It seems that msgcat may create an invalid header (at least "invalid" as in "cannot be parsed by poedit").
>>>>>>>>>>>>>>>>>>>>>> file1.po $ cat file1.po msgid "" msgstr "" "Project-Id-Version: PROJECT\n" "POT-Creation-Date: 2013-03-13 19:05+0100\n" "PO-Revision-Date: 2013-03-13 19:05+0100\n" "Content-Type: text/plain; charset=utf-8\n" #: /location:here msgid "str1" msgstr "none" <<<<<<<<<<<<<<<<<<<<<< A) concat file1.po with itself: $ msgcat file1.po file1.po # resulting header is ok, poedit is happy B) concat file1.po with a slightly modified version where only the Revision-Date has been modified: $ sed -i '/Revision-Date/s/19:05/19:06/' file1.po > file2.po >>>>>>>>>>>>>>>>>>>>>> $ msgcat file1.po file2.po #, fuzzy msgid "" msgstr "" "#-#-#-#-# file1.po (PROJECT) #-#-#-#-#\n" "Project-Id-Version: PROJECT\n" "POT-Creation-Date: 2013-03-13 19:05+0100\n" "PO-Revision-Date: 2013-03-13 19:05+0100\n" "Content-Type: text/plain; charset=utf-8\n" "#-#-#-#-# file2.po (PROJECT) #-#-#-#-#\n" "Project-Id-Version: PROJECT\n" "POT-Creation-Date: 2013-03-13 19:05+0100\n" "PO-Revision-Date: 2013-03-13 19:06+0100\n" "Content-Type: text/plain; charset=utf-8\n" #: /location:here msgid "str1" msgstr "none" <<<<<<<<<<<<<<<<<<<<<< 1) notice the "fuzzy" on the first line 2) try to open such a file using poedit, it will warn: > Malformed header: '#-#-#-#-# file1.po (PROJECT) #-#-#-#-#' Is poedit expected to parse such a heading line or is msgcat wrong in generating such a header ? (Using poedit 1.4.6.1 and gettext 0.18.1.1) regards
