Native File Translate {R}←{X}⎕NXLATE Y

This associates a character translation vector with a native file or, if Y is 0, with the use by ⎕DR.

A translate vector is a 256-element vector of integers from 0-255. Each element maps the corresponding ⎕AV position onto an ANSI character code.

For example, to map ⎕AV[17+⎕IO] onto ANSI 'a' (code 97), element 17 of the translate vector is set to 97.

⎕NXLATE is a non-Unicode (Classic Edition) feature and is retained in the Unicode Edition only for compatibility.

Y is either a negative integer tie number associated with a tied native file or 0. If Y is negative, monadic ⎕NXLATE returns the current translation vector associated with the corresponding native file. If specified, the left argument X is a 256-element vector of integers that specifies a new translate vector. In this case, the old translate vector is returned as a shy result. If Y is 0, it refers to the translate vector used by ⎕DR to convert to and from character data.

The system treats a translate vector with value (⍳256)-⎕IO as meaning no translation and thus provides raw input/output bypassing the whole translation process.

The default translation vector established at ⎕NTIE or ⎕NCREATE time is derived from the mapping defined in the current output translation table (normally WIN.DOT) and maps alphabetic, numeric and most other characters in ⎕AV to their corresponding ANSI positions. However, some characters are not resolved by this process and it is recommended that users define translate vectors to cover all cases.

Unicode Edition

⎕NXLATE is relevant in the Unicode Edition only to process Native Files that contain characters expressed as indices into ⎕AV, such as files written by the Classic Edition.

In the Unicode Edition, when reading data from a Native File using conversion code 82, incoming bytes are translated first to ⎕AV indices using the translation table specified by ⎕NXLATE, and then to type 80, 160 or 320 using ⎕AVU. When writing data to a Native File using conversion code 82, characters are converted using these two translation tables in reverse.