Archives/gnosygnu_xowa
1
0
Fork 0
mirror of https://github.com/gnosygnu/xowa.git synced 2026-03-02 03:49:30 +00:00
Code Issues Projects Releases Wiki Activity

Scribunto: Do not fail in ustring.find if negative bgn is large [#366]

Browse Source
This commit is contained in:
gnosygnu
2019-02-24 16:14:34 -05:00
parent 91cbb34fa5
commit 3fd759b020
7 changed files with 392 additions and 180 deletions
Download Patch File Download Diff File Expand all files Collapse all files

49
100_core/src/gplx/core/intls/Utf8_.java
View File

@@ -88,7 +88,7 @@ public class Utf8_ {
for (int i = pos - 1; i >= stop; i--) { // start at pos - 1, and move bwd; NOTE: pos - 1 to skip pos, b/c pos will never definitively yield any char_len info
byte b = bry[i];
int char_len = Len_of_char_by_1st_byte(b);
switch (char_len) { // if char_len is multi-byte and pos is at correct multi-byte pos (pos - i = # of bytes - 1), then pos0 found; EX: <EFBFBD> = {226,130,172}; 172 is skipped; 130 has len of 1 -> continue; 226 has len of 3 and is found at correct pos for 3 byte char -> return
switch (char_len) { // if char_len is multi-byte and pos is at correct multi-byte pos (pos - i = # of bytes - 1), then pos0 found; EX: € = {226,130,172}; 172 is skipped; 130 has len of 1 -> continue; 226 has len of 3 and is found at correct pos for 3 byte char -> return
case 2: if (pos - i == 1) return i; break;
case 3: if (pos - i == 2) return i; break;
case 4: if (pos - i == 3) return i; break;
@@ -111,7 +111,7 @@ public class Utf8_ {
// loop maximum of 4 times; note that UTF8 char has max of 4 bytes
for (int i = 0; i < 4; i++) {
int char_len = Len_of_char_by_1st_byte(b);
switch (char_len) { // if char_len is multi-byte and cur is at correct multi-byte pos (cur - i = # of bytes - 1), then pos0 found; EX: <EFBFBD> = {226,130,172}; 172 is skipped; 130 has len of 1 -> continue; 226 has len of 3 and is found at correct cur for 3 byte char -> return
switch (char_len) { // if char_len is multi-byte and cur is at correct multi-byte pos (cur - i = # of bytes - 1), then pos0 found; EX: € = {226,130,172}; 172 is skipped; 130 has len of 1 -> continue; 226 has len of 3 and is found at correct cur for 3 byte char -> return
case 2: if (i == 1) return pos; break;
case 3: if (i == 2) return pos; break;
case 4: if (i == 3) return pos; break;
@@ -141,3 +141,48 @@ public class Utf8_ {
, Codepoint_surrogate_end = 0xDFFF
;
}
/*
== Definitions ==
=== a7 vs u8 ===
* a7 -> ASCII (7 bits)
* u8 -> UTF-8 (8 bytes)
In retrospect, better abbreviations would have been:
* ascii -> ASCII
* utf08 -> UTF-8
* utf16 -> UTF-16
=== General ===
==== Byte ====
* Standard definition; 8 bits (2^8 or 256)
==== Codepoint ====
* Represents 1 atomic character but can be composed of multiple bytes
** Examples:
<pre>
1 byte : "a" (letter a)
2 bytes: "¢" (cent)
3 bytes: "€" (euro)
4 bytes: "𤭢" (Chinese character)
</pre>
* Defined by unicode as a sequence of 4 hexadecimals (2 bytes) or 8 hexadecimals (4 bytes); REF:http://www.unicode.org
** 4 hexadecimal is 2 bytes (2^(4 * 4) -> 2^16)
==== char ====
* Java definition of a codepoint which is encoded as 2 bytes (2^16 or 65,536)
* For Western langauges: 1 codepoint equals 1 char (2 bytes);
** For example, chars like "a", "œ", "é" are 1 Java char
* For Eastern langauges: 1 codepoint can equal 2 chars (4 bytes);
** For example, chars like "駣" are 2 Java chars though they represent 1 conceptual codepoint (in English terms, "駣" is a single letter just like the letter "a")
==== Supplementary characters ====
* Represents a codepoint which is defined by 3 or 4 bytes
* Is defined by 1 surrogate pair
** lo-surrogate : 2 bytes
** hi-surrogate : 2 bytes
=== Conventions ===
* Codepoints will be rendered as one int (4 bytes), not 4 hexadecimals (1 byte) 8 hexadecimal (4 bytes)
* The "char" datatype will rarely be used in code; instead byte arrays or codepoint-ints will be used
* The "character" word will not be used in comments; instead the "codepoint" word will be used
*/