readUtilPtr->report_error( "File truncated at line %d\n", line_number() ); else readUtilPtr->report_error( "I/O Error\n" ); return NULL; } nextToken = buffer; bufferEnd = buffer + count; } // If the current character is not a space, we've found a token. if (!isspace(*nextToken)) break; // If the current space character is a newline, // increment the line number count. if (*nextToken == '\n') ++lineNumber; ++nextToken; } // Store the start of the token in "result" and // advance "nextToken" to one past the end of the // token. char* result = nextToken; while (nextToken != bufferEnd && !isspace(*nextToken)) ++nextToken; // If we have reached the end of the buffer without finding // a whitespace character terminating the token, we need to // read more from the file. Only try once. If the token is // too large to fit in the buffer, give up. if (nextToken == bufferEnd) { // Shift the (possibly) partial token to the start of the buffer. size_t remaining = bufferEnd - result; memmove( buffer, result, remaining ); result = buffer; nextToken = result + remaining; // Fill the remainder of the buffer after the token. size_t count = fread( nextToken, 1, sizeof(buffer) - remaining - 1, filePtr ); if (!count && !feof(filePtr)) { readUtilPtr->report_error( "I/O Error\n" ); return NULL; } bufferEnd = nextToken + count; // Continue to advance nextToken until we find the space // terminating the token. while (nextToken != bufferEnd && !isspace(*nextToken)) ++nextToken; if (nextToken == bufferEnd) // EOF { *bufferEnd = '\0'; ++bufferEnd; } } // Save terminating whitespace character (or NULL char if EOF). lastChar = *nextToken; // Put null in buffer to mark end of current token. *nextToken = '\0'; // Advance nextToken to the next character to search next time. ++nextToken; return result; }
int moab::FileTokenizer::line_number | ( | ) | const [inline] |
Get the line number the last token was read from.
Definition at line 178 of file FileTokenizer.hpp.
{ return lineNumber; }
bool moab::FileTokenizer::match_token | ( | const char * | string, |
bool | print_error = true |
||
) |
Match current token to passed string. If token doesn't match, set error message.
Definition at line 362 of file FileTokenizer.cpp.
{ // Get a token const char *token = get_string( ); if (!token) return false; // Check if it matches if (0 == strcmp( token, str )) return true; // Construct error message if (print_error) readUtilPtr->report_error( "Syntax error at line %d: expected \"%s\", got \"%s\"", line_number(), str, token ); return false; } // namespace Mesquite
int moab::FileTokenizer::match_token | ( | const char *const * | string_list, |
bool | print_error = true |
||
) |
Match the current token to one of an array of strings. Sets the error message if the current token doesn't match any of the input strings.
string_list | A NULL-terminated array of strings. |
Definition at line 381 of file FileTokenizer.cpp.
{ // Get a token const char *token = get_string( ); if (!token) return false; // Check if it matches any input string const char* const* ptr; for (ptr = list; *ptr; ++ptr) if (0 == strcmp( token, *ptr )) return ptr - list + 1; if (!print_error) return false; // No match, constuct error message std::string message( "Parsing error at line " ); char lineno[16]; sprintf( lineno, "%d", line_number() ); message += lineno; message += ": expected one of {"; for (ptr = list; *ptr; ++ptr) { message += " "; message += *ptr; } message += " } got \""; message += token; message += "\""; readUtilPtr->report_error( message ); return false; }
void moab::FileTokenizer::unget_token | ( | ) |
char moab::FileTokenizer::buffer[512] [private] |
Input buffer
Definition at line 227 of file FileTokenizer.hpp.
char* moab::FileTokenizer::bufferEnd [private] |
One past the last used byte of the buffer
Definition at line 232 of file FileTokenizer.hpp.
std::FILE* moab::FileTokenizer::filePtr [private] |
Pointer to standard C FILE struct
Definition at line 221 of file FileTokenizer.hpp.
char moab::FileTokenizer::lastChar [private] |
The whitespace character marking the end of the last returned token. Saved here because if it is a newline, the line count will need to be incremented when the next token is returned.
Definition at line 242 of file FileTokenizer.hpp.
int moab::FileTokenizer::lineNumber [private] |
Line number of last returned token
Definition at line 235 of file FileTokenizer.hpp.
char* moab::FileTokenizer::nextToken [private] |
One past the end of the last token returned
Definition at line 230 of file FileTokenizer.hpp.
ReadUtilIface* moab::FileTokenizer::readUtilPtr [private] |
Pointer to MOAB ReadUtil Interface
Definition at line 224 of file FileTokenizer.hpp.