class conversion_pipe :?subst:int -> string -> in_enc:encoding -> out_enc:encoding -> unit ->
Netchannels.io_obj_channel
This pipeline class (see Netchannels
for more information) can be used
to recode a netchannel while reading or writing. The argument in_enc
is the input encoding, and out_enc
is the output encoding.
The channel must consist of a whole number of characters. If it
ends with an incomplete multi-byte character, however, this is
detected, and the exception Malformed_code
will be raised.
This exception is also raised for other encoding errors in the
channel data.
Example. Convert ISO-8859-1 to UTF-8 while writing to the file
"output.txt"
:
let ch = new output_channel (open_out "output.txt") in
let encoder =
new conversion_pipe ~in_enc:`Enc_iso88591 ~out_enc:`Enc_utf8 () in
let ch' = new output_filter encoder ch in
... (* write to ch' *)
ch' # close_out();
ch # close_out(); (* you must close both channels! *)
If you write as UTF-16, don't forget to output the byte order mark yourself, as the channel does not do this.
Example. Convert UTF-16 to UTF-8 while reading from the file
"input.txt"
:
let ch = new input_channel (open_in "input.txt") in
let encoder =
new conversion_pipe ~in_enc:`Enc_utf16 ~out_enc:`Enc_utf8 () in
let ch' = new input_filter ch encoder in
... (* read from ch' *)
ch' # close_in();
ch # close_in(); (* you must close both channels! *)
subst
: This function is invoked for code points of in_enc
that
cannot be represented in out_enc
, and the result of the function
invocation is substituted (directly, without any further conversion).
Restriction: The string returned by subst
must not be longer than 50
bytes.
If subst
is missing, Cannot_represent
is raised in this case.