summaryrefslogtreecommitdiffstats
path: root/gfx/harfbuzz/src/hb-bit-set.hh
diff options
context:
space:
mode:
Diffstat (limited to 'gfx/harfbuzz/src/hb-bit-set.hh')
-rw-r--r--gfx/harfbuzz/src/hb-bit-set.hh982
1 files changed, 982 insertions, 0 deletions
diff --git a/gfx/harfbuzz/src/hb-bit-set.hh b/gfx/harfbuzz/src/hb-bit-set.hh
new file mode 100644
index 0000000000..1dbcce5cbd
--- /dev/null
+++ b/gfx/harfbuzz/src/hb-bit-set.hh
@@ -0,0 +1,982 @@
+/*
+ * Copyright © 2012,2017 Google, Inc.
+ * Copyright © 2021 Behdad Esfahbod
+ *
+ * This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_BIT_SET_HH
+#define HB_BIT_SET_HH
+
+#include "hb.hh"
+#include "hb-bit-page.hh"
+
+
+struct hb_bit_set_t
+{
+ hb_bit_set_t () = default;
+ ~hb_bit_set_t () = default;
+
+ hb_bit_set_t (const hb_bit_set_t& other) : hb_bit_set_t () { set (other, true); }
+ hb_bit_set_t ( hb_bit_set_t&& other) : hb_bit_set_t () { hb_swap (*this, other); }
+ hb_bit_set_t& operator= (const hb_bit_set_t& other) { set (other); return *this; }
+ hb_bit_set_t& operator= (hb_bit_set_t&& other) { hb_swap (*this, other); return *this; }
+ friend void swap (hb_bit_set_t &a, hb_bit_set_t &b)
+ {
+ if (likely (!a.successful || !b.successful))
+ return;
+ hb_swap (a.population, b.population);
+ hb_swap (a.last_page_lookup, b.last_page_lookup);
+ hb_swap (a.page_map, b.page_map);
+ hb_swap (a.pages, b.pages);
+ }
+
+ void init ()
+ {
+ successful = true;
+ population = 0;
+ last_page_lookup = 0;
+ page_map.init ();
+ pages.init ();
+ }
+ void fini ()
+ {
+ page_map.fini ();
+ pages.fini ();
+ }
+
+ using page_t = hb_bit_page_t;
+ struct page_map_t
+ {
+ int cmp (const page_map_t &o) const { return cmp (o.major); }
+ int cmp (uint32_t o_major) const { return (int) o_major - (int) major; }
+
+ uint32_t major;
+ uint32_t index;
+ };
+
+ bool successful = true; /* Allocations successful */
+ mutable unsigned int population = 0;
+ mutable hb_atomic_int_t last_page_lookup = 0;
+ hb_sorted_vector_t<page_map_t> page_map;
+ hb_vector_t<page_t> pages;
+
+ void err () { if (successful) successful = false; } /* TODO Remove */
+ bool in_error () const { return !successful; }
+
+ bool resize (unsigned int count, bool clear = true, bool exact_size = false)
+ {
+ if (unlikely (!successful)) return false;
+
+ if (pages.length == 0 && count == 1)
+ exact_size = true; // Most sets are small and local
+
+ if (unlikely (!pages.resize (count, clear, exact_size) || !page_map.resize (count, clear, exact_size)))
+ {
+ pages.resize (page_map.length, clear, exact_size);
+ successful = false;
+ return false;
+ }
+ return true;
+ }
+
+ void alloc (unsigned sz)
+ {
+ sz >>= (page_t::PAGE_BITS_LOG_2 - 1);
+ pages.alloc (sz);
+ page_map.alloc (sz);
+ }
+
+ void reset ()
+ {
+ successful = true;
+ clear ();
+ }
+
+ void clear ()
+ {
+ resize (0);
+ if (likely (successful))
+ population = 0;
+ }
+ bool is_empty () const
+ {
+ unsigned int count = pages.length;
+ for (unsigned int i = 0; i < count; i++)
+ if (!pages[i].is_empty ())
+ return false;
+ return true;
+ }
+ explicit operator bool () const { return !is_empty (); }
+
+ uint32_t hash () const
+ {
+ uint32_t h = 0;
+ for (auto &map : page_map)
+ {
+ auto &page = pages.arrayZ[map.index];
+ if (unlikely (page.is_empty ())) continue;
+ h = h * 31 + hb_hash (map.major) + hb_hash (page);
+ }
+ return h;
+ }
+
+ private:
+ void dirty () { population = UINT_MAX; }
+ public:
+
+ void add (hb_codepoint_t g)
+ {
+ if (unlikely (!successful)) return;
+ if (unlikely (g == INVALID)) return;
+ dirty ();
+ page_t *page = page_for (g, true); if (unlikely (!page)) return;
+ page->add (g);
+ }
+ bool add_range (hb_codepoint_t a, hb_codepoint_t b)
+ {
+ if (unlikely (!successful)) return true; /* https://github.com/harfbuzz/harfbuzz/issues/657 */
+ if (unlikely (a > b || a == INVALID || b == INVALID)) return false;
+ dirty ();
+ unsigned int ma = get_major (a);
+ unsigned int mb = get_major (b);
+ if (ma == mb)
+ {
+ page_t *page = page_for (a, true); if (unlikely (!page)) return false;
+ page->add_range (a, b);
+ }
+ else
+ {
+ page_t *page = page_for (a, true); if (unlikely (!page)) return false;
+ page->add_range (a, major_start (ma + 1) - 1);
+
+ for (unsigned int m = ma + 1; m < mb; m++)
+ {
+ page = page_for (major_start (m), true); if (unlikely (!page)) return false;
+ page->init1 ();
+ }
+
+ page = page_for (b, true); if (unlikely (!page)) return false;
+ page->add_range (major_start (mb), b);
+ }
+ return true;
+ }
+
+ /* Duplicated here from hb-machinery.hh to avoid including it. */
+ template<typename Type>
+ static inline const Type& StructAtOffsetUnaligned(const void *P, unsigned int offset)
+ {
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-align"
+ return * reinterpret_cast<const Type*> ((const char *) P + offset);
+#pragma GCC diagnostic pop
+ }
+
+ template <typename T>
+ void set_array (bool v, const T *array, unsigned int count, unsigned int stride=sizeof(T))
+ {
+ if (unlikely (!successful)) return;
+ if (!count) return;
+ dirty ();
+ hb_codepoint_t g = *array;
+ while (count)
+ {
+ unsigned int m = get_major (g);
+ page_t *page = page_for (g, v); if (unlikely (v && !page)) return;
+ unsigned int start = major_start (m);
+ unsigned int end = major_start (m + 1);
+ do
+ {
+ if (g != INVALID && (v || page)) /* The v check is to optimize out the page check if v is true. */
+ page->set (g, v);
+
+ array = &StructAtOffsetUnaligned<T> (array, stride);
+ count--;
+ }
+ while (count && (g = *array, start <= g && g < end));
+ }
+ }
+
+ template <typename T>
+ void add_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+ { set_array (true, array, count, stride); }
+ template <typename T>
+ void add_array (const hb_array_t<const T>& arr) { add_array (&arr, arr.len ()); }
+
+ template <typename T>
+ void del_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+ { set_array (false, array, count, stride); }
+ template <typename T>
+ void del_array (const hb_array_t<const T>& arr) { del_array (&arr, arr.len ()); }
+
+ /* Might return false if array looks unsorted.
+ * Used for faster rejection of corrupt data. */
+ template <typename T>
+ bool set_sorted_array (bool v, const T *array, unsigned int count, unsigned int stride=sizeof(T))
+ {
+ if (unlikely (!successful)) return true; /* https://github.com/harfbuzz/harfbuzz/issues/657 */
+ if (unlikely (!count)) return true;
+ dirty ();
+ hb_codepoint_t g = *array;
+ hb_codepoint_t last_g = g;
+ while (count)
+ {
+ unsigned int m = get_major (g);
+ page_t *page = page_for (g, v); if (unlikely (v && !page)) return false;
+ unsigned int end = major_start (m + 1);
+ do
+ {
+ /* If we try harder we can change the following comparison to <=;
+ * Not sure if it's worth it. */
+ if (g < last_g) return false;
+ last_g = g;
+
+ if (g != INVALID && (v || page)) /* The v check is to optimize out the page check if v is true. */
+ page->add (g);
+
+ array = &StructAtOffsetUnaligned<T> (array, stride);
+ count--;
+ }
+ while (count && (g = *array, g < end));
+ }
+ return true;
+ }
+
+ template <typename T>
+ bool add_sorted_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+ { return set_sorted_array (true, array, count, stride); }
+ template <typename T>
+ bool add_sorted_array (const hb_sorted_array_t<const T>& arr) { return add_sorted_array (&arr, arr.len ()); }
+
+ template <typename T>
+ bool del_sorted_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+ { return set_sorted_array (false, array, count, stride); }
+ template <typename T>
+ bool del_sorted_array (const hb_sorted_array_t<const T>& arr) { return del_sorted_array (&arr, arr.len ()); }
+
+ void del (hb_codepoint_t g)
+ {
+ if (unlikely (!successful)) return;
+ page_t *page = page_for (g);
+ if (!page)
+ return;
+ dirty ();
+ page->del (g);
+ }
+
+ private:
+ void del_pages (int ds, int de)
+ {
+ if (ds <= de)
+ {
+ // Pre-allocate the workspace that compact() will need so we can bail on allocation failure
+ // before attempting to rewrite the page map.
+ hb_vector_t<unsigned> compact_workspace;
+ if (unlikely (!allocate_compact_workspace (compact_workspace))) return;
+
+ unsigned int write_index = 0;
+ for (unsigned int i = 0; i < page_map.length; i++)
+ {
+ int m = (int) page_map[i].major;
+ if (m < ds || de < m)
+ page_map[write_index++] = page_map[i];
+ }
+ compact (compact_workspace, write_index);
+ resize (write_index);
+ }
+ }
+
+
+ public:
+ void del_range (hb_codepoint_t a, hb_codepoint_t b)
+ {
+ if (unlikely (!successful)) return;
+ if (unlikely (a > b || a == INVALID)) return;
+ dirty ();
+ unsigned int ma = get_major (a);
+ unsigned int mb = get_major (b);
+ /* Delete pages from ds through de if ds <= de. */
+ int ds = (a == major_start (ma))? (int) ma: (int) (ma + 1);
+ int de = (b + 1 == major_start (mb + 1))? (int) mb: ((int) mb - 1);
+ if (ds > de || (int) ma < ds)
+ {
+ page_t *page = page_for (a);
+ if (page)
+ {
+ if (ma == mb)
+ page->del_range (a, b);
+ else
+ page->del_range (a, major_start (ma + 1) - 1);
+ }
+ }
+ if (de < (int) mb && ma != mb)
+ {
+ page_t *page = page_for (b);
+ if (page)
+ page->del_range (major_start (mb), b);
+ }
+ del_pages (ds, de);
+ }
+
+ bool get (hb_codepoint_t g) const
+ {
+ const page_t *page = page_for (g);
+ if (!page)
+ return false;
+ return page->get (g);
+ }
+
+ /* Has interface. */
+ bool operator [] (hb_codepoint_t k) const { return get (k); }
+ bool has (hb_codepoint_t k) const { return (*this)[k]; }
+ /* Predicate. */
+ bool operator () (hb_codepoint_t k) const { return has (k); }
+
+ /* Sink interface. */
+ hb_bit_set_t& operator << (hb_codepoint_t v)
+ { add (v); return *this; }
+ hb_bit_set_t& operator << (const hb_codepoint_pair_t& range)
+ { add_range (range.first, range.second); return *this; }
+
+ bool intersects (hb_codepoint_t first, hb_codepoint_t last) const
+ {
+ hb_codepoint_t c = first - 1;
+ return next (&c) && c <= last;
+ }
+ void set (const hb_bit_set_t &other, bool exact_size = false)
+ {
+ if (unlikely (!successful)) return;
+ unsigned int count = other.pages.length;
+ if (unlikely (!resize (count, false, exact_size)))
+ return;
+ population = other.population;
+
+ page_map = other.page_map;
+ pages = other.pages;
+ }
+
+ bool is_equal (const hb_bit_set_t &other) const
+ {
+ if (has_population () && other.has_population () &&
+ population != other.population)
+ return false;
+
+ unsigned int na = pages.length;
+ unsigned int nb = other.pages.length;
+
+ unsigned int a = 0, b = 0;
+ for (; a < na && b < nb; )
+ {
+ if (page_at (a).is_empty ()) { a++; continue; }
+ if (other.page_at (b).is_empty ()) { b++; continue; }
+ if (page_map[a].major != other.page_map[b].major ||
+ !page_at (a).is_equal (other.page_at (b)))
+ return false;
+ a++;
+ b++;
+ }
+ for (; a < na; a++)
+ if (!page_at (a).is_empty ()) { return false; }
+ for (; b < nb; b++)
+ if (!other.page_at (b).is_empty ()) { return false; }
+
+ return true;
+ }
+
+ bool is_subset (const hb_bit_set_t &larger_set) const
+ {
+ if (has_population () && larger_set.has_population () &&
+ population > larger_set.population)
+ return false;
+
+ uint32_t spi = 0;
+ for (uint32_t lpi = 0; spi < page_map.length && lpi < larger_set.page_map.length; lpi++)
+ {
+ uint32_t spm = page_map[spi].major;
+ uint32_t lpm = larger_set.page_map[lpi].major;
+ auto sp = page_at (spi);
+
+ if (spm < lpm && !sp.is_empty ())
+ return false;
+
+ if (lpm < spm)
+ continue;
+
+ auto lp = larger_set.page_at (lpi);
+ if (!sp.is_subset (lp))
+ return false;
+
+ spi++;
+ }
+
+ while (spi < page_map.length)
+ if (!page_at (spi++).is_empty ())
+ return false;
+
+ return true;
+ }
+
+ private:
+ bool allocate_compact_workspace (hb_vector_t<unsigned>& workspace)
+ {
+ if (unlikely (!workspace.resize_exact (pages.length)))
+ {
+ successful = false;
+ return false;
+ }
+
+ return true;
+ }
+
+ /*
+ * workspace should be a pre-sized vector allocated to hold at exactly pages.length
+ * elements.
+ */
+ void compact (hb_vector_t<unsigned>& workspace,
+ unsigned int length)
+ {
+ assert(workspace.length == pages.length);
+ hb_vector_t<unsigned>& old_index_to_page_map_index = workspace;
+
+ hb_fill (old_index_to_page_map_index.writer(), 0xFFFFFFFF);
+ for (unsigned i = 0; i < length; i++)
+ old_index_to_page_map_index[page_map[i].index] = i;
+
+ compact_pages (old_index_to_page_map_index);
+ }
+ void compact_pages (const hb_vector_t<unsigned>& old_index_to_page_map_index)
+ {
+ unsigned int write_index = 0;
+ for (unsigned int i = 0; i < pages.length; i++)
+ {
+ if (old_index_to_page_map_index[i] == 0xFFFFFFFF) continue;
+
+ if (write_index < i)
+ pages[write_index] = pages[i];
+
+ page_map[old_index_to_page_map_index[i]].index = write_index;
+ write_index++;
+ }
+ }
+ public:
+
+ void process_ (hb_bit_page_t::vector_t (*op) (const hb_bit_page_t::vector_t &, const hb_bit_page_t::vector_t &),
+ bool passthru_left, bool passthru_right,
+ const hb_bit_set_t &other)
+ {
+ if (unlikely (!successful)) return;
+
+ dirty ();
+
+ unsigned int na = pages.length;
+ unsigned int nb = other.pages.length;
+ unsigned int next_page = na;
+
+ unsigned int count = 0, newCount = 0;
+ unsigned int a = 0, b = 0;
+ unsigned int write_index = 0;
+
+ // Pre-allocate the workspace that compact() will need so we can bail on allocation failure
+ // before attempting to rewrite the page map.
+ hb_vector_t<unsigned> compact_workspace;
+ if (!passthru_left && unlikely (!allocate_compact_workspace (compact_workspace))) return;
+
+ for (; a < na && b < nb; )
+ {
+ if (page_map[a].major == other.page_map[b].major)
+ {
+ if (!passthru_left)
+ {
+ // Move page_map entries that we're keeping from the left side set
+ // to the front of the page_map vector. This isn't necessary if
+ // passthru_left is set since no left side pages will be removed
+ // in that case.
+ if (write_index < a)
+ page_map[write_index] = page_map[a];
+ write_index++;
+ }
+
+ count++;
+ a++;
+ b++;
+ }
+ else if (page_map[a].major < other.page_map[b].major)
+ {
+ if (passthru_left)
+ count++;
+ a++;
+ }
+ else
+ {
+ if (passthru_right)
+ count++;
+ b++;
+ }
+ }
+ if (passthru_left)
+ count += na - a;
+ if (passthru_right)
+ count += nb - b;
+
+ if (!passthru_left)
+ {
+ na = write_index;
+ next_page = write_index;
+ compact (compact_workspace, write_index);
+ }
+
+ if (unlikely (!resize (count)))
+ return;
+
+ newCount = count;
+
+ /* Process in-place backward. */
+ a = na;
+ b = nb;
+ for (; a && b; )
+ {
+ if (page_map.arrayZ[a - 1].major == other.page_map.arrayZ[b - 1].major)
+ {
+ a--;
+ b--;
+ count--;
+ page_map.arrayZ[count] = page_map.arrayZ[a];
+ page_at (count).v = op (page_at (a).v, other.page_at (b).v);
+ page_at (count).dirty ();
+ }
+ else if (page_map.arrayZ[a - 1].major > other.page_map.arrayZ[b - 1].major)
+ {
+ a--;
+ if (passthru_left)
+ {
+ count--;
+ page_map.arrayZ[count] = page_map.arrayZ[a];
+ }
+ }
+ else
+ {
+ b--;
+ if (passthru_right)
+ {
+ count--;
+ page_map.arrayZ[count].major = other.page_map.arrayZ[b].major;
+ page_map.arrayZ[count].index = next_page++;
+ page_at (count) = other.page_at (b);
+ }
+ }
+ }
+ if (passthru_left)
+ while (a)
+ {
+ a--;
+ count--;
+ page_map.arrayZ[count] = page_map.arrayZ[a];
+ }
+ if (passthru_right)
+ while (b)
+ {
+ b--;
+ count--;
+ page_map.arrayZ[count].major = other.page_map.arrayZ[b].major;
+ page_map.arrayZ[count].index = next_page++;
+ page_at (count) = other.page_at (b);
+ }
+ assert (!count);
+ resize (newCount);
+ }
+ template <typename Op>
+ static hb_bit_page_t::vector_t
+ op_ (const hb_bit_page_t::vector_t &a, const hb_bit_page_t::vector_t &b)
+ { return Op{} (a, b); }
+ template <typename Op>
+ void process (const Op& op, const hb_bit_set_t &other)
+ {
+ process_ (op_<Op>, op (1, 0), op (0, 1), other);
+ }
+
+ void union_ (const hb_bit_set_t &other) { process (hb_bitwise_or, other); }
+ void intersect (const hb_bit_set_t &other) { process (hb_bitwise_and, other); }
+ void subtract (const hb_bit_set_t &other) { process (hb_bitwise_gt, other); }
+ void symmetric_difference (const hb_bit_set_t &other) { process (hb_bitwise_xor, other); }
+
+ bool next (hb_codepoint_t *codepoint) const
+ {
+ if (unlikely (*codepoint == INVALID)) {
+ *codepoint = get_min ();
+ return *codepoint != INVALID;
+ }
+
+ const auto* page_map_array = page_map.arrayZ;
+ unsigned int major = get_major (*codepoint);
+ unsigned int i = last_page_lookup;
+
+ if (unlikely (i >= page_map.length || page_map_array[i].major != major))
+ {
+ page_map.bfind (major, &i, HB_NOT_FOUND_STORE_CLOSEST);
+ if (i >= page_map.length) {
+ *codepoint = INVALID;
+ return false;
+ }
+ last_page_lookup = i;
+ }
+
+ const auto* pages_array = pages.arrayZ;
+ const page_map_t &current = page_map_array[i];
+ if (likely (current.major == major))
+ {
+ if (pages_array[current.index].next (codepoint))
+ {
+ *codepoint += current.major * page_t::PAGE_BITS;
+ return true;
+ }
+ i++;
+ }
+
+ for (; i < page_map.length; i++)
+ {
+ const page_map_t &current = page_map_array[i];
+ hb_codepoint_t m = pages_array[current.index].get_min ();
+ if (m != INVALID)
+ {
+ *codepoint = current.major * page_t::PAGE_BITS + m;
+ last_page_lookup = i;
+ return true;
+ }
+ }
+ *codepoint = INVALID;
+ return false;
+ }
+ bool previous (hb_codepoint_t *codepoint) const
+ {
+ if (unlikely (*codepoint == INVALID)) {
+ *codepoint = get_max ();
+ return *codepoint != INVALID;
+ }
+
+ page_map_t map = {get_major (*codepoint), 0};
+ unsigned int i;
+ page_map.bfind (map, &i, HB_NOT_FOUND_STORE_CLOSEST);
+ if (i < page_map.length && page_map.arrayZ[i].major == map.major)
+ {
+ if (pages[page_map.arrayZ[i].index].previous (codepoint))
+ {
+ *codepoint += page_map.arrayZ[i].major * page_t::PAGE_BITS;
+ return true;
+ }
+ }
+ i--;
+ for (; (int) i >= 0; i--)
+ {
+ hb_codepoint_t m = pages.arrayZ[page_map.arrayZ[i].index].get_max ();
+ if (m != INVALID)
+ {
+ *codepoint = page_map.arrayZ[i].major * page_t::PAGE_BITS + m;
+ return true;
+ }
+ }
+ *codepoint = INVALID;
+ return false;
+ }
+ bool next_range (hb_codepoint_t *first, hb_codepoint_t *last) const
+ {
+ hb_codepoint_t i;
+
+ i = *last;
+ if (!next (&i))
+ {
+ *last = *first = INVALID;
+ return false;
+ }
+
+ /* TODO Speed up. */
+ *last = *first = i;
+ while (next (&i) && i == *last + 1)
+ (*last)++;
+
+ return true;
+ }
+ bool previous_range (hb_codepoint_t *first, hb_codepoint_t *last) const
+ {
+ hb_codepoint_t i;
+
+ i = *first;
+ if (!previous (&i))
+ {
+ *last = *first = INVALID;
+ return false;
+ }
+
+ /* TODO Speed up. */
+ *last = *first = i;
+ while (previous (&i) && i == *first - 1)
+ (*first)--;
+
+ return true;
+ }
+
+ unsigned int next_many (hb_codepoint_t codepoint,
+ hb_codepoint_t *out,
+ unsigned int size) const
+ {
+ // By default, start at the first bit of the first page of values.
+ unsigned int start_page = 0;
+ unsigned int start_page_value = 0;
+ if (unlikely (codepoint != INVALID))
+ {
+ const auto* page_map_array = page_map.arrayZ;
+ unsigned int major = get_major (codepoint);
+ unsigned int i = last_page_lookup;
+ if (unlikely (i >= page_map.length || page_map_array[i].major != major))
+ {
+ page_map.bfind (major, &i, HB_NOT_FOUND_STORE_CLOSEST);
+ if (i >= page_map.length)
+ return 0; // codepoint is greater than our max element.
+ }
+ start_page = i;
+ start_page_value = page_remainder (codepoint + 1);
+ if (unlikely (start_page_value == 0))
+ {
+ // The export-after value was last in the page. Start on next page.
+ start_page++;
+ start_page_value = 0;
+ }
+ }
+
+ unsigned int initial_size = size;
+ for (unsigned int i = start_page; i < page_map.length && size; i++)
+ {
+ uint32_t base = major_start (page_map[i].major);
+ unsigned int n = pages[page_map[i].index].write (base, start_page_value, out, size);
+ out += n;
+ size -= n;
+ start_page_value = 0;
+ }
+ return initial_size - size;
+ }
+
+ unsigned int next_many_inverted (hb_codepoint_t codepoint,
+ hb_codepoint_t *out,
+ unsigned int size) const
+ {
+ unsigned int initial_size = size;
+ // By default, start at the first bit of the first page of values.
+ unsigned int start_page = 0;
+ unsigned int start_page_value = 0;
+ if (unlikely (codepoint != INVALID))
+ {
+ const auto* page_map_array = page_map.arrayZ;
+ unsigned int major = get_major (codepoint);
+ unsigned int i = last_page_lookup;
+ if (unlikely (i >= page_map.length || page_map_array[i].major != major))
+ {
+ page_map.bfind(major, &i, HB_NOT_FOUND_STORE_CLOSEST);
+ if (unlikely (i >= page_map.length))
+ {
+ // codepoint is greater than our max element.
+ while (++codepoint != INVALID && size)
+ {
+ *out++ = codepoint;
+ size--;
+ }
+ return initial_size - size;
+ }
+ }
+ start_page = i;
+ start_page_value = page_remainder (codepoint + 1);
+ if (unlikely (start_page_value == 0))
+ {
+ // The export-after value was last in the page. Start on next page.
+ start_page++;
+ start_page_value = 0;
+ }
+ }
+
+ hb_codepoint_t next_value = codepoint + 1;
+ for (unsigned int i=start_page; i<page_map.length && size; i++)
+ {
+ uint32_t base = major_start (page_map[i].major);
+ unsigned int n = pages[page_map[i].index].write_inverted (base, start_page_value, out, size, &next_value);
+ out += n;
+ size -= n;
+ start_page_value = 0;
+ }
+ while (next_value < HB_SET_VALUE_INVALID && size) {
+ *out++ = next_value++;
+ size--;
+ }
+ return initial_size - size;
+ }
+
+ bool has_population () const { return population != UINT_MAX; }
+ unsigned int get_population () const
+ {
+ if (has_population ())
+ return population;
+
+ unsigned int pop = 0;
+ unsigned int count = pages.length;
+ for (unsigned int i = 0; i < count; i++)
+ pop += pages[i].get_population ();
+
+ population = pop;
+ return pop;
+ }
+ hb_codepoint_t get_min () const
+ {
+ unsigned count = pages.length;
+ for (unsigned i = 0; i < count; i++)
+ {
+ const auto& map = page_map[i];
+ const auto& page = pages[map.index];
+
+ if (!page.is_empty ())
+ return map.major * page_t::PAGE_BITS + page.get_min ();
+ }
+ return INVALID;
+ }
+ hb_codepoint_t get_max () const
+ {
+ unsigned count = pages.length;
+ for (signed i = count - 1; i >= 0; i--)
+ {
+ const auto& map = page_map[(unsigned) i];
+ const auto& page = pages[map.index];
+
+ if (!page.is_empty ())
+ return map.major * page_t::PAGE_BITS + page.get_max ();
+ }
+ return INVALID;
+ }
+
+ static constexpr hb_codepoint_t INVALID = page_t::INVALID;
+
+ /*
+ * Iterator implementation.
+ */
+ struct iter_t : hb_iter_with_fallback_t<iter_t, hb_codepoint_t>
+ {
+ static constexpr bool is_sorted_iterator = true;
+ static constexpr bool has_fast_len = true;
+ iter_t (const hb_bit_set_t &s_ = Null (hb_bit_set_t),
+ bool init = true) : s (&s_), v (INVALID), l(0)
+ {
+ if (init)
+ {
+ l = s->get_population () + 1;
+ __next__ ();
+ }
+ }
+
+ typedef hb_codepoint_t __item_t__;
+ hb_codepoint_t __item__ () const { return v; }
+ bool __more__ () const { return v != INVALID; }
+ void __next__ () { s->next (&v); if (l) l--; }
+ void __prev__ () { s->previous (&v); }
+ unsigned __len__ () const { return l; }
+ iter_t end () const { return iter_t (*s, false); }
+ bool operator != (const iter_t& o) const
+ { return s != o.s || v != o.v; }
+
+ protected:
+ const hb_bit_set_t *s;
+ hb_codepoint_t v;
+ unsigned l;
+ };
+ iter_t iter () const { return iter_t (*this); }
+ operator iter_t () const { return iter (); }
+
+ protected:
+
+ page_t *page_for (hb_codepoint_t g, bool insert = false)
+ {
+ unsigned major = get_major (g);
+
+ /* The extra page_map length is necessary; can't just rely on vector here,
+ * since the next check would be tricked because a null page also has
+ * major==0, which we can't distinguish from an actually major==0 page... */
+ unsigned i = last_page_lookup;
+ if (likely (i < page_map.length))
+ {
+ auto &cached_page = page_map.arrayZ[i];
+ if (cached_page.major == major)
+ return &pages.arrayZ[cached_page.index];
+ }
+
+ page_map_t map = {major, pages.length};
+ if (!page_map.bfind (map, &i, HB_NOT_FOUND_STORE_CLOSEST))
+ {
+ if (!insert)
+ return nullptr;
+
+ if (unlikely (!resize (pages.length + 1)))
+ return nullptr;
+
+ pages.arrayZ[map.index].init0 ();
+ memmove (page_map.arrayZ + i + 1,
+ page_map.arrayZ + i,
+ (page_map.length - 1 - i) * page_map.item_size);
+ page_map.arrayZ[i] = map;
+ }
+
+ last_page_lookup = i;
+ return &pages.arrayZ[page_map.arrayZ[i].index];
+ }
+ const page_t *page_for (hb_codepoint_t g) const
+ {
+ unsigned major = get_major (g);
+
+ /* The extra page_map length is necessary; can't just rely on vector here,
+ * since the next check would be tricked because a null page also has
+ * major==0, which we can't distinguish from an actually major==0 page... */
+ unsigned i = last_page_lookup;
+ if (likely (i < page_map.length))
+ {
+ auto &cached_page = page_map.arrayZ[i];
+ if (cached_page.major == major)
+ return &pages.arrayZ[cached_page.index];
+ }
+
+ page_map_t key = {major};
+ if (!page_map.bfind (key, &i))
+ return nullptr;
+
+ last_page_lookup = i;
+ return &pages.arrayZ[page_map[i].index];
+ }
+ page_t &page_at (unsigned int i)
+ {
+ assert (i < page_map.length);
+ return pages.arrayZ[page_map.arrayZ[i].index];
+ }
+ const page_t &page_at (unsigned int i) const
+ {
+ assert (i < page_map.length);
+ return pages.arrayZ[page_map.arrayZ[i].index];
+ }
+ unsigned int get_major (hb_codepoint_t g) const { return g >> page_t::PAGE_BITS_LOG_2; }
+ unsigned int page_remainder (hb_codepoint_t g) const { return g & page_t::PAGE_BITMASK; }
+ hb_codepoint_t major_start (unsigned int major) const { return major << page_t::PAGE_BITS_LOG_2; }
+};
+
+
+#endif /* HB_BIT_SET_HH */