build url web for a file name with accented letter

[jmburnod]

Hi All
I look for build a valid url for a file name with accented letters
On a server there is a file called "boîte.png", I noticed the http name is "bo%C3%AEte.png" and it works with this name

I thought I could use urlencode to get the same but
urlEncode(boîte) = "bo%94te"

How can I translate "boîte.png" to "bo%C3%AEte.png" ?
Thanks again for your lights
Jean-Marc

[zaxos]

Replace î with %C3% or the reverse? I`m guessing this char is allways represented like that.

Code: Select all

Put "http:\\blablabla\blabla\boîte.jpg" into tUrl
Replace "î" with "%C3%" in tUrl

Just a thought

[jmburnod]

Hi Zaxos
Thanks for reply

Replace î with %C3% or the reverse?

Unfortunately not
With your script I get "bo%C3%te.jpg" instead "bo%C3%AEte.jpg"
Best regards
Jean-Marc

[SparkOut]

then just
replace "î" with "%C3%AE" in tUrl

but it might be worthwhile trying to find our why the filename has been encoded like that. It looks like a problem interpreting between UTF-8 and ISO-8859-1 or Windows-1252 code sets. If you can find out the encoding used, you might be able to make it match from LiveCode by using the matching code set, rather than hard coding a replacement. But that might get pretty complicated.

[jacque]

If you are using LC 7.x you can use the built-in unicode support to translate:

Code: Select all

get "boîte" -- or any other text
put urlEncode(textEncode(it,"utf8"))

In LC 7, all text exported or imported from/to LiveCode should be textEncoded or textDecoded so it is compatible with external apps and servers that have unicode support. Text encoding will not alter non-unicode characters, so it is safe to use on all text.

If you are using LC 6.x then you'll need to use the more awkward uniEncode/uniDecode functions which are now deprecated.

[SparkOut]

Ah great, now I can test I see that is the unicode representation in UTF-8. I was unsure if the file name had been encoded with a different (eg Windows-1252) encoding. This works nicely though. Here's your solution Jean-Marc.

[jmburnod]

Thanks at all,
It works fine with LC 7.
Jaque said there is a way for LC 6.7 but I'm not yet able to use it
Is there an example of "uniEncode/uniDecode" way for 6.7 ?
Jean-Marc

[jmburnod]

I found a quick but dirty way to translate an accented file name.
We can create a list of translated accented chars in LC 7 using unicode.
This list can be used in LC 6.7 with a search/replace

[jacque]

You can use the older method this way:

Code: Select all

get "boîte"
  put urlEncode(unidecode(uniencode(it), "UTF8")) into tURL

[jmburnod]

Thanks a lot Jacque
That is the way for LC 6.7

[jmburnod]

Hi All,
It seems something changed about this.
url file name on server for "zéro_2.png" = "ze%CC%81ro_2.png"

I tried this

Code: Select all

  get "zéro_2.png"
  put urlEncode(unidecode(uniencode(it), "UTF8")) 

return "z%C3%A9ro_2.png" instead "ze%CC%81ro_2.png"
Thanks in advance for your help
Jean-Marc

[jacque]

Maybe you have to go back to textEncode?

[jmburnod]

Thanks again Jacqueline
We have tested on a ownCloud and it works with this:

Code: Select all

   get "zéro_2.png"
   put textencode(it,"UTF-8") into tTE
   put urlEncode(tTE) into tUrltTE
   replace "+" with "%20" in tUrltTE

No idea why our site does different encoding

P.S: my hight score on Klondike is 840. What is the max score ?

[jacque]

my hight score on Klondike is 840. What is the max score ?

LOL. I can't remember but I think you're close.

[richmond62]

La Boîte Diabolique:

https://youtu.be/mE0takm9TX0