void read_checkpoint ( char *filename, int *gen, multipop **mpop )
{
FILE *f;
char *buffer;
ephem_const **eind;
int random_state_bytes;
int i, j;
char *rand_state;
/* miscellaneous buffer for reading. */
buffer = (char *)MALLOC ( MAXCHECKLINELENGTH );
/* open the file. */
f = fopen ( filename, "rb" );
if ( f == NULL )
{
error ( E_FATAL_ERROR, "couldn't read checkpoint \"%s\".",
filename );
}
oprintf ( OUT_SYS, 30, "reading from checkpoint \"%s\".\n",
filename );
/** confirm the magic word that starts every checkpoint file. **/
fgets ( buffer, MAXCHECKLINELENGTH, f );
if ( strcmp ( buffer, CK_MAGIC ) )
error ( E_FATAL_ERROR,
"\"%s\" is not a lil-gp v1.0 checkpoint file.", filename );
/* skip the human-readable id line. */
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "id line: %s", buffer );
#endif
/** read and print the timestamp. **/
fscanf ( f, "%*s " );
fgets ( buffer, MAXCHECKLINELENGTH, f );
/* chop the newline. */
buffer[strlen(buffer)-1] = 0;
oprintf ( OUT_SYS, 30, " checkpoint timestamp: [%s].\n", buffer );
/* skip the "section: global" line. */
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "should be global section: %s", buffer );
#endif
/* read the generation number. */
fscanf ( f, "%*s %d\n", gen );
/** read the random number state encoded as a string of hex chars. **/
/* first read the length. */
fscanf ( f, "%*s %d ", &random_state_bytes );
#ifdef DEBUG
fprintf ( stderr, "%d random state bytes.\n", random_state_bytes );
#endif
/* allocate the buffer. */
rand_state = (char *)MALLOC ( random_state_bytes+1 );
/* read the hex data into the buffer. */
read_hex_block ( rand_state, random_state_bytes, f );
/* set the state. */
random_set_state ( rand_state );
/* free the buffer. */
FREE ( rand_state );
/* slurp the newline character following the hex data. */
fgetc ( f );
/** skip the "section: parameter" line. **/
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "should be parameter section: %s", buffer );
#endif
/* read the parameter database. */
read_parameter_database ( f );
/* make internal copies of function set(s). */
if ( app_build_function_sets() )
error ( E_FATAL_ERROR, "app_build_function_sets() failure." );
/** skip the "section: erc" line. **/
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "should be erc section: %s", buffer );
#endif
/* read the list of ephemeral constants, and index them */
eind = read_ephem_list ( f );
/** skip the "section: erc" line. **/
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "should be population section: %s", buffer );
#endif
/** read the population **/
/* allocate memory. */
*mpop = (multipop *)MALLOC ( sizeof ( multipop ) );
/* read number of subpops. */
fscanf ( f, "%*s %d\n", &((**mpop).size) );
/* allocate subpop list. */
(**mpop).pop = (population **)MALLOC ( (**mpop).size *
sizeof ( population * ) );
for ( i = 0; i < (**mpop).size; ++i )
{
/** skip each "subpop: #" line. **/
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "should be subpop %d: %s", i, buffer );
#endif
/* read the population. */
(**mpop).pop[i] = read_population ( eind, f );
}
/** skip the "section: application" line. **/
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "should be application section: %s", buffer );
#endif
/* read application-specific stuff. */
app_read_checkpoint ( f );
/** skip the "section: statistics" line. **/
fgets ( buffer, MAXCHECKLINELENGTH, f );
#ifdef DEBUG
printf ( "should be statistics section: %s", buffer );
#endif
/* read the statistics. */
read_stats_checkpoint ( *mpop, eind, f );
/* close'n'free. */
FREE ( eind );
FREE ( buffer );
fclose ( f );
oprintf ( OUT_SYS, 30, "population read from checkpoint \"%s\".\n",
filename );
}
int main(int argc, char **argv)
{
extern int optind, opterr;
extern char *optarg;
int optchar;
char *treefile_name = NULL, *genomefile_name = NULL, *outfile_basename = NULL, g_outfile_name[FILENAME_MAX];
FILE *treefile, *genomefile, *outfile;
long generation, g, num_trees, i;
int ret_code;
genomefile_name = NULL;
treefile_name = NULL;
outfile_basename = NULL;
generation = -1;
phyl_init_tree(&phyltree);
while ((optchar = getopt(argc, argv, "t:g:o:h")) != -1)
{
switch (optchar)
{
case 'g':
genomefile_name = optarg;
break;
case 't':
treefile_name = optarg;
break;
case 'o':
outfile_basename = optarg;
break;
case 'h':
printf("distcorr -- a hack to visualize correlation between\n");
printf(" phylogenetic and edit distances\n");
printf("Usage of command line options:\n");
printf("-g <filename>: Specify genome file\n");
printf("-t <filename>: Specify tree file\n");
printf("-o <outfile>: specify output file basename:");
printf("-h: Print this info and exit\n");
exit (EXIT_SUCCESS);
}
}
if (treefile_name == NULL)
{
fprintf(stderr, "no tree file specified -- exit\n");
exit (EXIT_FAILURE);
}
if ((treefile = fopen(treefile_name, "r")) == NULL)
{
fprintf(stderr, "failed to open tree file \"%s\" -- exit\n", treefile_name);
exit (EXIT_FAILURE);
}
if (genomefile_name == NULL)
{
fprintf(stderr, "no tree file specified -- exit\n");
exit (EXIT_FAILURE);
}
if ((genomefile = fopen(genomefile_name, "r")) == NULL)
{
fprintf(stderr, "failed to open genome file \"%s\" -- exit\n", genomefile_name);
exit (EXIT_FAILURE);
}
while (!feof(treefile) && !ferror(treefile) && !feof(genomefile) && !ferror(genomefile))
{
get_line(buf, 256, treefile);
if (feof(treefile) || ferror(treefile))
break;
if (buf[0] != 'g')
{
fprintf(stderr, "error in header of treefile after generation %ld -- exit\n", generation);
exit (EXIT_FAILURE);
}
generation = strtol(buf + 1, (char **) NULL, 10);
get_line(buf, 256, genomefile);
if (feof(genomefile) || ferror(genomefile))
break;
if (buf[0] != 'g')
{
fprintf(stderr, "error in header of genomefile after generation %ld -- exit\n", g);
exit (EXIT_FAILURE);
}
g = strtol(buf + 1, (char **) NULL, 10);
if (g != generation)
{
fprintf(stderr, "treefile: g=%ld, genomefile: g=%ld -- out of sync\n", generation, g);
break;
}
get_line(buf, 256, treefile);
num_trees = strtol(buf, NULL, 10);
if (num_trees < 1)
{
fprintf(stderr, "%ld trees in treefile at generation %ld\n", num_trees, generation);
break;
}
if (num_trees > 1)
{
fprintf(stderr, "distcorr: multiple trees in treefile\n");
for (i = 0; i < num_trees; i++)
{
if ((ret_code = phyl_read_tree(treefile, &phyltree)) < 0)
fprintf(stderr, "error #%d while reading tree #%ld of generation %ld\n", ret_code, i, generation);
phyl_free_tree(&phyltree);
}
read_population(genomefile, g);
free_population();
continue;
}
if ((ret_code = phyl_read_tree(treefile, &phyltree)) < 0)
{
fprintf(stderr, "error #%d while reading tree #%ld of generation %ld\n", ret_code, i, generation);
break;
}
if ((ret_code = read_population(genomefile, g)) < 0)
{
fprintf(stderr, "error %d reading genomes of generation %ld\n", ret_code, g);
phyl_free_tree(&phyltree);
break;
}
if (outfile_basename)
{
sprintf(g_outfile_name, "%s-%06ld-dc.gpd", outfile_basename, generation);
if ((outfile = fopen(g_outfile_name, "w")) == NULL)
{
fprintf(stderr, "failed to open \"%s\" for output -- exit\n", g_outfile_name);
exit (EXIT_FAILURE);
}
}
else
outfile = stdout;
ret_code = distcorr(outfile, generation);
if (ret_code < 0)
fprintf(stderr, "error %d in distcorr\n", ret_code);
if (outfile_basename)
fclose(outfile);
free_population();
phyl_free_tree(&phyltree);
}
fclose(treefile);
fclose(genomefile);
return (EXIT_SUCCESS);
}
