Laboratoire ICAR » Plateforme TXM

// Copyright © 2022 ENS de Lyon - https://www.textometrie.org

// @author sheiden

package org.txm.macro.txt

// STANDARD DECLARATIONS

import java.lang.Character

import java.awt.Font

import java.awt.event.KeyEvent

def TypeCode2TypeName(tc) {

	types = [ Character.COMBINING_SPACING_MARK, Character.CONNECTOR_PUNCTUATION, Character.CONTROL,

                Character.CURRENCY_SYMBOL, Character.DASH_PUNCTUATION, Character.DECIMAL_DIGIT_NUMBER,

                Character.ENCLOSING_MARK, Character.END_PUNCTUATION, Character.FORMAT, Character.LETTER_NUMBER,

                Character.LINE_SEPARATOR, Character.LOWERCASE_LETTER, Character.MATH_SYMBOL,

                Character.MODIFIER_SYMBOL, Character.NON_SPACING_MARK, Character.OTHER_LETTER,

                Character.OTHER_NUMBER, Character.OTHER_PUNCTUATION, Character.OTHER_SYMBOL,

                Character.PARAGRAPH_SEPARATOR, Character.PRIVATE_USE, Character.SPACE_SEPARATOR,

                Character.START_PUNCTUATION, Character.SURROGATE, Character.TITLECASE_LETTER, Character.UNASSIGNED,

                Character.UPPERCASE_LETTER ]

	typeNames = [ "Combining spacing mark", "Connector punctuation", "Control", "Currency symbol",

                "Dash punctuation", "Decimal digit number", "Enclosing mark", "End punctuation", "Format",

                "Letter number", "Line separator", "Lowercase letter", "Math symbol", "Modifier symbol",

                "Non spacing mark", "Other letter", "Other number", "Other punctuation", "Other symbol",

                "Paragraph separator", "Private use", "Space separator", "Start punctuation", "Surrogate",

                "Titlecase letter", "Unassigned", "Uppercase letter" ]

	for (int i = 0; i < types.size(); i++) {

        	if (tc == types[i]) {

            	return(typeNames[i])

            	}

	}

	return "Unknown"

}

def DirectionalityCode2DirectionalityName(dc) {

	directionalities = [

		Character.DIRECTIONALITY_ARABIC_NUMBER,

		Character.DIRECTIONALITY_BOUNDARY_NEUTRAL,

		Character.DIRECTIONALITY_COMMON_NUMBER_SEPARATOR,

		Character.DIRECTIONALITY_EUROPEAN_NUMBER,

		Character.DIRECTIONALITY_EUROPEAN_NUMBER_SEPARATOR,

		Character.DIRECTIONALITY_EUROPEAN_NUMBER_TERMINATOR,

		Character.DIRECTIONALITY_LEFT_TO_RIGHT,

		Character.DIRECTIONALITY_LEFT_TO_RIGHT_EMBEDDING,

		Character.DIRECTIONALITY_LEFT_TO_RIGHT_OVERRIDE,

		Character.DIRECTIONALITY_NONSPACING_MARK,

		Character.DIRECTIONALITY_OTHER_NEUTRALS,

		Character.DIRECTIONALITY_PARAGRAPH_SEPARATOR,

		Character.DIRECTIONALITY_POP_DIRECTIONAL_FORMAT,

		Character.DIRECTIONALITY_RIGHT_TO_LEFT,

		Character.DIRECTIONALITY_RIGHT_TO_LEFT_ARABIC,

		Character.DIRECTIONALITY_RIGHT_TO_LEFT_EMBEDDING,

		Character.DIRECTIONALITY_RIGHT_TO_LEFT_OVERRIDE,

		Character.DIRECTIONALITY_SEGMENT_SEPARATOR,

		Character.DIRECTIONALITY_UNDEFINED,

		Character.DIRECTIONALITY_WHITESPACE

	]

	directionalitiesNames = [

		"Arabic Number",

		"Boundary Neutral",

		"Common Number Separator",

		"European Number",

		"European Number Separator",

		"European Number Terminator",

		"Left To Right",

		"Left To Right Embedding",

		"Left To Right Override",

		"Nonspacing Mark",

		"Other Neutrals",

		"Paragraph Separator",

		"Pop Directional Format",

		"Right To Left",

		"Right To Left Arabic",

		"Right To Left Embedding",

		"Right To Left Override",

		"Segment Separator",

		"Undefined",

		"Whitespace"

	]

	for (int i = 0; i < directionalities.size(); i++) {

        	if (dc == directionalities[i]) {

            	return(directionalitiesNames[i])

            	}

	}

	return "Unknown"

}

def isPrintableChar(c) {

    Character.UnicodeBlock block = Character.UnicodeBlock.of( c )

    return (!Character.isISOControl(c)) &&

            c != KeyEvent.CHAR_UNDEFINED &&

            block != null &&

            block != Character.UnicodeBlock.SPECIALS;

}

monitor.syncExec(new Runnable() {

	public void run() {

		try {

			s = org.txm.rcp.utils.IOClipboard.read()

			if (s == null || s.length() == 0) {

				println "No text in clipboard. Aborting."

				return

			}

			l = s.length()

			maxcodeL = 0

			maxnameL = 0

			maxtypeL = 0

			maxscriptL = 0

			maxblockL = 0

			maxdirectL = 0

			for (int i = 0; i < l; i++) {

				c = s.substring(i, i+1)

				cp = s.codePointAt(i)

				name = Character.getName(cp)

				type = TypeCode2TypeName(Character.getType(cp))

				script = Character.UnicodeScript.of(cp).name()

				block = Character.UnicodeBlock.of(cp)

				direct = DirectionalityCode2DirectionalityName(Character.getDirectionality(cp))

				cpsl = cp.toString().length()

				if (cpsl > maxcodeL) maxcodeL = cpsl

				namesl = name.toString().length()

				if (namesl > maxnameL) maxnameL = namesl

				typesl = type.toString().length()

				if (typesl > maxtypeL) maxtypeL = typesl

				scriptsl = script.toString().length()

				if (scriptsl > maxscriptL) maxscriptL = scriptsl

				blocksl = block.toString().length()

				if (blocksl > maxblockL) maxblockL = blocksl

				directsl = direct.toString().length()

				if (directsl > maxdirectL) maxdirectL = directsl

			}

			if (maxcodeL < 4) maxcodeL = 4

			if (maxnameL < 4) maxnameL = 4

			if (maxtypeL < 4) maxtypeL = 4

			if (maxscriptL < 6) maxscriptL = 6

			if (maxblockL < 5) maxblockL = 5

			if (maxdirectL < 14) maxdirectL = 14

			println sprintf("%-15s %${maxcodeL}s %-${maxnameL}s %-${maxtypeL}s %-${maxscriptL}s %-${maxblockL}s %-${maxdirectL}s", "Char", "Code", "Name", "Type", "Script", "Block", "Directionality")

			for (int i = 0; i < l; i++) {

				c = s.substring(i, i+1)

				cp = s.codePointAt(i)

				name = Character.getName(cp)

				type = TypeCode2TypeName(Character.getType(cp))

				script = Character.UnicodeScript.of(cp).name()

				block = Character.UnicodeBlock.of(cp)

				direct = DirectionalityCode2DirectionalityName(Character.getDirectionality(cp))

				if (isPrintableChar(c.charAt(0))) {

					println sprintf("%-15s %${maxcodeL}s %-${maxnameL}s %-${maxtypeL}s %-${maxscriptL}s %-${maxblockL}s %-${maxdirectL}s", c, cp, name, type, script, block, direct)

				} else {

					println sprintf("<not printable> %${maxcodeL}s %-${maxnameL}s %-${maxtypeL}s %-${maxscriptL}s %-${maxblockL}s %-${maxdirectL}s", cp, name, type, script, block, direct)

				}				

			}

/*

			for (char c : s) {

				charValue = c.charValue()

				charName = java.lang.Character.getName(c)

				numericValue = java.lang.Character.getNumericValue(c)

				charType = java.lang.Character.getType(c)

			}

*/

		} catch (Exception e) {

			e.printStackTrace()

		}

	}

})