improved whitespace handling in msi2lmp for force fields and topologies

tools/msi2lmp/README

@@ -1,13 +1,18 @@
 Axel Kohlmeyer is the current maintainer of the msi2lmp tool.
 Please send any inquiries about msi2lmp to the lammps-users mailing list.
 
+06 Oct 2016 Axel Kohlmeyer <akohlmey@gmail.com>
+
+Improved whitespace handling in parsing topology and force field
+files to avoid bogus warnings about type name truncation.
+
 24 Oct 2015 Axel Kohlmeyer <akohlmey@gmail.com>
 
 Added check to make certain that force field files
 are consistent with the notation of non-bonded parameters
 that the msi2lmp code expects. For Class 1 and OPLS-AA
 the A-B notation with geometric mixing is expected and for
-Class 2 the r-eps notation with sixthpower mixing.(
+Class 2 the r-eps notation with sixthpower mixing.
 
 11 Sep 2014 Axel Kohlmeyer <akohlmey@gmail.com>
 

tools/msi2lmp/src/ReadMdfFile.c

@@ -74,9 +74,9 @@ void ReadMdfFile(void)
         at_end = 1;
       } else if (strncmp(line,"@column",7) == 0) {
 
-        temp_string = strtok(line," ");
-        col_no = strtok(NULL," ");
-        col_name = strtok(NULL," ");
+        temp_string = strtok(line,WHITESPACE);
+        col_no = strtok(NULL,WHITESPACE);
+        col_name = strtok(NULL,WHITESPACE);
         if (strncmp(col_name,"charge",6) == 0) {
           if (strlen(col_name) < 8) {
             q_col_no = atoi(col_no);
@@ -285,7 +285,7 @@ int get_molecule(char *line, int connect_col_no, int q_col_no,
     /* Get atom name */
     cur_field = strtok(line,":");
     sscanf(cur_field, "%s", atoms[*counter].residue_string);
-    cur_field = strtok(NULL," ");
+    cur_field = strtok(NULL,WHITESPACE);
     /* Compare atom name with that in .car file */
     if (strcmp(atoms[*counter].name, cur_field)) {
       printf("Names %s from .car file and %s from .mdf file do not match\n",
@@ -297,18 +297,18 @@ int get_molecule(char *line, int connect_col_no, int q_col_no,
 
     /* Skip unwanted fields until charge column, then update charge */
 
-    for (i=1; i < q_col_no; i++) strtok(NULL," ");
-    cur_field = strtok(NULL, " ");
+    for (i=1; i < q_col_no; i++) strtok(NULL,WHITESPACE);
+    cur_field = strtok(NULL, WHITESPACE);
     atoms[*counter].q = atof(cur_field);
 
     /* Continue skipping unwanted fields until connectivity records begin */
 
-    for ( i = (q_col_no + 1); i < connect_col_no; i++) strtok(NULL," ");
+    for ( i = (q_col_no + 1); i < connect_col_no; i++) strtok(NULL,WHITESPACE);
 
     /* Process connections */
 
     connect_no = 0; /* reset connections counter */
-    while ((cur_field = strtok(NULL," ")) && (connect_no < MAX_CONNECTIONS)) {
+    while ((cur_field = strtok(NULL,WHITESPACE)) && (connect_no < MAX_CONNECTIONS)) {
       sscanf(cur_field, "%s", atoms[*counter].connections[connect_no++]);
     }
     atoms[*counter].no_connect = connect_no;

tools/msi2lmp/src/SearchAndFill.c

@@ -45,9 +45,9 @@ const char *SearchAndCheck(const char *keyword)
       exit(1);
     }
     if (line[0] == '@') {
-      if (string_match(strtok(line+1," '\t\n'("),keyword)) {
+      if (string_match(strtok(line+1," '\t\n\r\f("),keyword)) {
         got_it = 1;
-        status = strtok(NULL," '\t\n(");
+        status = strtok(NULL," '\t\n\r\f(");
         if (status != NULL)
           return strdup(status);
       }
@@ -79,7 +79,7 @@ void SearchAndFill(struct FrcFieldItem *item)
       exit(1);
     }
     if (line[0] == '#') {
-      if (string_match(strtok(line," '\t'("),item->keyword)) got_it = 1;
+      if (string_match(strtok(line," '\t\r\n("),item->keyword)) got_it = 1;
     }
     /*     if (strncmp(line, item->keyword,strlen(item->keyword))==0) got_it = 1; */
   }
@@ -132,13 +132,13 @@ void SearchAndFill(struct FrcFieldItem *item)
 
     /* version number and reference number */
 
-    version = atof(strtok(line, " "));
-    reference = atoi(strtok(NULL, " "));
+    version = atof(strtok(line, WHITESPACE));
+    reference = atoi(strtok(NULL, WHITESPACE));
 
     /* equivalences */
 
     for(i = 0; i < item->number_of_members; i++ ) {
-      charptr = strtok(NULL, " ");
+      charptr = strtok(NULL, WHITESPACE);
       if (strlen(charptr) > 4) {
         fprintf(stderr,"Warning: type name overflow for '%s'. "
                 "Truncating to 4 characters.\n",charptr);
@@ -150,7 +150,7 @@ void SearchAndFill(struct FrcFieldItem *item)
        endbontor have to be treated carefully */
 
     for( i = 0; i < item->number_of_parameters; i++ ) {
-      charptr = strtok(NULL, " ");
+      charptr = strtok(NULL, WHITESPACE);
       if(charptr == NULL) {
         for ( j = i; j < item->number_of_parameters; j++ )
           parameters[j] = parameters[j-i];

tools/msi2lmp/src/msi2lmp.c

@@ -2,6 +2,9 @@
 *
 *  msi2lmp.exe
 *
+*   v3.9.8 AK- Improved whitespace handling in parsing topology and force
+*              field files to avoid bogus warnings about type name truncation
+*
 *   v3.9.7 AK- Add check to enforce that Class1/OPLS-AA use A-B parameter
 *              conventions in force field file and Class2 us r-eps conventions
 *

tools/msi2lmp/src/msi2lmp.h

@@ -36,7 +36,7 @@
 
 # include <stdio.h>
 
-#define MSI2LMP_VERSION "v3.9.7 / 24 Oct 2015"
+#define MSI2LMP_VERSION "v3.9.8 / 06 Oct 2016"
 
 #define PI_180  0.01745329251994329576
 
@@ -45,6 +45,8 @@
 #define MAX_STRING        64
 #define MAX_NAME          16
 
+#define WHITESPACE " \t\r\n\f"
+
 #define MAX_ATOM_TYPES         100
 #define MAX_BOND_TYPES         200
 #define MAX_ANGLE_TYPES        300