On 7/20/19 5:41 PM, Troy A. Griffitts wrote:
>
> Michael, I would be interested to hear your logic for internationalizing a
> numeric parser. I always assumed it wasn't as simply as providing alternate
> symbol codes for 0-9. Think Roman Numerals (SWORD already supports Roman
> Numerals). If it was as simply as swapping out, say 47 with an alternate
> symbol for '4' and an alternate symbol for '7', it wouldn't be much of a
> problem to solve, but I suspect other languages do things like: XLVII.
>
Digit swapping works for most languages, but not for Ethiopic, Roman, and
Coptic. There may be others that I haven't encountered, yet. LGPL C# code that
I wrote follows for going from Hindu-Arabic (0-9, like you are most used to) to
a variety of number systems:
// The following linguistically diverse digit sets might not display
correctly unless
// you have appropriate Unicode fonts installed.
public const string ArabicDigits = "٠١٢٣٤٥٦٧٨٩";
public const string BengaliDigits = "০১২৩৪৫৬৭৮৯";
public const string ChineseSimplifiedDigits = "〇一二三四五六七八九";
public const string ChineseTraditionalDigits = "零壹貳參肆伍陸柒捌玖";
public const string ChineseHuaMaDigits = "〇〡〢〣〤〥〦〧〨〩";
public const string CopticUnits =
"\u2c81\u0305\u2C83\u0305\u2C85\u0305\u2C87\u0305\u2C89\u0305\u2C8B\u0305\u2C8D\u0305\u2C8F\u0305\u2C91\u0305";
// First 9 coptic letters with single overbar
public const string CopticTens =
"\u2C93\u0305\u2C95\u0305\u2C97\u0305\u2C99\u0305\u2C9B\u0305\u2C9D\u0305\u2C9F\u0305\u2CA1\u0305\u03E5\u0305";
// 10th thru 18th coptic letters with single overbar
public const string CopticHundreds =
"\u2CA3\u0305\u2CA5\u0305\u2CA7\u0305\u2CA9\u0305\u2CAB\u0305\u2CAD\u0305\u2CAF\u0305\u2CB1\u0305\u2CB3\u0305\u2CB5\u0305";
// 19th thru 27th coptic letters with single overbar
public const string CopticThousands =
"\u2c81\u033F\u2C83\u033F\u2C85\u033F\u2C87\u033F\u2C89\u033F\u2C8B\u033F\u2C8D\u033F\u2C8F\u033F\u2C91\u033F";
// Thousands are the same as units, but double overbar.
public const string CopticTenThousands =
"\u2C93\u033F\u2C95\u033F\u2C97\u033F\u2C99\u033F\u2C9B\u033F\u2C9D\u033F\u2C9F\u033F\u2CA1\u033F\u03E5\u033F";
// Pattern continues: one bar added per period (10^3)
public const string DevangariDigits = "०१२३४५६७८९";
public const string EthiopicDigits = " ፩፪፫፬፭፮፯፰፱";
public const string EthiopicTens = " ፲፳፴፵፶፷፸፹፺";
public const string EthiopicHundred = "፻";
public const string EthiopicTenThousand = "፼";
public const string GugaratiDigits = "૦૧૨૩૪૫૬૭૮૯";
public const string GurmukhiDigits = "੦੧੨੩੪੫੬੭੮੯";
public const string KannadaDigits = "೦೧೨೩೪೫೬೭೮೯";
public const string KhmerDigits = "០១២៣៤៥៦៧៨៩";
public const string LaoDigits = "໐໑໒໓໔໕໖໗໘໙";
public const string LimbuDigits = "᥆᥇᥈᥉᥊᥋᥌᥍᥎᥏";
public const string MalayalamDigits = "൦൧൨൩൪൫൬൭൮൯";
public const string MongolianDigits = "᠐᠑᠒᠓᠔᠕᠖᠗᠘᠙";
public const string BurmeseDigits = "၀၁၂၃၄၅၆၇၈၉";
public const string OriyaDigits = "୦୧୨୩୪୫୬୭୮୯";
public const string PersianDigits = "۰۱۲۳۴۵۶۷۸۹"; // Same as Urdu
digits
public const string TamilDigits = "௦௧௨௩௪௫௬௭௮௯";
public const string TeluguDigits = "౦౧౨౩౪౫౬౭౮౯";
public const string ThaiDigits = "๐๑๒๓๔๕๖๗๘๙";
public const string TibetanDigits = "༠༡༢༣༤༥༦༧༨༩";
public const string UrduDigits = "۰۱۲۳۴۵۶۷۸۹";
public const string RomanDigits = " ⅠⅡⅢⅣⅤⅥⅦⅧⅨ";
protected static string CurrentDigits = String.Empty;
public static string NumberSample()
{
if (!String.IsNullOrEmpty(CurrentDigits))
return CurrentDigits;
else
return "0123456789";
}
/// <summary>
/// true iff we are changing digits to an alternate writing system
/// </summary>
public static bool LocalizingDigits
{
get { return CurrentDigits != String.Empty; }
}
/// <summary>
/// Set the locale for localizing digits for display in Bibles for
verse numbers, etc.
/// </summary>
/// <param name="digitPlace">string with one of the exact names of
supported digit sets</param>
/// <returns>the set string if successful, or "Default"
otherwise</returns>
public static string SetDigitLocale(string digitPlace)
{
switch (digitPlace)
{
case "Arabic":
CurrentDigits = ArabicDigits;
break;
case "Bengali":
CurrentDigits = BengaliDigits;
break;
case "Burmese (Myanmar)":
CurrentDigits = BurmeseDigits;
break;
case "Chinese (Simplified)":
CurrentDigits = ChineseSimplifiedDigits;
break;
case "Chinese (Traditional)":
CurrentDigits = ChineseTraditionalDigits;
break;
case "Chinese (hua ma)":
CurrentDigits = ChineseHuaMaDigits;
break;
case "Coptic":
CurrentDigits = CopticUnits;
break;
case "Devangari":
CurrentDigits = DevangariDigits;
break;
case "Ethiopic (Ge'ez)":
CurrentDigits = EthiopicDigits;
break;
case "Gujarati":
CurrentDigits = GugaratiDigits;
break;
case "Gurmukhi":
CurrentDigits = GurmukhiDigits;
break;
case "Kannada":
CurrentDigits = KannadaDigits;
break;
case "Khmer":
CurrentDigits = KhmerDigits;
break;
case "Lao":
CurrentDigits = LaoDigits;
break;
case "Limbu":
CurrentDigits = LimbuDigits;
break;
case "Malayalam":
CurrentDigits = MalayalamDigits;
break;
case "Mongolian":
CurrentDigits = MongolianDigits;
break;
case "Oriya":
CurrentDigits = OriyaDigits;
break;
case "Roman":
CurrentDigits = RomanDigits;
break;
case "Tamil":
CurrentDigits = TamilDigits;
break;
case "Telugu":
CurrentDigits = TeluguDigits;
break;
case "Thai":
CurrentDigits = ThaiDigits;
break;
case "Tibetan":
CurrentDigits = TibetanDigits;
break;
case "Persian":
case "Urdu":
CurrentDigits = UrduDigits;
break;
case "Hindu-Arabic":
case "Default":
default:
CurrentDigits = String.Empty;
digitPlace = "Default";
break;
}
return digitPlace;
}
/// <summary>
/// Replaces all numbers with appropriate numbers in the current
writing system
/// </summary>
/// <param name="s">string that might include numbers</param>
/// <returns>string with numbers localized</returns>
public static string LocalizeDigits(string s)
{
return ReplaceDigits(s, CurrentDigits);
}
/// <summary>
/// Some writing systems just have exact equivalents for 0 through 9
and the same place values.
/// Those are easy, with a simple digit-for-digit substitution. Others
require some logic beyond
/// that.
/// </summary>
/// <param name="s">String that may have digits to localize</param>
/// <param name="newDigits">one of the supported digit strings</param>
/// <returns></returns>
public static string ReplaceDigits(string s, string newDigits)
{ // TODO: implement logic for the different Chinese numeral systems,
which require more than simple digit substitution, and which have many dialect
and usage options.
if ((newDigits == null) || (newDigits.Length < 10))
{ // Nothing to do; no conversion specified
return s;
}
if (newDigits == EthiopicDigits)
{ // Gotta count differently
return EthiopicNumerals(s);
}
else if (newDigits == RomanDigits)
{ // Seriously old school
return RomanNumerals(s);
}
else if (newDigits == CopticUnits)
{ // Older than Roman Numerals, but not the same as Heirogliphic
numbers or ancient Egyptian numbers
return CopticNumerals(s);
}
else
{ // Simple digit substitution with normal place values
StringBuilder sb = new StringBuilder();
int n;
foreach (char c in s)
{
n = ((int)c) - ((int)'0');
if ((n >= 0) && (n <= 9))
{
sb.Append(newDigits[n]);
}
else
{
sb.Append(c);
}
}
return sb.ToString();
}
}
/// <summary>
/// If the input ch is a localized digit in the string localDigits,
return a digit in the range '0'-'9',
/// otherwise return the input character.
/// </summary>
/// <param name="ch">Possible localized digit</param>
/// <param name="localDigits">String of 0-9 in local digits.</param>
/// <returns>Standardized digit or input character</returns>
public static char StandardDigit(char ch)
{
char result = ch;
int i = CurrentDigits.IndexOf(ch);
if (i >= 0)
result = (char)(i + (int)'0');
return result;
}
/// <summary>
/// Coptic numbers have no 0, but have different symbols for units,
tens, hundreds, thousands, etc.
/// </summary>
/// <param name="s">Digits to convert less than or equal to 999</param>
/// <returns>String with coptic numerals</returns>
public static string CopticNumerals(string s)
{
StringBuilder sb = new StringBuilder();
int i, n;
int place = 0;
for (i = s.Length - 1; i >= 0; i--)
{
if (Char.IsDigit(s[i]))
{
if (s[i] == '0')
{
place++;
}
else
{
n = 2 * (((int)s[i]) - ((int)'1'));
if (place == 0)
{
sb.Insert(0, CopticUnits[n + 1]); // Letter
sb.Insert(0, CopticUnits[n]); // Combining overbar
place++;
}
else if (place == 1)
{
sb.Insert(0, CopticTens[n + 1]);
sb.Insert(0, CopticTens[n]);
place++;
}
else if (place == 2)
{
sb.Insert(0, CopticHundreds[n + 1]);
sb.Insert(0, CopticHundreds[n]);
place++;
}
else if (place == 3)
{
sb.Insert(0, CopticThousands[n + 1]);
sb.Insert(0, CopticThousands[n]);
place++;
}
else if (place == 4)
{
sb.Insert(0, CopticTenThousands[n + 1]);
sb.Insert(0, CopticTenThousands[n]);
place++;
}
else if (place >= 5)
{
return s; // Give up and fail gracefully for
numbers bigger than we designed for.
}
}
}
else
{ // Non-digit: just copy it.
place = 0;
sb.Insert(0, s[i]);
}
}
return sb.ToString();
}
/// <summary>
/// Writing big numbers without a 0 is different.
/// </summary>
/// <param name="s">String with numbers to localize to Ethiopic</param>
/// <returns>String with numbers localized to Ethiopic Ge'ez.</returns>
public static string EthiopicNumerals(string s)
{
StringBuilder sb = new StringBuilder();
int i, n;
int place = 0;
for (i = s.Length - 1; i >= 0; i--)
{
if (Char.IsDigit(s[i]))
{
n = ((int)s[i]) - ((int)'0');
if (place == 0)
{
if (n > 0)
{
sb.Insert(0, EthiopicDigits[n]);
}
place++;
}
else if (place == 1)
{
if (n > 0)
{
sb.Insert(0, EthiopicTens[n]);
}
place++;
}
else if (place == 2)
{
sb.Insert(0, EthiopicHundred);
if (n > 0)
{
sb.Insert(0, EthiopicDigits[n]);
}
place++;
}
else if (place == 3)
{
if (n > 0)
{
sb.Insert(0, EthiopicTens[n]);
}
place++;
}
else if (place == 4)
{
sb.Insert(0, EthiopicTenThousand);
if (n > 0)
{
sb.Insert(0, EthiopicDigits[n]);
}
place++;
}
else if (place == 5)
{
if (n > 0)
{
sb.Insert(0, EthiopicDigits[n]);
}
place = 0;
}
}
else
{
place = 0;
sb.Insert(0, s[i]);
}
}
return sb.ToString();
}
/// <summary>
/// Roman numerals kind of break down after 3,000 (MMM) in terms of
common use,
/// which seems to be limited to chapters and years in this decade.
This function
/// uses Unicode Roman numerals, but an alternate routine could easily
be created
/// that uses plain letters I, V, X, L, C, and M or i, v, x, l, c, and
m by replacing
/// or providing choices for the strings in RomanUnits in this method.
/// </summary>
/// <param name="s">string that may have numbers to "Romanize"</param>
/// <returns>string with numbers as Roman numerals</returns>
public static string RomanNumerals(string s)
{
string[,] RomanUnits = {{"","Ⅰ","Ⅱ","Ⅲ","Ⅳ","Ⅴ","Ⅵ","Ⅶ","Ⅷ","Ⅸ"},
{"", "Ⅹ","ⅩⅩ","ⅩⅩⅩ","ⅩⅬ","Ⅼ","ⅬⅩ","ⅬⅩⅩ","ⅬⅩⅩⅩ","ⅩⅭ"},
{"", "Ⅽ", "ⅭⅭ", "ⅭⅭⅭ", "ⅭⅮ", "Ⅾ", "ⅮⅭ", "ⅮⅭⅭ", "ⅮⅭⅭⅭ", "ⅩⅯ"},
{ "", "Ⅿ", "ⅯⅯ", "ⅯⅯⅯ", "ⅯV̅", "V̅", "V̅Ⅿ", "V̅ⅯⅯ", "V̅ⅯⅯⅯ",
"ⅯX̅̅"}};
StringBuilder sb = new StringBuilder();
int i, n;
int place = 0;
for (i = s.Length - 1; i >= 0; i--)
{
if (Char.IsDigit(s[i]))
{
n = ((int)s[i]) - ((int)'0');
sb.Insert(0, RomanUnits[place, n]);
place++;
if (place > 3)
place = 0;
}
else
{
place = 0;
sb.Insert(0, s[i]);
}
}
sb.Replace("ⅩⅡ", "Ⅻ");
sb.Replace("ⅩⅠ", "Ⅺ");
return sb.ToString();
}
> On 7/20/19 1:49 PM, Michael Johnson wrote:
>> It is an important question. I have a way to handle it in other formats
>> using Haiola. I suspect that proper handling it in SWORD will take a design
>> change. In Haiola, the source chapter and verse numbers are always as in
>> English. I can select display chapter and verse numbers from many options.
>>
>> Aloha,
>> Michael
>> http://mpj.us
>>
>>
>> On Jul 20, 2019, at 10:39, Cyrille <lafricai...@gmail.com
>> <mailto:lafricai...@gmail.com>> wrote:
>>
>>> Hello,
>>> No body else have a proposition for this question? Should I open a bug
>>> report?
>>> Is it not an important question?
>>>
>>> Best regards, Br Cyrille
>>>
>>> Le 17/07/2019 à 23:14, David Haslam a écrit :
>>>> Several other non-Roman scripts have their own digit characters
>>>> corresponding to our 0-9.
>>>>
>>>> IMHO the possibilities for using non-Roman digits ought to be facilitated
>>>> in the back-end.
>>>>
>>>> Even so, each front-end would then require a new UI control to select
>>>> which script should be used to display the chapter and verse numbers.
>>>>
>>>> An alternate idea would be to specify the non-Roman digits as a ten
>>>> character UTF-8 string in a module .conf file.
>>>>
>>>> Adapting the back-end to use this module specific configuration key might
>>>> be much simpler.
>>>>
>>>> Front-ends would still require adapting for the UI features that require
>>>> chapter and verse numbers to be input or displayed or adjusted by
>>>> controls.
>>>>
>>>> Best regards,
>>>>
>>>> David
>>>>
>>>> Sent from ProtonMail Mobile
>>>>
>>>>
>>>> On Wed, Jul 17, 2019 at 22:00, Cyrille <lafricai...@gmail.com
>>>> <mailto:lafricai...@gmail.com>> wrote:
>>>>> Hello,
>>>>> I'm still working on a modern NT-Ps-Pr translation in Burmese. My friends
>>>>> from Myanmar send me the text. But they don't use the arab numbers, they
>>>>> hava their own numbers.
>>>>> It could be very important for them to write in they own numbers (If I
>>>>> had tu use their I will be lost ;) ).
>>>>> Is it possible to add this possibility to the frontend, or this should be
>>>>> in sword directly?
>>>>> Need I to open a new issue on the bug tracker?
>>>>>
>>>>> Example of text, in bold the chapter and verses:
>>>>>
>>>>> ၃ ၁။ ထိုနေ့ရက်တို့၌ ယောဟန်ဘတ္တိဇံသည် ရောက်လာ၍ ဂျူဒေးယပြည်၊ တောကန္တာရတွင်
>>>>> ဟော
>>>>>
>>>>> *၂။* ပြောသည်မှာ၊- နောင်တရကြလော့၊ အကြောင်းမူကား ကောင်းကင်နိုင်ငံတော်သည်
>>>>> ရောက်လုနီးပြီဟူ၍တည်း။-
>>>>>
>>>>> *၃။* ပရောဖက်အီဇာယဟောထားခဲ့သည့်အတိုင်း၊ ထာ၀ရ
>>>>> ဘုရားသခင်ကြွလာတော်မူမည့်လမ်းကိုပြင်ဆင်ကြ လော့၊ ကိုယ်တော်၏ လမ်းများကို
>>>>> ဖြောင့်တန်းစေကြ လော့ဟူ၍ တောကန္တာရ၌ ကြွေးကြော်သောသူ၏အသံသည်ကား
>>>>> ဤသူပင်ဖြစ်သတည်း။-
>>>>>
>>>>> *၄။* ယောဟန်သည် ကုလားအုတ်မွေးဖြင့် ရက်လုပ်သောအဝတ်ကိုဝတ်ဆင်ကာ ခါး၌
>>>>> သားရေခါးစည်းကြိုးကိုစည်းထား၏။ သူ၏အစာသည်ကား ကျိုင်းကောင်နှင့်
>>>>> တောပျားရည်တို့သာဖြစ်၏။-
>>>>>
>>>>> *၅။* ထိုအခါ ဂျေရုဆလင်မြို့မှစ၍ ဂျူဒေးယနယ်တစ်နယ် လုံးနှင့်
>>>>> ဂျော်ဒန်မြစ်တစ်လျှောက်ရှိဒေသမှ လူအပေါင်း တို့သည် သူ့ထံသို့ လာကြ၏။-
>>>>>
>>>>> ၆။ ထိုသူတို့သည် မိမိတို့၏ အပြစ်များကို ထုတ်ဖော်ဝန်ခံကြလျက်
>>>>> ဂျော်ဒန်မြစ်တွင် သူ့အားဖြင့် ဆေးကြောခြင်းကို ခံယူကြ၏။
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> sword-devel mailing list: sword-devel@crosswire.org
>>>> http://www.crosswire.org/mailman/listinfo/sword-devel
>>>> Instructions to unsubscribe/change your settings at above page
>>>
>>> _______________________________________________
>>> sword-devel mailing list: sword-devel@crosswire.org
>>> <mailto:sword-devel@crosswire.org>
>>> http://www.crosswire.org/mailman/listinfo/sword-devel
>>> Instructions to unsubscribe/change your settings at above page
>>
>> _______________________________________________
>> sword-devel mailing list: sword-devel@crosswire.org
>> http://www.crosswire.org/mailman/listinfo/sword-devel
>> Instructions to unsubscribe/change your settings at above page
>
> _______________________________________________
> sword-devel mailing list: sword-devel@crosswire.org
> http://www.crosswire.org/mailman/listinfo/sword-devel
> Instructions to unsubscribe/change your settings at above page
--
signature
Aloha,
*/Michael Johnson/**
PO BOX 881143 • PUKALANI HI 96788-1143*• USA
mljohnson.org <http://mljohnson.org> • Phone: +1 808-333-6921 • Skype:
kahunapule
_______________________________________________
sword-devel mailing list: sword-devel@crosswire.org
http://www.crosswire.org/mailman/listinfo/sword-devel
Instructions to unsubscribe/change your settings at above page
