## Samples.catalog Ucode sample viewer, by Ken Shillito Sample Unicode strings.... (click any of those below to view it) Simple placement of lone diacritics Hebrew I - א/ל + markings Hebrew II - שׁ/שׂ/terminals/rafe Hebrew III - dagesh, metheg, etc. Arabic I - character forming Arabic II - punctuation mirroring ##10 Indic Scripts Merging Diacritics Kerning and wrapping diacritics Double diacritics Individual under/overlines Multiple diacritics Decomposing Forced under/overlines Small caps, old-style nums Tilting, Emboldening ##20 Monospacing Justification, ASCII Tabbing Jam 1 Kerning Dingbats, etc. Unicode bidi Soft hyphens Font grabbing Manual input ##30 Simple placement of lone diacritics. Diacritics are shown here with a dummy glyph in the centre. ~Common placements are: above left/middle/right; left; right; and ~below left/middle/right. Each of these tight or loose. ~(there is no left tight). Hebrew I - א/ל + pointings Hebrew has complex rules for placing dagesh/mapiq within ~characters, relative positions of vowel points, cantillation ~marks and siluq. Lamedh has special placement rules. Hebrew II -שׁ/שׂ, Terminal forms, Rafe Sin and Shin present problems with holam (the placement of ~holam here looks ugly but is in accord with Unicode rules). ~Ucode must use terminal forms where required, and sort rafe ~and other diacritical marks into the correct order. Hebrew III - Dagesh/Mappiq, metheg, etc. Ucode must put dagesh/mappiq into its right place, place ~metheg into its right place, and deal with further sorting ~out of cantillation marks. ~I included some underlining and a box in the 4th line. Arabic I - Character shaping Arabic has complex rules for character shaping. Line 1 ~is an alphabet with forming suppressed; line 2 with shaping ~normal. Line 3 shows alif/lam ligatures. ##40 Arabic II - rtl mirroring of punctuation rtl languages (Hebrew, Arabic) mirror certain individual and ~paired punctuation. ltr do not. Ucode detects when mirroring ~is required, and does it automatically. The 3rd line has some ~enclosing diacritics, to show that Ucode does the joins ok. Indic scripts Some Indic scripts hang characters down from the line, with ~character forms varying depending on the context. Here ~are samples from these scripts. Some enclosing diacritics are ~included in order to show that Ucode joins through them ok. Merging diacritics Here are diacritics which occur over Latin or Greek ~characters, which interact with each others’ positioning. ~Ucode detects these, regardless of their order/context. Kerning and wrapping diacritics Some diacritics kern with their principal characters, or ~enclose them horizontally or vertically. Here are examples ~of each. In some cases, Ucode must also join the characters. Double diacritics Diacritics 0360 to 0362 cover two characters. Double ~diacritics are drawn in both character boxes, and have pixels ~between the boxes as in Arabic and Indic joiners. You will note ~that ucode.library does double diacritics ok for rtl Unicode. ##50 Individual under/overlines, strike throughs You can direct Ucode to do under/overlines/strike throughs on a ~whole passage, which can be a different colour/pattern from ~the text (as here). Or, the under/overline/strike through can be ~character by character. Or, a combination of both. Multiple diacritics This sample shows the letter “O” with various ~combinations of diacritics built around it. Some of these are ~rather odd, but they put ucode.library to the test. Decomposing ucode.library “decomposes” letters whose Unicodes have ~a built-in diacritic, e.g. 00C0 = 0041 + 0300. This gives ~better-looking but less compact results, especially for ~small glyph sizes or where resizing has taken place. Line 1 ~has decomposing suppressed, while line 2 is decomposed. Forced under/overlines (“rulings”), strike throughs Unlike “Individual under/overlines”, this sample ~forces entire lines to be under/overlined, etc. ~This is what the Unicode book calls a “higher ~protocol”, as Unicode itself only provides for ~rulings character by character. So the Unicodes do not ~contain the rulings; rather they are inserted as tags to ~TLUformat calls. Small capitals, old-style numerals Like “Forced under/overlines”, small caps are a higher ~protocol than Unicode. Small capitals are needed to ~implement CSS2, and old-style numerals look better in text. ~I shoe-horned their glyphs into glyphless Unicode positions. ~They are implemented as tags to TLUpreview calls. ##60 Emboldening and Tilting (Italic) ucode.library can recognise a limited set of embedded CTRL codes or ~HTML tags (e.g. <i> </i>), if they do NOT have ~attributes. These can be done algorithmically (as here), ~or using alternate fonts supplied by the caller. Monospacing You can direct TLUpreview to monospace the characters. If a ~character is too wide for the specified monospace, ~it will be trimmed. It is of course of course better to use a ~font specifically designed to be monospaced. Justification, ASCII The first line is centred. Subsequent lines are full justified. ~Justification can also be left or right. Unlike previous ~samples, which had input in UTF16 format, this sample ~has input in ASCII format. ucode.library does justification ~when a program calls TLUformat. Tabbing The Unicode standard says it “recognises” the TAB ~character 0009, but gives no specifics. ucode.libeary makes ~a default set of tab positions, and the caller can override ~these if required. LTR strings count tabs from the left, and ~RTL from the right. Jam1 ucode.library can overlay a background picture jam1 style ~if required. This example draws some stripes as background, ~and then calls ucode.library to overlay it with text. ##70 Kerning ** not yet implemented ** When kerning is implemented, there will be a tag to request ~kerning when TLUformat is called. Dingbats, etc. Unicode has all kinds of wierd and wonderful characters. ~Here are some of them. Unicode bidi Here is the same set of Unicodes 3 times. The first line has ~bidi suppressed by LRO...PDF; the second has normal bidi; the ~third has "AB...CD" embedded by LRE...PDF which causes the AB...CD ~to appear in the right order. Note also ~mirroring of punctuation. Soft hyphens This text contains some soft hyphens. Note also the ~HTML-style tags, which ucode.library can do, ~provided they do not have attributes. In order to see ~how the soft hyphens work, you may need to adjust the ~window width to cause the words to be broken. Font grabbing - Ruby.font must be in FONTS: In this example, I have grab an Amiga font into 0021-007E. ~Ruby.font must be in FONTS:, or else this Sample will not ~work. Note that all characters composed from a grabbed font ~also use the grabbed glyph. The font gets grabbed when TLUset ~is called. ##80 Manual input Ucode made your input into a UTF16 string. Any illegal characters ~were made into 0s. ~Finally, Ucode appended a null word delimiter. If you should ~prefer to input in ASCII (with &...; as per HTML), then put ~a ! at the start of your input, e.g. ~ !&ldquo;Hello, world&rdquo; Manual Input of a Unicode string... You are required to input 1 or more 4-digit hex numbers, ~optionally separated by spaces. The numbers may contain the characters 0-9 or a-f or A-F. Any ~invalid characters you specify will be replaced by 0s. You can interpolate spaces between numbers for readability. ~Do not include zero words (i.e. 0000) in your input. ~e.g. the Unicode for the string "abc" is "0061 0062 0063". Alternately, you can put a ‘!’ as the first ~character, and then input in ASCII format, with &...; ~characters, and the tags that ucode.library supports. ##90 e.g. !Hello, <i>world</i><br>Ciao, ~<b>&kappa;&omicron;&sigma; ~&mu;&omicron;&sigmaf; Now type your input: Error: TLUpreview / TLUformat failed Out of memory / required glyph files unavailable (Click to acknowledge) View Sample View UTF16/ASCII Menu ##100 Here is the UTF16/ASCII for this sample string (Click when finished viewing) Error: TLUmake failed - out of memory