3 # A script checking the po-files for translators. Final goal is updating AUTHORS.
5 # To be copied and run in the git directory for having "ls po/*po" find the files.
6 # It will generate a subdirectory temp_translators
8 echo Searching the headers of *.po
10 mkdir temp_translators
11 # Should check for existance, for the real use should be running a diff next time for reducing the lines to read
13 git describe > temp_translators/start.txt
14 date >> temp_translators/start.txt
15 # to have the date visible
18 for i in $( ls po/*po ); do
20 echo $i >>temp_translators/start.txt
25 echo $i is just a link >>temp_translators/start.txt
29 # translationlangcode=$(echo $i | sed 's/\([a-z]*\).po*/\1/')
30 # longlanguage=$(sed -n '1,30 s/[*]*[t,T]ranslati[.]*/&/p' < $i)
31 # poeditlanguage=$(sed -n '1,30 s/[.]*X-Poedit-Language:[.]*/&/p' < $i)
32 # echo $longlanguage $poeditlanguage
33 # echo $translationlangcode >> ausgabe/start.txt
34 # I used this to have more output on running the script. Gives also the language name instead of the po-file's name only
37 sed '/#: include/ q' < $i >> temp_translators/start.txt
38 # The header should be done by now.
39 #TODO: A real safe way would take only the wanted meta parts (i.e. "Last-Translator:) and all commented lines #
46 # Removing every \n The result will have to be read by humans, so there it's no use in keeping them
47 sed 's:\\n::g' < temp_translators/start.txt > temp_translators/start_without_backslash_n.txt
49 echo '"Plural-Forms:.*' > temp_translators/doubles.txt
50 # TODO: This can be more than one line ^^
51 echo '"&& n%100<=10 ? 3 : n%100>=11 && n%100<=99 ? 4 : 5;"' >> temp_translators/doubles.txt
52 echo '"%100<10 || n%100>=20) ? 1 : 2);"' >> temp_translators/doubles.txt
53 echo '"|| n%100>=20) ? 1 : 2);"' >> temp_translators/doubles.txt
54 echo '"%100==4 ? 3 : 0);"' >> temp_translators/doubles.txt
55 # The only two liners so far. Quick fix for the above TODO...
57 echo '"PO-Revision-Date: .*"' >> temp_translators/doubles.txt
58 echo '"X-Poedit-Bookmark.*' >> temp_translators/doubles.txt
59 echo '"Project-Id-Version:.*' >> temp_translators/doubles.txt
60 echo '"X-Generator: .*' >> temp_translators/doubles.txt
61 echo '"Language: .*' >> temp_translators/doubles.txt
62 echo ' *[Cc]opyright *([cC])[0-9 ,-]*t*h*e* *VideoLAN$* *t*e*a*m*[0-9 ,-]*\.*' >> temp_translators/doubles.txt
63 echo '[ ]*\$[ ]*[i,I][d,D][ ]*[:]*[ ]*\$' >> temp_translators/doubles.txt
64 echo '"X-Poedit-Country:.*' >> temp_translators/doubles.txt
65 echo '"X-Project-Style:.*' >> temp_translators/doubles.txt
66 echo ' *<videolan@videolan.org> *' >> temp_translators/doubles.txt
67 # Whatever line occurs twice or more is most probably not a translators name. However, you can check it. The file will not be deleted
68 #TODO: If I knew sed better, I would put all the removed parts in a logfile. sort uniq would give a list to check for everything deleted
69 sort -r temp_translators/start_without_backslash_n.txt | uniq -d >> temp_translators/doubles.txt
70 # With the reverse Order the # will be removed at the end. Else the doubles beginning with # will not be matched
72 # Just check a "sort results.txt|less" for more to remove
74 # Changing the strings to sed commands removing the doubles
75 sed 's:.*:s^&^^g:' < temp_translators/doubles.txt > temp_translators/generated_com
78 # Removing all doubles
79 sed -f temp_translators/generated_com temp_translators/start_without_backslash_n.txt >temp_translators/results.txt
82 # Now, we are going to mark the already mentioned translators..
83 # Some names are written in CAPITALS. A modified marking script would be good.
85 sed -n '/Localization/,/^$/ p' <AUTHORS | sed -n '3,$ s/\(.*\) --* .*/\1/p' >temp_translators/localization_part.txt
86 #sed -n '/Localization/,/^$/ p' <AUTHORS | sed -n '3,$ s/\(.*\) - .*/\1/p' >temp_translators/sect.txt
87 #The second line is what we want. I added -* to have "Éric Lassauge -- French" included, too. But I really think this is a typo in AUTHORS
89 sed -n '/Programming/,/^$/ p' <AUTHORS | sed -n '3,$ p' >temp_translators/pro_part.txt
90 sed 's:.*:s^&^YYY-- & --YYY^g:' < temp_translators/pro_part.txt > temp_translators/replace_prog_names
93 # Changing the strings to sed commands removing the doubles
94 sed 's:.*:s^&^XXX-- & --XXX^g:' < temp_translators/localization_part.txt > temp_translators/replacenames
98 #mkdir -p temp_twice_used_adress
99 #echo "Checking all git logs"
100 #git shortlog -sne > temp_twice_used_adress/all_shortlog_sne.txt
103 #echo "Removing everything but email addresses"
104 git shortlog -sne |sed 's/[^<]*\(.*\)/\1/g' |sort > temp_translators/git_addresses_only.txt
106 sed 's:.*:s^&^ZZZ-- & --ZZZ^g:' < temp_translators/git_addresses_only.txt > temp_translators/replace_git_commiters
110 #sed -f temp_translators/replacenames <temp_translators/results.txt |uniq >temp_translators/review.txt
112 #sed -f temp_translators/replacenames <temp_translators/results.txt |sed -f temp_translators/replace_prog_names |uniq >temp_translators/review.txt
113 sed -f temp_translators/replacenames <temp_translators/results.txt |sed -f temp_translators/replace_prog_names |sed -f temp_translators/replace_git_commiters |uniq >temp_translators/review.txt
115 echo "Now temp_translators/review.txt should be reviewed. I don't think this can be done automatically, so I have done it already (not on your computer). Feedback is appreciated."
116 echo "XXX are named translators, YYY are named programmers, ZZZ commited with git"
118 # temp_translators/review.txt has to be reviewed manually. I don't think you would simply remove the names from AUTHORS found because then you would
119 # see an email adress and would not know, if it has to be added or how (with what name).
120 # The simpler to handle git commiter check script gives better names and dates, but of course not for all contributors. However, cross checking is recommended.
122 # The idea is to use a diff next time, but I did not work out yet what files would be best for that. Probably after finding the doubles would be a good time to remove everything visible in older versions.