26.2 Listing rules

The parse diagnostics should include a facility for displaying the rules in a grammar. The rules will have IDs from 0 to the highest rule ID, which can be found using the marpa_g_highest_rule_id() method (see marpa_g_highest_rule_id).

It is up to the application whether to have names for rules. In some grammar generators, the LHS of a rule serves as its name, but does not produce unique names for each rule in Libmarpa, since many rules may have the same LHS. Currently the most used upper layer for Libmarpa, Marpa::R2, does not have rule names. An application that does want to name its rules must decide on a rule naming convention.

At the highest verbosity level, the rule listing facility should display

• the rule ID;
• the LHS of the rule (see marpa_g_rule_lhs);
• the RHS of the rule, that is, each RHS symbol (see marpa_g_rule_rhs) for the length of the rule (see marpa_g_rule_length);
• whether the rule is accessible (see marpa_g_rule_is_accessible);
• whether the rule is nullable (see marpa_g_rule_is_nullable);
• whether the rule is nulling (see marpa_g_rule_is_nulling);
• whether the rule is productive (see marpa_g_rule_is_productive);
• the rank of the rule (see marpa_g_rule_rank);
• whether the rule is a sequence rule (see marpa_g_sequence_min);
• if the rule is a sequence rule, the minimum count of the sequence (see marpa_g_sequence_min); and
• if the rule is a sequence rule, the sequence separator, if any, (see marpa_g_sequence_separator)
• if the rule is a sequence rule, whether the rule uses proper separation, (see marpa_g_rule_is_proper_separation);