//~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
// PBook::ReadFile(): read in a file.
errorT
PBook::ReadFile ()
{
ASSERT (FileName != NULL);
ReadOnly = false;
MFile fp;
if (fp.Open (FileName, FMODE_Both) != OK) {
ReadOnly = true;
if (fp.Open (FileName, FMODE_ReadOnly) != OK) {
return ERROR_FileOpen;
}
}
LineCount = 1;
Position * pos = new Position;
DString * line = new DString;
ReadLine(line, &fp);
DString dstr;
while (line->Length() || ! fp.EndOfFile()) {
if (pos->ReadFromFEN (line->Data()) != OK) {
fprintf (stderr, "Error reading line: %u\n", LineCount);
LineCount++;
line->Clear();
ReadLine(line, &fp);
continue;
//exit (1);
}
char * s = (char *) line->Data();
// Skip over first four fields, which were the position:
while (*s == ' ') { s++; }
for (uint i=0; i < 4; i++) {
while (*s != ' ' && *s != 0) { s++; }
while (*s == ' ') { s++; }
}
// Now process each field in turn:
while (*s == ';' || *s == ' ') { s++; }
dstr.Clear();
while (*s != 0) {
while (*s == ';' || *s == ' ') { s++; }
bool seenCode = false;
while (*s != ';' && *s != 0) {
seenCode = true;
char ch = *s;
// Check for backslash (escape) character:
if (ch == '\\') {
s++;
ch = *s;
// "\s" -> semicolon within a field:
if (ch == 's') { ch = ';'; }
}
dstr.AddChar (ch);
s++;
}
if (seenCode) { dstr.AddChar ('\n'); }
}
if (Insert (pos, dstr.Data()) != OK) {
//fprintf (stderr, "Warning: position already exists! Line %u\n",
// LineCount);
}
LineCount++;
line->Clear();
ReadLine(line, &fp);
}
delete pos;
delete line;
Altered = false;
NextIndex = NodeListCount - 1;
return OK;
}
int
main (int argc, char * argv[])
{
setbuf(stdout, NULL); // Make stdout unbuffered.
gameNumberT gNumber;
bool option_ForceReplace = false;
bool option_PreGameComments = true;
uint pgnFileSize = 0;
char *progname = argv[0];
fileNameT fname;
fileNameT baseName;
uint argsleft = argc - 1;
char ** nextArg = argv + 1;
// Parse command-line argments:
while (argsleft > 0 && nextArg[0][0] == '-') {
if (! strCompare (*nextArg, "-f")) {
option_ForceReplace = true;
} else if (! strCompare (*nextArg, "-x")) {
option_PreGameComments = false;
} else {
usage (progname);
}
argsleft--;
nextArg++;
}
if (argsleft != 1 && argsleft != 2) {
usage (progname);
}
char * pgnName = *nextArg;
MFile pgnFile;
pgnFileSize = fileSize (pgnName, "");
// Ensure positive file size counter to avoid division by zero:
if (pgnFileSize < 1) {
pgnFileSize = 1;
}
// Make baseName from pgnName if baseName is not provided:
if (argsleft == 1) {
strCopy (baseName, pgnName);
// If a gzip file, remove two suffixes, the first being ".gz":
const char * lastSuffix = strFileSuffix (baseName);
if (lastSuffix != NULL && strEqual (lastSuffix, GZIP_SUFFIX)) {
strTrimFileSuffix (baseName);
}
// Trim the ".pgn" suffix:
strTrimFileSuffix (baseName);
} else {
strCopy (baseName, nextArg[1]);
}
// Check for existing database, avoid overwriting it:
if (! option_ForceReplace) {
if (fileSize (baseName, INDEX_SUFFIX) > 0) {
// Scid index file already exists:
fprintf (stderr, "%s: database already exists: %s\n", progname,
baseName);
fprintf (stderr, "You can use: %s -f %s to overwrite"
" the existing database.\n", progname, pgnName);
exit(1);
}
}
if (pgnFile.Open (pgnName, FMODE_ReadOnly) != OK) {
fprintf (stderr, "%s: could not open file %s\n", progname, pgnName);
exit(1);
}
// Try opening the log file:
strCopy (fname, baseName);
strAppend (fname, ".err");
FILE * logFile = fopen (fname, "w");
if (logFile == NULL) {
fprintf (stderr, "%s: could not open log file: %s\n", progname, fname);
exit(1);
}
printf ("Converting file %s to Scid database %s:\n", pgnName, baseName);
printf ("Errors/warnings will be written to %s.\n\n", fname);
scid_Init();
GFile * gameFile = new GFile;
if ((gameFile->Create (baseName, FMODE_WriteOnly)) != OK) {
fprintf (stderr, "%s: could not create the file %s%s\n",
progname, baseName, GFILE_SUFFIX);
fprintf (stderr, "The file may already exist and be read-only, or\n");
fprintf (stderr, "you may not have permission to create this file.\n");
exit(1);
}
NameBase * nb = new NameBase;
Index * idx = new Index;
IndexEntry * ie = new IndexEntry;
idx->SetFileName (baseName);
idx->CreateIndexFile (FMODE_WriteOnly);
Game * game = new Game;
ProgBar progBar(stdout);
progBar.Start();
ByteBuffer *bbuf = new ByteBuffer;
bbuf->SetBufferSize (BBUF_SIZE); // 32768
PgnParser pgnParser (&pgnFile);
pgnParser.SetErrorFile (logFile);
pgnParser.SetPreGameText (option_PreGameComments);
// TODO: Add command line option for ignored tags, rather than
// just hardcoding PlyCount as the only ignored tag.
pgnParser.AddIgnoredTag ("PlyCount");
// Add each game found to the database:
while (pgnParser.ParseGame(game) != ERROR_NotFound) {
ie->Init();
if (idx->AddGame (&gNumber, ie) != OK) {
fprintf (stderr, "\nLimit of %d games reached!\n", MAX_GAMES);
exit(1);
}
// Add the names to the namebase:
idNumberT id = 0;
if (nb->AddName (NAME_PLAYER, game->GetWhiteStr(), &id) != OK) {
fatalNameError (NAME_PLAYER);
}
nb->IncFrequency (NAME_PLAYER, id, 1);
ie->SetWhite (id);
if (nb->AddName (NAME_PLAYER, game->GetBlackStr(), &id) != OK) {
fatalNameError (NAME_PLAYER);
}
nb->IncFrequency (NAME_PLAYER, id, 1);
ie->SetBlack (id);
if (nb->AddName (NAME_EVENT, game->GetEventStr(), &id) != OK) {
fatalNameError (NAME_EVENT);
}
nb->IncFrequency (NAME_EVENT, id, 1);
ie->SetEvent (id);
if (nb->AddName (NAME_SITE, game->GetSiteStr(), &id) != OK) {
fatalNameError (NAME_SITE);
}
nb->IncFrequency (NAME_SITE, id, 1);
ie->SetSite (id);
if (nb->AddName (NAME_ROUND, game->GetRoundStr(), &id) != OK) {
fatalNameError (NAME_ROUND);
}
nb->IncFrequency (NAME_ROUND, id, 1);
ie->SetRound (id);
bbuf->Empty();
if (game->Encode (bbuf, ie) != OK) {
fprintf (stderr, "Fatal error encoding game!\n");
abort();
}
uint offset = 0;
if (gameFile->AddGame (bbuf, &offset) != OK) {
fprintf (stderr, "Fatal error writing game file!\n");
abort();
}
ie->SetOffset (offset);
ie->SetLength (bbuf->GetByteCount());
idx->WriteEntries (ie, gNumber, 1);
// Update the progress bar:
if (! (gNumber % 100)) {
int bytesSeen = pgnParser.BytesUsed();
int percentDone = 1 + ((bytesSeen) * 100 / pgnFileSize);
progBar.Update (percentDone);
}
}
uint t = 0; // = time(0);
nb->SetTimeStamp(t);
nb->SetFileName (baseName);
if (nb->WriteNameFile() != OK) {
fprintf (stderr, "Fatal error writing name file!\n");
exit(1);
}
progBar.Finish();
printf ("\nDatabase `%s': %d games, %d players, %d events, %d sites.\n",
baseName, idx->GetNumGames(), nb->GetNumNames (NAME_PLAYER),
nb->GetNumNames (NAME_EVENT), nb->GetNumNames (NAME_SITE));
fclose (logFile);
if (pgnParser.ErrorCount() > 0) {
printf ("There were %u errors or warnings; ", pgnParser.ErrorCount());
printf ("examine the file \"%s.err\"\n", baseName);
} else {
printf ("There were no warnings or errors.\n");
removeFile (baseName, ".err");
}
gameFile->Close();
idx->CloseIndexFile();
// If there is a tree cache file for this database, it is out of date:
removeFile (baseName, TREEFILE_SUFFIX);
#ifdef ASSERTIONS
printf("%d asserts were tested\n", numAsserts);
#endif
return 0;
}
//~~~~~~~~~~~~~~~~~~~~~~~~~~~~
// PBook::ReadEcoFile():
// Read an ECO (not EPD) format file.
errorT
PBook::ReadEcoFile ()
{
MFile fp;
if (fp.Open (FileName, FMODE_ReadOnly) != OK) {
return ERROR_FileOpen;
}
ReadOnly = true;
LineCount = 1;
Position std_start;
std_start.StdStart();
DString text;
DString moves;
ecoStringT ecoStr;
ecoT ecoCode;
int ch;
errorT err = OK;
bool done = false;
// Loop to read in and add all positions:
while (!done) {
// Find the next ECO code:
while (true) {
ch = fp.ReadOneByte();
if (ch == EOF) { done = true; break; }
if (ch == '\n') { LineCount++; }
if (ch >= 'A' && ch <= 'E') { break; }
if (ch == '#') {
while (ch != '\n' && ch != EOF) {
ch = fp.ReadOneByte();
}
if (ch == EOF) { done = true; }
LineCount++;
}
}
if (done) { break; }
// Read in the rest of the ECO code:
ecoStr[0] = ch;
ch = fp.ReadOneByte();
if (ch < '0' || ch > '9') { goto corrupt; }
ecoStr[1] = ch;
ch = fp.ReadOneByte();
if (ch < '0' || ch > '9') { goto corrupt; }
ecoStr[2] = ch;
ecoStr[3] = 0;
// Now check for optional extra part of code, e.g. "A00a1":
ch = fp.ReadOneByte();
if (ch >= 'a' && ch <= 'z') {
ecoStr[3] = ch; ecoStr[4] = 0;
ch = fp.ReadOneByte();
if (ch >= '1' && ch <= '4') {
ecoStr[4] = ch; ecoStr[5] = 0;
}
}
// Now put ecoCode in the text string and read the text in quotes:
ecoCode = eco_FromString (ecoStr);
eco_ToExtendedString (ecoCode, ecoStr);
text.Clear();
text.Append ("eco ", ecoStr, " [");
// Find the start of the text:
while ((ch = fp.ReadOneByte()) != '"') {
if (ch == EOF) { goto corrupt; }
}
while ((ch = fp.ReadOneByte()) != '"') {
if (ch == EOF) { goto corrupt; }
text.AddChar ((char) ch);
}
text.Append ("]\n");
// Now read the position:
moves.Clear();
char prev = 0;
while ((ch = fp.ReadOneByte()) != '*') {
if (ch == EOF) { goto corrupt; }
if (ch == '\n') {
ch = ' ';
LineCount++;
}
if (ch != ' ' || prev != ' ') {
moves.AddChar ((char) ch);
}
prev = ch;
}
Position pos (std_start);
err = pos.ReadLine (moves.Data());
if (err != OK) { goto corrupt; }
text.Append ("moves ", strTrimLeft (moves.Data()), "\n");
if (Insert (&pos, text.Data()) != OK) {
// Position already exists: just ignore it.
}
}
return OK;
corrupt:
return ERROR_Corrupt;
}