Mirror of forum.nim-lang.org

790 :: How to echo unicode string

• DarkBlue (orginal) [2015-01-25T06:26:54+01:00] view original

  System: Linux

  In python2.7 I do this:

  Run

  and get: 彼は銀行で働いています。

  in Nim 0.10.3: echo ?? how do

  Thanks

---

• Varriount (orginal) [2015-01-25T07:19:49+01:00] view original

  Um, what's wrong with

  
  echo("彼は銀行で働いています。")
  Run

  ?

---

• DarkBlue (orginal) [2015-01-25T08:03:32+01:00] view original

  Thank you for looking at it.

  There is nothing wrong with what you did.

  Try this:

  echo ("u'\u5f7c\u306f\u9280\u884c\u3067\u50cd\u3044\u3066\u3044\u307e\u3059\u3002'")
  Run

  or any variation of it.

  I am getting this string as result from a database cursor and try

  to actually display the contents rather than codes.

---

• nodrygo (orginal) [2015-01-25T13:10:16+01:00] view original

  Not sure,I may mistake, but I am afraid that u".." doesn't exist in Nim

  you can transform an unicode char with something like

  
  import unicode
  Rune(0x5f7c).toUTF8
  Run

  the Nim unicode module is far to be complete for the moment (if I compare to say: Elixir unicode module). Probably Nim have yet to evolve in this area. Even runeAt(...) work at area index level and that's bad, but Nim is a new language and have still an important need for contributor

---

• gradha (orginal) [2015-01-25T15:38:01+01:00] view original

  DarkBlue, that's not a unicode string, that's an ASCII string containing some form of unicode escaping sequence. The u prefix suggests you have some python code escaping unicode characters writing to the database and later decoding them when reading, quite a performance problem. Still, you can use libraries like iconv to decode the unicode escape sequences. The following example works under unix:

  
  import encodings
  
  proc test() =
    let input = r"u'\u5f7c\u306f\u9280\u884c\u3067\u50cd\u3044\u3066\u3044\u307e\u3059\u3002'"
    echo input.convert("UTF-8", "C99")
    # --> u'彼は銀行で働いています。'
  
  when isMainModule: test()
  Run

  The example might fail under Windows if it doesn't support the C99 encoding. In any case, the output still contains the u'…' markers. If you can't get rid of this encoding in the database you will have to write your own escaping of python unicode strings. Or maybe use something like nimborg to do the string decoding.

---

• LeuGim (orginal) [2015-01-25T19:11:55+01:00] view original

  Iа you need to parse such escaped string into UTF-8 encoded:

  import unicode, strutils
  proc escStrToUTF8(s: string): string =
    result=newStringOfCap(s.len/%6*4)
    for i in 0..(s.len/%6-1):
      let c=parseHexInt(s[(i*6+2)..(i*6+5)]).Rune.toUTF8
      result.add c
  echo escStrToUTF8(r"\u5f7c\u306f\u9280\u884c\u3067\u50cd\u3044\u3066\u3044\u307e\u3059\u3002")
  Run

---

• DarkBlue (orginal) [2015-01-26T03:10:17+01:00] view original

  Thank you for your replies, which helped me to understand what's needed.

  I am already using nimborg to get the data from a Firebird 2.5.3

  database, however, the unpacking of these escaped strings in Nim

  had me stumped .

  Doing everything on the python side obviously works.

---