This is an automated email from the ASF dual-hosted git repository.
delei pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/fesod.git
The following commit(s) were added to refs/heads/main by this push:
new 0add5679 refactor: simplify hex escape pattern scanning (#1003)
0add5679 is described below
commit 0add5679c01c5a400b9766a231cd96873b98f0cd
Author: Nikita Kuprins <[email protected]>
AuthorDate: Sun Aug 16 15:02:13 2026 +0300
refactor: simplify hex escape pattern scanning (#1003)
* refactor: optimize pattern matching and avoid unnecessary allocations
* refactor: extract pattern offset constants
* refactor: simplify hex escape scanning
* chore: update the comment
Co-authored-by: Copilot Autofix powered by AI
<[email protected]>
* refactor: extract escaped prefix constant
---------
Co-authored-by: ian zhang <[email protected]>
Co-authored-by: Copilot Autofix powered by AI
<[email protected]>
Co-authored-by: DeleiGuo <[email protected]>
---
.../write/handler/EscapeHexCellWriteHandler.java | 114 ++++++++-------------
1 file changed, 42 insertions(+), 72 deletions(-)
diff --git
a/fesod-sheet/src/main/java/org/apache/fesod/sheet/write/handler/EscapeHexCellWriteHandler.java
b/fesod-sheet/src/main/java/org/apache/fesod/sheet/write/handler/EscapeHexCellWriteHandler.java
index 6e5de6c2..dc172e2b 100644
---
a/fesod-sheet/src/main/java/org/apache/fesod/sheet/write/handler/EscapeHexCellWriteHandler.java
+++
b/fesod-sheet/src/main/java/org/apache/fesod/sheet/write/handler/EscapeHexCellWriteHandler.java
@@ -38,6 +38,21 @@ import org.apache.poi.xssf.streaming.SXSSFCell;
*/
public class EscapeHexCellWriteHandler implements CellWriteHandler {
+ // ASCII hex digits only. Not Character.digit(c, 16), which also accepts
non-ASCII
+ // digits such as U+0663 that OOXML never uses.
+ private static final boolean[] HEX_TABLE = new boolean[128];
+
+ static {
+ for (char c = '0'; c <= '9'; c++) HEX_TABLE[c] = true;
+ for (char c = 'A'; c <= 'F'; c++) HEX_TABLE[c] = true;
+ for (char c = 'a'; c <= 'f'; c++) HEX_TABLE[c] = true;
+ }
+
+ private static final String PREFIX = "_x";
+ private static final String ESCAPED_PREFIX = "_x005F" + PREFIX;
+ private static final int PREFIX_LENGTH = PREFIX.length();
+ private static final int HEX_DIGIT_COUNT = 4;
+
@Override
public void afterCellDataConverted(
WriteSheetHolder writeSheetHolder,
@@ -57,101 +72,56 @@ public class EscapeHexCellWriteHandler implements
CellWriteHandler {
}
}
- // Static hex lookup table for O(1) character validation
- private static final boolean[] HEX_TABLE = new boolean[128];
-
- static {
- for (char c = '0'; c <= '9'; c++) HEX_TABLE[c] = true;
- for (char c = 'A'; c <= 'F'; c++) HEX_TABLE[c] = true;
- for (char c = 'a'; c <= 'f'; c++) HEX_TABLE[c] = true;
- }
-
/**
- * Escapes hexadecimal-encoded strings with optimized performance Replaces
_xHHHH_ with _x005F_xHHHH_ to prevent POI
- * from decoding them
+ * Replaces every _xHHHH_ sequence with _x005F_xHHHH_ to prevent POI from
decoding them.
*/
private String escapeHex(String originalString) {
int length = originalString.length();
- // Fast path: if string is too short to contain pattern, return
original
- if (length < 7) {
- return originalString;
- }
-
- // Fast path: search for first potential pattern
+ // Lazily allocated: stays null (no allocation) when no valid pattern
is found
+ StringBuilder result = null;
+ int lastEnd = 0;
int searchStart = 0;
int patternIndex;
- while ((patternIndex = originalString.indexOf("_x", searchStart)) !=
-1) {
- // Check if we have enough characters for full pattern
- if (patternIndex + 6 >= length) {
+ while ((patternIndex = originalString.indexOf(PREFIX, searchStart)) !=
-1) {
+ int hexStart = patternIndex + PREFIX_LENGTH;
+ int suffixIndex = hexStart + HEX_DIGIT_COUNT;
+ int patternEnd = suffixIndex + 1;
+ // Too few characters left for a full pattern, and any later match
has even fewer
+ if (patternEnd > length) {
break;
}
- // Quick validation: check if it ends with '_' and has valid hex
- if (originalString.charAt(patternIndex + 6) == '_' &&
isValidHexFast(originalString, patternIndex + 2)) {
-
- // Found at least one pattern, proceed with full processing
- return processWithPatterns(originalString, patternIndex);
- }
-
- searchStart = patternIndex + 2;
- }
-
- // No valid patterns found
- return originalString;
- }
-
- /**
- * Process string when we know it contains at least one valid pattern
- */
- private String processWithPatterns(String originalString, int
firstPatternIndex) {
- int length = originalString.length();
- StringBuilder result = new StringBuilder(length + 64); // More
generous pre-allocation
- int lastEnd;
-
- // Process the first known pattern
- result.append(originalString, 0, firstPatternIndex);
- result.append("_x005F_x");
- result.append(originalString, firstPatternIndex + 2, firstPatternIndex
+ 6);
- result.append('_');
- lastEnd = firstPatternIndex + 7;
-
- // Continue searching for more patterns
- int searchStart = firstPatternIndex + 7;
- int patternIndex;
- while ((patternIndex = originalString.indexOf("_x", searchStart)) !=
-1) {
- if (patternIndex + 6 >= length) {
- break;
- }
-
- if (originalString.charAt(patternIndex + 6) == '_' &&
isValidHexFast(originalString, patternIndex + 2)) {
-
- // Append content between patterns
+ if (originalString.charAt(suffixIndex) == '_' &&
isHexDigits(originalString, hexStart)) {
+ if (result == null) {
+ result = new StringBuilder(length + 64);
+ }
+ // Append content since the previous match, then the escaped
pattern
result.append(originalString, lastEnd, patternIndex);
- // Append escaped pattern
- result.append("_x005F_x");
- result.append(originalString, patternIndex + 2, patternIndex +
6);
+ result.append(ESCAPED_PREFIX);
+ result.append(originalString, hexStart, suffixIndex);
result.append('_');
- lastEnd = patternIndex + 7;
- searchStart = patternIndex + 7;
+ lastEnd = patternEnd;
+ searchStart = patternEnd;
} else {
- searchStart = patternIndex + 2;
+ searchStart = hexStart;
}
}
- // Append remaining content
- if (lastEnd < length) {
- result.append(originalString, lastEnd, length);
+ // No valid patterns found
+ if (result == null) {
+ return originalString;
}
+ result.append(originalString, lastEnd, length);
return result.toString();
}
/**
- * Fast hex validation using lookup table - O(1) per character
+ * Checks whether the four characters starting at {@code startIndex} are
all ASCII hex digits.
*/
- private static boolean isValidHexFast(String str, int startIndex) {
- for (int i = 0; i < 4; i++) {
+ private static boolean isHexDigits(String str, int startIndex) {
+ for (int i = 0; i < HEX_DIGIT_COUNT; i++) {
char c = str.charAt(startIndex + i);
if (c >= 128 || !HEX_TABLE[c]) {
return false;
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
