Scott Moore wrote:
Frank Heckenbach wrote:
My understanding of the ISO pages is that the characters outside of ASCII are in the > 127 codes. So, for example, IP Pascal specifically leaves the 8th bit unmolested, so would I/O other ISO code pages ok, and would accept ISO pages as source, since it treats c < 32 or c > 127 as characters to be ignored.
That's a valid decision, according to ISO Pascal, but not the one we've made, or which I personally like. Ignoring control characters is not exactly my idea, and characters > 127 (usually interpreted in ISO-8859-n) have been in use for a long time ...
Probally that needs clarification. When reading the source, IP "ignores" c < 32 and c > 127, as in does not try to parse or otherwise deal with them, but allows them to appear in the source without an error. That means that c < 32 and c > 127 characters can appear in the listing without fault.
This means that ISO code page characters can be used in comments. Since Pascal does not require any characters for the language outside 32 <= c <= 127, this gives ISO code page capability without relying on any particular code page format.
Ah, you're talking about source code. The problem is not so serious here (except for string literals). Data compatibility is the main concern, AFAICS.
Frank