Fix Alembic upgrades.

A non-existent constraint was being referenced in the upgrade script.
This patch corrects the problem by removing the reference.

In addition, the head of the alembic branch referred to a non-existent
revision. This has been fixed by referring to the proper revision.

This patch fixes another realtime problem as well. Our Alembic scripts
store booleans as yes or no values. However, Sorcery tries to insert
"true" or "false" instead. This patch introduces a new boolean type that
translates to "yes" or "no" instead.

ASTERISK-26128 #close

Change-Id: I51574736a881189de695a824883a18d66a52dcef

include/asterisk/config_options.h

@@ -445,6 +445,29 @@ enum aco_option_type {
 	 * {endcode}
 	 */
 	OPT_UINT_T,
+
+	/*! \brief Type for default option handler for bools (ast_true/ast_false)
+	 * \note aco_option_register flags:
+	 *   non-zero : process via ast_true
+	 *   0        : process via ast_false
+	 * aco_option_register varargs:
+	 *   FLDSET macro with the field of type int. It is important to note that the field
+	 *   cannot be a bitfield. If bitfields are required, they must be set via a custom handler.
+	 *
+	 * This is exactly the same as OPT_BOOL_T. The only difference is that when
+	 * translated to a string, OPT_BOOL_T becomes "true" or "false"; OPT_YESNO_T becomes
+	 * "yes" or "no".
+	 *
+	 * Example:
+	 * {code}
+	 * struct test_item {
+	 *     int enabled;
+	 * };
+	 * aco_option_register(&cfg_info, "enabled", ACO_EXACT, my_types, "no", OPT_YESNO_T, 1, FLDSET(struct test_item, enabled));
+	 * {endcode}
+	 */
+	OPT_YESNO_T,
+
 };
 
 /*! \brief A callback function for handling a particular option