Hello William Blum,

I recently found out your plug-in NppColumnSort.dll ( version 1.0.0.2 ) for N++. I really like it because :

  • the sort process is rapid ( On an old Windows XP notebook, with only 1MB of RAM, 5000 lines, with length till 130 characters, were sorted in about 30s )

  • the plug-in is highly configurable ( alphabetic/numeric, ascending/descending, sensitive/insensitive, choice of number, start and length of each sort column )

  • Sort process concerns all the lines of the current file, by default, or only the lines of a normal user selection

However, unless I misunderstood something, I think there are some odd sort results, when accentuated characters are involved !

For example, consider the original list, of 15 lines, below :

Image
Albert
Étienne Jean-Louis
Šerbedžija Rade
Etienne Louis Malus
Álvarez Nieves
Åkerlund Jonas
Šprem Karolina
Ímaz Manuela
Özpetek Ferzan
Oscar
Ösel Lama
Serbe
Çakir Deniz
Óscar Andrés Rodríguez Maradiaga

( Although I'm French, I very often use this list, with Turkish and/or Eastern Europe names, to test sort softwares. Don't ask me why ! )

All sort tests, below, are case sensitive


1)

If you performs a sort on 1 column of 35 positions, beginning at position 1, you obtain :

Albert
Etienne Louis Malus
Image
Oscar
Serbe
Álvarez Nieves
Åkerlund Jonas
Çakir Deniz
Étienne Jean-Louis
Ímaz Manuela
Óscar Andrés Rodríguez Maradiaga
Ösel Lama
Özpetek Ferzan
Šerbedžija Rade
Šprem Karolina

=> The sort is correct for standard and accentuated characters but you notice that words, beginning with an accentuated character, are placed AFTER the words with standard letters

Would it be possible to get on option to mix standard and accentuated characters ?

So, the list would be :

Albert
Álvarez Nieves
Åkerlund Jonas
Çakir Deniz
Etienne Louis Malus
Étienne Jean-Louis
Image
Ímaz Manuela
Oscar
Óscar Andrés Rodríguez Maradiaga
Ösel Lama
Özpetek Ferzan
Serbe
Šerbedžija Rade
Šprem Karolina

It looks nice also , isn't it ?


2)

If you performs a sort on 1 column of 1 position, beginning at position 1, you obtain :

Albert
Etienne Louis Malus
Image
Oscar
Serbe
Étienne Jean-Louis
Álvarez Nieves
Åkerlund Jonas
Ímaz Manuela
Özpetek Ferzan
Ösel Lama
Çakir Deniz
Óscar Andrés Rodríguez Maradiaga
Šerbedžija Rade
Šprem Karolina

=> Why the sort results are quite different from sort results, in point 1) ? I can't guess the logic of current sort !


3)

If you performs a sort on 1 column of 35 position, beginning at position 2, you obtain :

Serbe
Albert
Image
Oscar
Etienne Louis Malus
Álvarez Nieves
Åkerlund Jonas
Çakir Deniz
Étienne Jean-Louis
Ímaz Manuela
Óscar Andrés Rodríguez Maradiaga
Ösel Lama
Özpetek Ferzan
Šerbedžija Rade
Šprem Karolina

=> Sort seems correct for the first five words, with standard letters ( second letter = e, l, m, s, t )

But, with words beginning with an accentuated character, I can't see any logic in the actual sort !?


4)

Is there a possibility to consider sort on a column mode selection ONLY ( rectangular selection ). I fully understand that this feature should be much more complicated to program and I don't want you feel under pressure because of my suggestion !

Particularly, in that case, it would rather to allow one criteria of sort ONLY !

In advance, many thanks for your answer.

Best regards,

guy038