Regression in gettext-0.13.1: produces corrupt pot-file

[Erik Sigra]

I upgraded to gettext-0.13.1 but it failed. The reason is that it does not 
recognize the comment "/* xgettext:no-c-format */" in one place in the 
sourcecode. Therefore xgettext refuses to create the gmo-file, because it 
believes that there is something wrong with the format arguments. The 
sourcecode is:

  GEN_INT("diplchance", game.diplchance, SSET_RULES_FLEXIBLE, SSET_TO_CLIENT,
          N_("Chance in diplomat/spy contests"),
          /* xgettext:no-c-format */
          N_("A Diplomat or Spy acting against a city which has one or "
             "more defending Diplomats or Spies has a diplchance "
             "(percent) chance to defeat each such defender.  Also, the "
             "chance of a Spy returning from a successful mission is "
             "diplchance percent.  (Diplomats never return.)  Also, a "
             "basic chance of success for Diplomats and Spies. "
             "Defending Spys are generally twice as capable as "
             "Diplomats, veteran units 50% more capable than "
             "non-veteran ones."), NULL, 
          GAME_MIN_DIPLCHANCE, GAME_MAX_DIPLCHANCE, GAME_DEFAULT_DIPLCHANCE)

Gettext 0.11.5 produces a correct pot-file containing:
#: server/stdinhand.c:644
#, no-c-format
msgid ""
"A Diplomat or Spy acting against a city which has one or more defending "
"Diplomats or Spies has a diplchance (percent) chance to defeat each such "
"defender.  Also, the chance of a Spy returning from a successful mission is "
"diplchance percent.  (Diplomats never return.)  Also, a basic chance of "
"success for Diplomats and Spies. Defending Spys are generally twice as "
"capable as Diplomats, veteran units 50% more capable than non-veteran ones."
msgstr ""


But gettext 0.13.1 produces a corrupt pot-file containing:
#: server/stdinhand.c:644
#, c-format
msgid ""
"A Diplomat or Spy acting against a city which has one or more defending "
"Diplomats or Spies has a diplchance (percent) chance to defeat each such "
"defender.  Also, the chance of a Spy returning from a successful mission is "
"diplchance percent.  (Diplomats never return.)  Also, a basic chance of "
"success for Diplomats and Spies. Defending Spys are generally twice as "
"capable as Diplomats, veteran units 50% more capable than non-veteran ones."
msgstr ""