allow comment line to be overrriden so that plain xyz files can be read

libextxyz/extxyz.c

@@ -568,8 +568,8 @@ char *read_line(char **line, unsigned long *line_len, FILE *fp) {
     return *line;
 }
 
-int extxyz_read_ll(cleri_grammar_t *kv_grammar, FILE *fp, int *nat, DictEntry **info, DictEntry **arrays) {
-    char *line;
+int extxyz_read_ll(cleri_grammar_t *kv_grammar, FILE *fp, int *nat, DictEntry **info, DictEntry **arrays, char *comment) {
+    char *line, *temp_line;
     unsigned long line_len;
     unsigned long line_len_init = 1024;
 
@@ -603,8 +603,15 @@ int extxyz_read_ll(cleri_grammar_t *kv_grammar, FILE *fp, int *nat, DictEntry **
         free(line);
         return 0;
     }
-    // actually parse
+    // actually parse - optionally replace line read from file with `comment` argument
+    if (comment != NULL) {
+      temp_line = line;
+      line = comment;
+    }
     cleri_parse_t * tree = cleri_parse(kv_grammar, line);
+    if (comment != NULL) {
+        line = temp_line;
+    }
     if (! tree->is_valid) {
         fprintf(stderr, "Failed to parse string at pos %zd\n", tree->pos);
         cleri_parse_free(tree);

libextxyz/extxyz.h

@@ -1,13 +1,14 @@
 /* interface for the the low-level C wrapper around libcleri parsed extxyz comment lines.
 
-   int extxyz_read_ll(kv_grammar, fp, nat, info, arrays)
+   int extxyz_read_ll(kv_grammar, fp, nat, info, arrays, comment)
    Parameters:
      cleri_grammar_t *kv_grammar: grammar, passed in so it does not have to be compiled in every time.
         calling routine should compile once and save indefinitely
      FILE *fp: file pointer to object to be read from, positioned at start of natoms line
      int *nat: storage for number of atoms
      DictEntry **info: pointer to allocated storage for info dict, will return pointer to first entry in linked list
      DictEntry **arrays: pointer to allocated storage for arrays dict, will return pointer to first entry in linked list
+     char *comment: NULL or pointer to replacement comment line. Useful if a previous call failed due to parse error.
    Returns
      int 0 for failure and 1 for success.
 
@@ -67,6 +68,6 @@ typedef struct dict_entry_struct {
 
 void print_dict(DictEntry *dict);
 void free_dict(DictEntry *dict);
-int extxyz_read_ll(cleri_grammar_t *kv_grammar, FILE *fp, int *nat, DictEntry **info, DictEntry **arrays);
+int extxyz_read_ll(cleri_grammar_t *kv_grammar, FILE *fp, int *nat, DictEntry **info, DictEntry **arrays, char *comment);
 int extxyz_write_ll(FILE *fp, int nat, DictEntry *info, DictEntry *arrays);
-void* extxyz_malloc(size_t nbytes);
+void* extxyz_malloc(size_t nbytes);

python/extxyz/cextxyz.py

@@ -49,7 +49,8 @@ class Dict_entry_struct(ctypes.Structure):
 extxyz.extxyz_read_ll.args = [ctypes.c_void_p, ctypes.c_void_p,
                               ctypes.POINTER(ctypes.c_int),
                               ctypes.POINTER(Dict_entry_ptr),
-                              ctypes.POINTER(Dict_entry_ptr)]
+                              ctypes.POINTER(Dict_entry_ptr),
+                              ctypes.c_void_p]
 
 extxyz.extxyz_write_ll.args = [ctypes.c_void_p, ctypes.c_int, Dict_entry_ptr, Dict_entry_ptr]
 
@@ -209,12 +210,13 @@ def cfclose(fp):
     fclose(fp)
 
 
-def read_frame_dicts(fp, verbose=False):
+def read_frame_dicts(fp, verbose=False, comment=None):
     """Read a single frame using extxyz_read_ll() C function
 
     Args:
         fp (FILE_ptr): open file pointer, as returned by `cfopen()`
         verbose (bool, optional): Dump C dictionaries to stdout. Defaults to False.
+        comment (str, optional): Overrride comment line with specified string.
 
     Returns:
         nat, info, arrays: int, dict, dict
@@ -225,11 +227,14 @@ def read_frame_dicts(fp, verbose=False):
     eof = False
 
     try:
+        if comment is not None:
+            comment = comment.encode('utf-8')
         if not extxyz.extxyz_read_ll(_kv_grammar,
-                                    fp,
-                                    ctypes.byref(nat),
-                                    ctypes.byref(info),
-                                    ctypes.byref(arrays)):
+                                     fp,
+                                     ctypes.byref(nat),
+                                     ctypes.byref(info),
+                                     ctypes.byref(arrays),
+                                     comment):
             eof = True
             raise EOFError()
 

python/extxyz/cli.py

@@ -19,6 +19,7 @@ def main():
     parser.add_argument('-R', '--round-trip', action='store_true')
     parser.add_argument('-P', '--profile', action='store_true')
     parser.add_argument('-C', '--cextxyz', action='store_true')
+    parser.add_argument('--comment', action='store', default=None)
     args = parser.parse_args()
     if args.round_trip:
         args.write = True # -R implies -w too
@@ -37,7 +38,8 @@ def main():
                     use_regex=args.regex,
                     create_calc=args.create_calc,
                     calc_prefix=args.calc_prefix,
-                    use_cextxyz=args.cextxyz)
+                    use_cextxyz=args.cextxyz,
+                    comment=args.comment)
     tr = time.time() - t0
     if args.verbose:
         print("main output of read()")

python/extxyz/extxyz.py

@@ -584,22 +584,24 @@ def read_frame_dicts(file, verbose=0, use_regex=True):
 
 
 def read_frame(file, verbose=0, use_cextxyz=True,
-               use_regex=True, create_calc=False, calc_prefix=''):
+               use_regex=True, create_calc=False, calc_prefix='', 
+               comment=None):
     """
     Read a single frame in extxyz format from `file`.
     """
 
     try:
         if use_cextxyz:
-            natoms, info, arrays = cextxyz.read_frame_dicts(file, verbose=verbose)        
+            natoms, info, arrays = cextxyz.read_frame_dicts(file, verbose=verbose, 
+                                                            comment=comment)        
             properties = info.pop('Properties', 'species:S:1:pos:R:3')
             properties = Properties(property_string=properties)
             data = np.zeros(natoms, properties.dtype_vector)
             for name, value in arrays.items():
                 data[name] = value            
         else:
             natoms, info, data, properties = read_frame_dicts(file, verbose=verbose, 
-                                                                use_regex=use_regex)
+                                                              use_regex=use_regex)
     except EOFError:
         return None
 

setup.py

@@ -126,7 +126,7 @@ def get_ext_filename(self, ext_name):
 
 setup(
     name='extxyz',
-    version='0.1.3',
+    version='0.1.4',
     author='various',
     packages=['extxyz'],
     package_dir={'': 'python'},