Source   Edit

Routines for converting between different character encodings. On UNIX, this uses the iconv library, on Windows the Windows API.

The following example shows how to change character encodings.


import std/encodings
when defined(windows):
    orig = "öäüß"
    # convert `orig` from "UTF-8" to "CP1252"
    cp1252 = convert(orig, "CP1252", "UTF-8")
    # convert `cp1252` from "CP1252" to "ibm850"
    ibm850 = convert(cp1252, "ibm850", "CP1252")
    current = getCurrentEncoding()
  assert orig == "\195\182\195\164\195\188\195\159"
  assert ibm850 == "\148\132\129\225"
  assert convert(ibm850, current, "ibm850") == orig
The example below uses a reuseable EncodingConverter object which is created by open with destEncoding and srcEncoding specified. You can use convert on this object multiple times.


import std/encodings
when defined(windows):
  var fromGB2312 = open("utf-8", "gb2312")
  let first = "\203\173\197\194\163\191\210\187" &
  assert fromGB2312.convert(first) == "谁怕?一蓑烟雨任平生"

  let second = "\211\208\176\215\205\183\200\231" &
  assert fromGB2312.convert(second) == "有白头如新,倾盖如故"


EncodingConverter = object
  dest, src: CodePage
  Source   Edit
EncodingError = object of ValueError
Exception that is raised for encoding errors.   Source   Edit


proc close(c: EncodingConverter) {....raises: [], tags: [].}
Frees the resources the converter c holds.   Source   Edit
proc codePageToName(c: CodePage): string {....raises: [], tags: [].}
  Source   Edit
proc convert(c: EncodingConverter; s: string): string {.
    ...raises: [EncodingError, OSError], tags: [].}
  Source   Edit
proc convert(s: string; destEncoding = "UTF-8"; srcEncoding = "CP1252"): string {.
    ...raises: [ValueError, EncodingError, EncodingError, OSError], tags: [].}
Converts s to destEncoding. It assumed that s is in srcEncoding. This opens a converter, uses it and closes it again and is thus more convenient but also likely less efficient than re-using a converter.
Warning: UTF-16BE and UTF-32 conversions are not supported on Windows.
  Source   Edit
proc getCurrentEncoding(uiApp = false): string {....raises: [], tags: [].}
Retrieves the current encoding. On Unix, "UTF-8" is always returned. The uiApp parameter is Windows specific. If true, the UI's code-page is returned, if false, the Console's code-page is returned.   Source   Edit
proc nameToCodePage(name: string): CodePage {....raises: [ValueError], tags: [].}
  Source   Edit
proc open(destEncoding = "UTF-8"; srcEncoding = "CP1252"): EncodingConverter {.
    ...raises: [ValueError, EncodingError], tags: [].}
Opens a converter that can convert from srcEncoding to destEncoding. Raises IOError if it cannot fulfill the request.   Source   Edit