TrimAll Roman / non-Roman ?

[LaRetta]

Hi Everyone!

I use TrimAll( field; 1; 1 ) because ... well ... because I saw -Queue- use it (and he's XP and American and just knows mostly everything) so I figure it's correct for me.

But I'm not satisfied with basing my understanding on ditto-behavior. I've studied everything I can find on this issue (and re-read Help a hundred times) and my confusion remains. You ever do a Google search for Unicode? It's a lateral universe. Code charts don't even list Verdana. And the Index doesn't even list FONTS.

From FM Help, I deduced that I am non-Roman (simply because my fonts didn't appear in the list qualified as Roman). Help says "Any character whose unicode value is greater than or equal to U+2F00 is considered non-Roman. " I can't even find a list of Unicode numbers that look like those numbers. FM says non-Roman is CJK. CJK pdf gives 8 pages of characters and numbers - none of which make any sense to me.

Is there a list of all fonts and their unicode values? As I use different fonts, I don't want to be surprised and find out it's Roman. So, basing my understanding on the fuzzy theory that I'm non-Roman, I continue reading ...

"Set trimSpaces to True (1) if you want to include the removal of full-width spaces between non-Roman and Roman characters. Set trimSpaces to False (0) if you do not."

But ... but ... what else would it do? If 0, does that mean it removes half-width spaces only? If it doesn't remove half-width then it doesn't remove any spaces and what is the purpose of using TrimAll()? And it refers to 'characters' and 'text' instead of words and so (in the first table, based upon me being non-Roman), 0 and 3 are identical!?

And then the topper, just when I think I'm becoming clear (NOT), as I chart the theory out on paper ...

"In all cases, spaces between non-Roman characters are removed."

Oh. Dear me. Half-spaces too? Between characters? Text? Words? Even if trimspaces is 0? I think it would have helped if FM would have used Let[] in their description of TrimAll(). Comment! Contact them! Maybe it would have made more sense to me.

Can someone help me get clarity on this or direct me to some reading about fonts and unicode? I just want to make sure I thoroughly understand what-the-heck this all means. Because after all my attempts (and study) and resulting cross-eyed blindness, I still use TrimAll(field; 1; 1) simply because I saw -Queue- use it.

LaRetta

Abstract-Characterless in Oregon

[comment]

A short practical answer:

As long as you're using the Latin alphabet, you can use TrimAll ( text ; 0 ; 0 ) to leave a single space between words, or TrimAll ( text ; 0 ; 3 ) to remove ALL spaces.

To answer some of your musings:

Unicode is a code, in the same category as ASCII. It PRESCRIBES which character should appear when a code is called. A font is something else. A font is (or can be) an IMPLEMENTATION of a code. A font complies with a code, if it has the prescribed shapes at each code slot.

Unicode, compared to ASCII, has a very large number of slots. The slots are divided into blocks. The Roman block includes practically everything except Chinese, Japanese and Korean (CJK). So unless you're doing a menu for the restaurant down the block, consider yourself a member of the Roman empire.

[LaRetta]

Oh what a relief. I feel like you've lifted 20 pounds from my brain.

"As long as you're using the Latin alphabet..."

Strange but true ... in 20 years working with computers, I've never known I was Latin alphabet. And FileMaker makes no reference to it. It DOES refer to Roman language (File Options/Text) but it only says I'm Western fonts (Edit/Preferences/Fonts) and the word 'Latin' doesn't even exist in Help!

"You can use TrimAll ( text ; 0 ; 0 ) to leave a single space between words, or TrimAll ( text ; 0 ; 3 ) to remove ALL spaces."

Questions please? What if there are spaces between characters? FM should say words not characters - what am I to think? I'd like to talk to their proof reader. And what if I import text which has half-spaces? How would I know? Extend & Condense don't apply to spaces?

"... consider yourself a member of the Roman empire."

So I'm a Latin Roman Westerner who should only care about (and focus on) full-spaces between words and not be concerned with the rest of that stuff?

Looks like I've spent 4 months researching unicode (off and on) for nothing. Well ... that's not true ... I DID learn that I don't want to work with unicode. I'm glad I work with FileMaker instead.

Thank you, Michael!!

LaRetta

[comment]

in 20 years working with computers, I've never known I was Latin alphabet.

It has nothing to do with computers. This is the Latin alphabet*, and has been so for the last 27 centuries (approximately).

What if there are spaces between characters?

In Filemaker, a space is considered a word deliminator, so spaces are ALWAYS between words - if they are between anything. Most of the help text about TrimALL() deals with mixing Roman with CJK. There, it is important which CHARACTERS are at each side of the space. At least in theory, a word could begin with a Chinese character and continue with Roman ones.

what if I import text which has half-spaces?

I don't know that there is a half-space character in the Roman block. Extend and Condense are STYLES - they are applied to the existing characters, not changing them in any way - only the way they are rendered.

----

(*) A musing of my own: Latin and Roman were considered synonyms in this context, until the fathers of Unicode decided that the Roman block includes such decidedly non-Latin alphabets as Greek/Cyrillic, Hebrew, Arabic, and so on.

[LaRetta]

Oh. I was going to zip my lip but ...

"and has been so for the last 27 centuries (approximately)."

Oh how time flies when you're having fun!

L

[LaRetta]

I need to apologize to -Queue-

I said his example showed TrimAll(text; 1;1) but ...

I looked at my Principles table and, errrr, an example he gave using TrimAll() used 0;0 NOT 1;1. I am unsure why I planted the 1;1 in my mind - I guess that's why I keep these tid-bits in a table for constant review.

-Queue- had it right as I would have expected.

It's a good thing I don't base my understanding on parroting behavior (and prefer understanding the principles) ... I mis-parroted.

Sorry for mis-quoting you, JT!

[-Queue-]

No problem. I think I've used all of the above in samples at one time or another.