walkero
/
harfbuzz
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[repacker] cleanup.

This commit is contained in:
Garret Rieger 2022-07-29 18:12:49 +00:00
parent 674f0194a3
commit 2a5902ee50
1 changed files with 30 additions and 34 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

64
src/hb-repacker.hh
View File

@ -46,39 +46,39 @@ struct lookup_size_t
unsigned lookup_index; unsigned lookup_index;
size_t size; size_t size;
unsigned num_subtables; unsigned num_subtables;
};
inline int compare_sizes (const void* a, const void* b)
{
lookup_size_t* size_a = (lookup_size_t*) a;
lookup_size_t* size_b = (lookup_size_t*) b;
double subtables_per_byte_a = (double) size_a->num_subtables / (double) size_a->size;
double subtables_per_byte_b = (double) size_b->num_subtables / (double) size_b->size;
if (subtables_per_byte_a == subtables_per_byte_b) {
return size_b->lookup_index - size_a->lookup_index;
static int cmp (const void* a, const void* b)
{
return cmp ((const lookup_size_t*) a,
(const lookup_size_t*) b);
} }
double cmp = subtables_per_byte_b - subtables_per_byte_a;
if (cmp < 0) return -1; static int cmp (const lookup_size_t* a, const lookup_size_t* b)
if (cmp > 0) return 1; {
return 0; double subtables_per_byte_a = (double) a->num_subtables / (double) a->size;
} double subtables_per_byte_b = (double) b->num_subtables / (double) b->size;
if (subtables_per_byte_a == subtables_per_byte_b) {
return b->lookup_index - a->lookup_index;
}
double cmp = subtables_per_byte_b - subtables_per_byte_a;
if (cmp < 0) return -1;
if (cmp > 0) return 1;
return 0;
}
};
static inline static inline
bool _presplit_subtables_if_needed (graph::gsubgpos_graph_context_t& ext_context) bool _presplit_subtables_if_needed (graph::gsubgpos_graph_context_t& ext_context)
{ {
// Algorithm: // For each lookup this will check the size of subtables and split them as needed
// 1. Scan all sub-tables and compute their sizes. // so that no subtable is at risk of overflowing. (where we support splitting for
// 2. For any PairPos subtables that are >64kb: // that subtable type).
// a. split into two more pieces that are <64kb //
// b. De-dup subgraphs (optional) (can this be done during serialization?) // TODO(grieger): de-dup newly added nodes as necessary. Probably just want a full de-dup
// pass after this processing is done. Not super necessary as splits are
// TODO: Add a split subtables method at the lookup level, it could scan it's subtables and split as // only done where overflow is likely, so de-dup probably will get undone
// needed. // later anyways.
// TODO: rename gsubgpos_graph_context_t to be more general.
for (unsigned lookup_index : ext_context.lookups.keys ()) for (unsigned lookup_index : ext_context.lookups.keys ())
{ {
graph::Lookup* lookup = ext_context.lookups.get(lookup_index); graph::Lookup* lookup = ext_context.lookups.get(lookup_index);
@ -98,24 +98,20 @@ bool _promote_extensions_if_needed (graph::gsubgpos_graph_context_t& ext_context
{ {
// Simple Algorithm (v1, current): // Simple Algorithm (v1, current):
// 1. Calculate how many bytes each non-extension lookup consumes. // 1. Calculate how many bytes each non-extension lookup consumes.
// 2. Select up to 64k of those to remain as non-extension (greedy, smallest first). // 2. Select up to 64k of those to remain as non-extension (greedy, highest subtables per byte first)
// 3. Promote the rest. // 3. Promote the rest.
// //
// Advanced Algorithm (v2, not implemented): // Advanced Algorithm (v2, not implemented):
// 1. Perform connected component analysis using lookups as roots. // 1. Perform connected component analysis using lookups as roots.
// 2. Compute size of each connected component. // 2. Compute size of each connected component.
// 3. Select up to 64k worth of connected components to remain as non-extensions. // 3. Select up to 64k worth of connected components to remain as non-extensions.
// (greedy, smallest first) // (greedy, highest subtables per byte first)
// 4. Promote the rest. // 4. Promote the rest.
// TODO(garretrieger): support extension demotion, then consider all lookups. Requires advanced algo. // TODO(garretrieger): support extension demotion, then consider all lookups. Requires advanced algo.
// TODO(garretrieger): also support extension promotion during iterative resolution phase, then // TODO(garretrieger): also support extension promotion during iterative resolution phase, then
// we can use a less conservative threshold here. // we can use a less conservative threshold here.
// TODO(grieger): skip this for the 24 bit case. // TODO(grieger): skip this for the 24 bit case.
// TODO(grieger): sort by # subtables / size instead (high to low). Goal is to get as many subtables
// as possible into space 0 to minimize the number of extension subtables added.
// A fully optimal solution will require a backpack problem dynamic programming type
// solution.
if (!ext_context.lookups) return true; if (!ext_context.lookups) return true;
hb_vector_t<lookup_size_t> lookup_sizes; hb_vector_t<lookup_size_t> lookup_sizes;
@ -132,7 +128,7 @@ bool _promote_extensions_if_needed (graph::gsubgpos_graph_context_t& ext_context
}); });
} }
lookup_sizes.qsort (compare_sizes); lookup_sizes.qsort ();
size_t lookup_list_size = ext_context.graph.vertices_[ext_context.lookup_list_index].table_size (); size_t lookup_list_size = ext_context.graph.vertices_[ext_context.lookup_list_index].table_size ();
size_t l2_l3_size = lookup_list_size; // Lookup List + Lookups size_t l2_l3_size = lookup_list_size; // Lookup List + Lookups