A fork of https://github.com/crosspoint-reader/crosspoint-reader
fix: Prevent line breaks on common English contractions (#1405)
authored by
Zach Nelson
and committed by
GitHub
dc394803
b5df6cb2
+4
-4
+2
-2
lib/Epub/Epub/hyphenation/Hyphenator.cpp
+2
-2
lib/Epub/Epub/hyphenation/Hyphenator.cpp
···
102
102
void appendApostropheContractionBreaks(const std::vector<CodepointInfo>& cps,
103
103
std::vector<Hyphenator::BreakInfo>& outBreaks) {
104
104
constexpr size_t kMinLeftSegmentLen = 3;
105
-
constexpr size_t kMinRightSegmentLen = 2;
105
+
constexpr size_t kMinRightSegmentLen = 3;
106
106
size_t segmentStart = 0;
107
107
108
108
for (size_t i = 0; i < cps.size(); ++i) {
···
123
123
}
124
124
}
125
125
126
-
// Avoid stranding short clitics like "l'"/"d'" or tiny suffixes like "'t".
126
+
// Avoid stranding short clitics like "l'"/"d'" or contraction tails like "'ve"/"'re"/"'ll".
127
127
if (leftPrefixLen >= kMinLeftSegmentLen && rightSuffixLen >= kMinRightSegmentLen) {
128
128
outBreaks.push_back({cps[i + 1].byteOffset, false});
129
129
}
+2
-2
lib/Epub/Epub/hyphenation/Hyphenator.h
+2
-2
lib/Epub/Epub/hyphenation/Hyphenator.h
···
26
26
// 2. Apostrophe contractions between letters (e.g. all'improvviso).
27
27
// Liang patterns are run per alphabetic segment around apostrophes.
28
28
// A direct break at the apostrophe boundary is allowed only when the left
29
-
// segment has at least 3 letters and the right segment has at least 2 letters,
30
-
// avoiding short clitics (e.g. l', d') and short contraction tails (e.g. can't).
29
+
// segment has at least 3 letters and the right segment has at least 3 letters,
30
+
// avoiding short clitics (e.g. l', d') and contraction tails (e.g. 've, 're, 'll).
31
31
// 3. Language-specific Liang patterns (e.g. German de_patterns).
32
32
// Example: "Quadratkilometer" -> Qua|drat|ki|lo|me|ter.
33
33
// 4. Fallback every-N-chars splitting (only when includeFallback is true AND no