Everyone! How are you getting on?
Can I borrow your wisdom?
I am currently compiling information on musical literature.
I am researching something called “harmonia”,
This naturally applies to Plato’s works.
I am having trouble getting Greek words printed correctly.
I search for a font that seems to be applicable on PopChar.app etc., but it still doesn’t work.
It is not displaying as expected.
What kind of font do you use to display ancient Greek letters?
Garbled characters occur when copying and pasting the corresponding PDF document using OCR to make it readable.
If there is any way to use it, I would appreciate it if you could let me know.
The documents I am currently referencing are as follows:
Epic poem by Homer
Plato’s Treatise on the State, Volume 3
Augustine’s Theory of Music, Volume 1
Yours, WAKAMATSU
P.S.
The font I always use is Times New Roman.
macOS’s Times New Roman is certainly able to supply (ancient) greek characters and diacriticals. You can check that by opening macOS System Settings ▸ Keyboard ▸ Text Input and clicking the Edit… button. Then in the panel that opens click the ‘+’, lower left.
1.pick ‘Greek’
2. in the filtered content top right click ‘Greek - Polytonic’. Note the bottom right panel shows the characters/accents added by which key when this input method is selected in macOS settings
3. Add.
Now, by setting Greek - Polytonic as you Mac input method you should be able to type greek characters (in Times New Roman), if only to prove the font is likely not the issue.
So what might be the cause? A very common misconception is that PDFs store the text as you see it on screen. Sadly no. You are looking at individual PostScript shapes (representing characters) drawn onto an imagined paper sheet. Thus is shown PDFs true paper-print heritage. When you drag select ‘screen’ text, the PDF viewer/editor you are using has to go read those bits of Postscript and turn them back into the text it can then copy to the OS clipboard. But, what it detects might be incorrect and garbled so when you paste the receiving app & font get asked for a character they cannot draw , an incorrect character or data they can’ interpret.
So, if possible don’t use PDF alone for this work. An actual textual source might be more use, using the PDF just to show the original layout/styler of the source greek.
Dear Dr. Mark Anderson,
Thanks a lot.
I have already added the input settings for Greek on Mac that you mentioned.
I am currently practicing how to type on the keyboard. Especially with modifier keys, and learning how to add organic sounds and anaphoric sounds in combination.
As you said, it is difficult to obtain font information etc. from the PDF information, so
I am looking for ways to use fonts based on the resources of pages displayed in browsers such as Safari. Is it written like LaTeX? This is about whether the description is written for HTML.
Now, I have a question!
How did your advice “original layout/styler of the source greek.”
Should I enable it?
Yours, WAKAMATSU
Dear David Bertenshaw,
Thank you for your advice.
I got Gentium-7000.zip.
I often use Geogia with my favorite type of font, but from now on
I would like to display it on Gentium.
Yours, WAKAMATSU
My apologies, I expressed this poorly. I was meaning that, if possible, find a text-based source for the greek text so the characters are correct. However, the text may give little indication of how the original text looked. Separately, a PDF that might be a scan of an original papyrus, will show you how the text looked in the original source. But, as discussed the PDF is most likely not able to give you the correct character data for the greek text you wish to use elsewhere. This is why I’m suggesting two separate source formats might not be mere duplication.
If you are looking at a webpage in Safari—and it is not using a built-in viewer to show a PDF, you will be looking at HTML. These days, most (almost all) pages use UTF-8 encoding to the font used by the web page ought (like macOS Times Roman) to be able to render a very wider range of characters from multiple languages/scrips, including their diacriticals†. The easiest way to get only the text from a web page is to drag-select on the webpage and copy. this is because in the underlying HTML, the HTML codes are inter-mixed with the actual text. But, the clipboard will hold a styled text version of the webpage text. Doing a paste-and-match-style (⌘+⌥+⇧+V) into Tinderbox‡ will paste just the ‘text’ without any other styling (bolding, unrelenting, links, font variations, etc.).
The latter should enable you to paste the ancient greek into a Tinderbox note. However, you might want to first check that the note’s $TextFont (i.e. the default font used in the $Text area of the text pane) can show the ancient greek characters you reuire. So, test this before doing any such pasting as above.
I’m glad @brookter recommended the Gentium font. I’d seen it suggested but having no personal experience of that thought it best not to give an untested suggestion.
†. An exception, though not pertinent here, is that some Asian and East Asian scripts require specialist fonts to contain their larger ‘alphabet’ of characters and which do not encode easily into UTF-8.
‡. Or any other app with this feature, e.g. TextEdit but not MS Word. Or paste into a plain-text text editor like BBEdit that strips the style info as it only renders plain text.
Dear Dr.Mark Anderson,
Thank you for always providing detailed answers and consideration.
Regarding note 2, use BBEdit or TextSoap.app.
I would like to delete the HTML related strings.
Yours, WAKAMATSU
I fear a possible translation slip, are you saying that dar-select+copy text from a web page, when pasted into a BBEdit new file shows HTML mark-up? IF so I cannot replicate that. For me, using ⌘+V paste into BBEdit I see only the selected text of the source webpage and no HTML markup (i.e. no <p>, etc.).
If you do see HTML, I suspect you are selecting+copying from the ‘raw’ HTML source of the web-page and not the rendered version of the page as seen in the web browser when loading that URL. Above in my previous answers I am assuming you will use the browser’s rendered tend and not the ‘raw’ HTML.
I appreciate we are communicating through some translation, so some things may not come through as intended. Do please ask if wstill not getting the result you need. slight_smile:
Dear Dr. Mark Anderson,
Thank you for your careful consideration.
“I am assuming you will use the browser’s rendered tend and not the ‘raw’ HTML.”
The method you suggested is to copy the web page itself displayed in the browser in front of me. Is this my understanding correct?
I will use the “Copy and paste” along your method.
Yours, WAKAMATSU
Yes. I know that it will seem a bit counter-intuitive. What gets copied is the 'output of the HTML+CSS, i.e. styled text (font type, embedded links, etc.). Paste-and-match-style (Edit menu or ⌘+⌥+⇧+v), strips any of that style giving you only the text characters themselves.
Dear Dr. Mark Anderson,
Thanks a lot for your detailed explanation.
(Edit menu or ⌘+⌥+⇧+v),
（It is a bit of a nuisance, but）
As explained in other apps
Paste and Match Style (Scrivener, DEVONthink4.1.1, Melle 6.5.0, MacJournal 7)
Paste and Match Identation (BBEdit v15.5.3)
Paste as Plain Text (nvUltra v1.0.0)
Yours ,WAKAMATSU
Having discovered this un-styled form of paste, in apps that support it I use it more than ‘normal’ paste as generally I do not want the source style.
Tip: if you want paste without style in an app without that option, first paste your text into a plaintext editor like BBEdit (or similar). Then in the latter window select all, copy and paste the now un-styled text into your target application. so much easier than pasting in all sorts of weird styles you don’t want and then figuring out how to remove them.
A different approach is to use the lovely small utility TextSniper.
Nest. The TReX app is new to me but I see it at https://trex.ameba.co, in case of use to others also unaware.
It looks like TReX services a similar utility role as where I use TextSniper. One difference between the two that I see is TReX offers an automation hook whereby the detected text is send to some other app/process. Both apps allow addition of a custom dictionary (list of words) the user wishes detected as words, e.g. things like brand names or domain terms of art.
