No, you cannot add “ignore” stuff. ID can easily search for things but not for not-something.

To not find a space in the above, try this one:

(?s)S(?!.)

The GREP code S is the inverse of “every kind of whitespace”. This will locate the black-followed-by-nots, try this for the reverse order:

(?s)(?<!.)S

These codes also find the end of footnotes, table cells, and stories, not because “the formatting is different” but because that also qualifies as “text, followed by nothing”. I think it stopping right before a footnote could qualify as a Real bug — why doesn't it “eat” the footnote marker as a single character and continue!? (Someone forgot to treat the code as “special”, I guess.)

Theoretically, you can put all both queries above in one single long GREP, like this:

(?s)(S(?!.))|((?<!.)S))

I hope I got the right number of opening and closing parens! If I got it wrong, it'll just tell you “Not Found”.

A tricky one indeed.

I could search for anything black that touches something not-black and change the black thing to the style of the not0black thing, that would also work.

A good idea. Try this: put [Black] in the Find formatting field, and look for

(?s).(?!.)

This will look for Black characters where there is “nothing” following. It should work, because it seems ID treats “not the same formatting” as “nothing”.

It needs the Ignore Hard Return flag (?s) because the end of a paragraph is also considered “nothing” in default mode.

The presence or absence of characters in a font have nothing to do with Opentype (well … hardly), and, the suffix “Pro” does not indicate that as well. You'd have to ask Adobe, but I think the “Pro” gets added to a pre-existing font that is converted to Opentype format, and in which several previously distinct font files are merged. Pre-Opentype fonts had to have separate files for extra accented characters and ligatures, small capitals, old style digits — stuff like that. (That's just an observation. Since this is just Adobe's own private font naming convention, other font foundries are not obliged or forbidden to add or leave off the “Pro”.)

So there is your answer. Cyrillic is in Times “because the designers added it”, and the reverse is true for Adobe Garamond Pro. (The actual language you assigned to your text has nothing to do with this …)

Garamond Premier Pro, on the other hand, has Cyrillic characters. Perhaps you could switch fonts for your entire publication, or (mixing different Garamonds … a bit like blasphemy …) use Premier Pro for only the Cyrillic text …

Something similar is going on with Stylistic Sets. There are no “rules” to what these ought to do for every font. In one font, for example, it might just change the appearance of the digit '0' to a slashed variant. In another, the lowercase 'a' and/or 'g' may switch from one story to a two story version; or something else may or may not happen, to only one character or to lots of them. For these variants you may want to read the documentation that came with each font.

No such feature in this 21st century software …

You could try “Go Back” in the Page menu, it sort of positions the screen — but not the cursor! — to the previous location.

I would like very much to learn more about prepress, can you please suggest me some books I could use?

A good background knowledge of prepress is always useful, and I would gladly recommend something to read — except that I know of no book or online document that you could start with! I learned my prepress “the hard way”, when computer-to-film was still in its infancy. (I fondly remember manually tinkering with PostScript files that failed to RIP properly to get 'em to work. I could probably write a book about that.)

Maybe one of the other forum members has a good suggestion for you.

First thing if you are going to compare files is to export to IDML, open again, save as INDD and then compare their size. (But make sure you don't accindentally overwrite your originals! Sure, this ought to be quite safe, but You Never Know.)

The reason for this fairly crucial step is that InDesign does not “overwrite” existing data when you edit a file; instead, changes get appended to the end, which is much faster and also has the benefit that if ID crashes on you, all previous editing is still there. And hard disk space is cheap. So it's possible your Frech file “contains” all English text as well, and the Arabic file contais both! And everything else you changed, by the way, such as images.

Arabic characters, by the way, fall outside the regular 1-byte range for basic Latin-plus-some-accents — each aingle character will always use two bytes, so even if you have the exact same length of text, it'll take up twice the space. But I don't think that should account for the major difference you are seeing … No matter how much text you have, the number of bytes it uses is usually dwarfed by a single image. The saying goes, “a picture is worth a thousand words”, well, a thousand words is about 5, maybe 6 or 7 KB. A typical press quality image should be 100 times that size :)

What is the biggest acceptable file size for a printer?

You are talking about InDesign file sizes here, and as per rules outlined above, not everything that is “in” that file will end up in the final PDF. The inverse is also true: if you use linked graphic images, the final PDF could be much larger. You are sending PDFs to your printer, right?

As for “biggest acceptable” … it depends on your printer. Smaller files can be send by mail, but we've had a Guide to Amsterdam, big book, heavy on full-page hq photo's, which ended up as a multi-gigabyte PDF that had to be sent to Hungary for printing. Solution: ask for FTP details, set up the transfer at the end of the day, and the following morning the file had gone off to the presses.

Doesn't sound too hard (except for the fact that you really should not judge color by looking at your monitor — except when you have it calibrated, of course).

Can you mock up an image of what the result should look like? I don't think there is much room for all the variations you list, esp. in a mere 2% increase … Or the squares could be really small, perhaps :)