[Oberon] Re (2): Explaining Texts.

[Jörg]

Lyall E



Here an example:

- Texts.Text represents this string "Hi Lyall, this is a text.". The length "len" of this Text is 25, the character positions "pos" can be 0..24.

- character at position pos=0 is "H", character at position pos=8 is ","  character at position pos=13 is an "s".

- if this string was stored in contiguous memory (let's call it "mem"), you could easily find these characters with a simply array access like "ch := mem[pos];"

- But the implementation uses pieces instead of one big chunk of contiguous memory. So the easy array access mentioned above is not possible anymore.

- Let's assume the above Text string is broken down in 4 pieces:

  piece 1: "Hi Lyal",  pce.len=7   (piece holds character pos=0..6         piece's origin org=0)

  piece 2: "l, th",       pce.len=5   (piece holds character pos=7..11      piece's origin org=7)

  piece 3: "is is a t"   pce.len=9   (piece holds character pos=12..20    piece's origin org=12)

  piece 4: "ext."        pce.len=4   (piece holds character pos=21..24    piece's origin org=21)

- As you can see, if you sum up the len of all pieces you get the len of the text = 25. "org" is the position in the overall text of the first character in the piece. 



Instead of a simple memory access, you a) need a procedure to locate the right piece "pce" where "pos" is in, and b) locate the character in that piece, that's pos - org.

The first step a) is done by procedure "FindPiece(T, pos, org, pce)", it translates pos to pce/org.

The result of the last call to FIndPiece is stored (or you can call it cached) in "org" and "pce".



What about the following wording?

- pce: found piece of last search for character position "pos"

- org: character position of first character in "pce"



br

Jörg



Am 28.07.18, 00:43 schrieb "oberon-bounces at lists.inf.ethz.ch im Auftrag von peter at easthope.ca" <oberon-bounces at lists.inf.ethz.ch im Auftrag von peter at easthope.ca>:



    From:	=?UTF-8?B?SsO2cmc=?= <joerg.straube at iaeth.ch>

    Date:	Thu, 26 Jul 2018 08:53:12 +0200

    > BTW: the two fields "pce" and "org" are not "pointers in a translation 

    > cache", they are the cache. 

    

    Is this any better?  (Elliminate the preposition.  "translation cache" becomes 

    attributive to piece.)

    

    org: Offset in characters[0,len) of translation cache piece.

    

    Regards,                          ... Lyall E.

    -- 

    Message composed and transmitted by software designed to avoid the 

    need, overhead and vulnerability of antivirus software.

    

    123456789 123456789 123456789 123456789 123456789 123456789 123456789

    Tel: +1 360 639 0202                                                 

    http://easthope.ca/Peter.html              Bcc: peter at easthope. ca

    --

    Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems

    https://lists.inf.ethz.ch/mailman/listinfo/oberon