Optimize KConfigIniBackend::parseConfig by reducing allocations.

Yet another awesome application of the Qt implicit sharing trick.
Since config files often contain only few different keys and even
value strings, we can share them. This reduces memory consumption
and also speeds up parsing, as we do not have to allocate the
duplicated strings, but can simply reuse the previous values.

The most extreme case for this of my knowledge, is KatePart:
katesyntaxhighlightingrc has more than 20k lines which triggered
nearly 30k allocations on startup. With this patch applied, this
value goes down dramatically. I added a simple static counter for
the cache hit/miss ratio, which resulted in 5442 cache misses compared
to 43624 cache hits across all KConfig files parsed by kwrite.

REVIEW: 118587

This is a forward-port of b8aaeff128233cfaecf67899168887572589dde8.

1 files changed, 19 insertions, 2 deletions

diff --git a/src/core/kconfigini.cpp b/src/core/kconfigini.cpp
index a882ee31..f4ce9410 100644
--- a/src/core/kconfigini.cpp
+++ b/src/core/kconfigini.cpp
@@ -44,6 +44,15 @@
 
 KCONFIGCORE_EXPORT bool kde_kiosk_exception = false; // flag to disable kiosk restrictions
 
+static QByteArray lookup(const KConfigIniBackend::BufferFragment &fragment, QHash<KConfigIniBackend::BufferFragment, QByteArray> *cache)
+{
+    QHash<KConfigIniBackend::BufferFragment, QByteArray>::iterator it = cache->find(fragment);
+    if (it == cache->end()) {
+        it = cache->insert(fragment, fragment.toByteArray());
+    }
+    return it.value();
+}
+
 QString KConfigIniBackend::warningProlog(const QFile &file, int line)
 {
     return QString::fromLatin1("KConfigIni: In file %2, line %1: ")
@@ -100,6 +109,14 @@ KConfigIniBackend::parseConfig(const QByteArray &currentLocale, KEntryMap &entry
     unsigned int len = contents.length();
     unsigned int startOfLine = 0;
 
+    // Reduce memory overhead by making use of implicit sharing
+    // This assumes that config files contain only a small amount of
+    // different fragments which are repeated often.
+    // This is often the case, especially sub groups will all have
+    // the same list of keys and similar values as well.
+    QHash<BufferFragment, QByteArray> cache;
+    cache.reserve(4096);
+
     while (startOfLine < len) {
         BufferFragment line = contents.split('\n', &startOfLine);
         line.trim();
@@ -265,9 +282,9 @@ KConfigIniBackend::parseConfig(const QByteArray &currentLocale, KEntryMap &entry
                 rawKey.reserve(aKey.length() + locale.length() + 2);
                 rawKey.append(aKey.toVolatileByteArray());
                 rawKey.append('[').append(locale.toVolatileByteArray()).append(']');
-                entryMap.setEntry(currentGroup, rawKey, line.toByteArray(), entryOptions);
+                entryMap.setEntry(currentGroup, rawKey, lookup(line, &cache), entryOptions);
             } else {
-                entryMap.setEntry(currentGroup, aKey.toByteArray(), line.toByteArray(), entryOptions);
+                entryMap.setEntry(currentGroup, lookup(aKey, &cache), lookup(line, &cache), entryOptions);
             }
         }
     next_line: