duckdb 1.0.1-dev22.0 → 1.0.1-dev27.0

package/src/duckdb/third_party/brotli/enc/backward_references.cpp

@@ -0,0 +1,3775 @@
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* Function to find backward reference copies. */
+
+#include "backward_references.h"
+
+#include <brotli/types.h>
+
+#include "../common/brotli_constants.h"
+#include "../common/dictionary.h"
+#include "../common/brotli_platform.h"
+#include "command.h"
+#include "compound_dictionary.h"
+#include "dictionary_hash.h"
+#include "encoder_dict.h"
+#include "memory.h"
+#include "quality.h"
+
+using namespace duckdb_brotli;
+
+static BROTLI_INLINE size_t ComputeDistanceCode(size_t distance,
+                                                size_t max_distance,
+                                                const int* dist_cache) {
+  if (distance <= max_distance) {
+    size_t distance_plus_3 = distance + 3;
+    size_t offset0 = distance_plus_3 - (size_t)dist_cache[0];
+    size_t offset1 = distance_plus_3 - (size_t)dist_cache[1];
+    if (distance == (size_t)dist_cache[0]) {
+      return 0;
+    } else if (distance == (size_t)dist_cache[1]) {
+      return 1;
+    } else if (offset0 < 7) {
+      return (0x9750468 >> (4 * offset0)) & 0xF;
+    } else if (offset1 < 7) {
+      return (0xFDB1ACE >> (4 * offset1)) & 0xF;
+    } else if (distance == (size_t)dist_cache[2]) {
+      return 2;
+    } else if (distance == (size_t)dist_cache[3]) {
+      return 3;
+    }
+  }
+  return distance + BROTLI_NUM_DISTANCE_SHORT_CODES - 1;
+}
+
+#define EXPAND_CAT(a, b) CAT(a, b)
+#define CAT(a, b) a ## b
+#define FN(X) EXPAND_CAT(X, HASHER())
+#define EXPORT_FN(X) EXPAND_CAT(X, EXPAND_CAT(PREFIX(), HASHER()))
+
+#define PREFIX() N
+#define ENABLE_COMPOUND_DICTIONARY 0
+
+#define HASHER() H2
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H3
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H4
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H5
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H6
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H40
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H41
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H42
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H54
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H35
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H55
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#define HASHER() H65
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#undef ENABLE_COMPOUND_DICTIONARY
+#undef PREFIX
+#define PREFIX() D
+#define ENABLE_COMPOUND_DICTIONARY 1
+
+#define HASHER() H5
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+#define HASHER() H6
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+#define HASHER() H40
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+#define HASHER() H41
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+#define HASHER() H42
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+#define HASHER() H55
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+#define HASHER() H65
+/* NOLINTNEXTLINE(build/include) */
+/* NOLINT(build/header_guard) */
+/* Copyright 2013 Google Inc. All Rights Reserved.
+
+   Distributed under MIT license.
+   See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
+*/
+
+/* template parameters: EXPORT_FN, FN */
+
+static BROTLI_NOINLINE void EXPORT_FN(CreateBackwardReferences)(
+    size_t num_bytes, size_t position,
+    const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  HASHER()* privat = &hasher->privat.FN(_);
+  /* Set maximum distance, see section 9.1. of the spec. */
+  const size_t max_backward_limit = BROTLI_MAX_BACKWARD_LIMIT(params->lgwin);
+  const size_t position_offset = params->stream_offset;
+
+  const Command* const orig_commands = commands;
+  size_t insert_length = *last_insert_len;
+  const size_t pos_end = position + num_bytes;
+  const size_t store_end = num_bytes >= FN(StoreLookahead)() ?
+      position + num_bytes - FN(StoreLookahead)() + 1 : position;
+
+  /* For speed up heuristics for random data. */
+  const size_t random_heuristics_window_size =
+      LiteralSpreeLengthForSparseSearch(params);
+  size_t apply_random_heuristics = position + random_heuristics_window_size;
+  const size_t gap = params->dictionary.compound.total_size;
+
+  /* Minimum score to accept a backward reference. */
+  const score_t kMinScore = BROTLI_SCORE_BASE + 100;
+
+  FN(PrepareDistanceCache)(privat, dist_cache);
+
+  while (position + FN(HashTypeLength)() < pos_end) {
+    size_t max_length = pos_end - position;
+    size_t max_distance = BROTLI_MIN(size_t, position, max_backward_limit);
+    size_t dictionary_start = BROTLI_MIN(size_t,
+        position + position_offset, max_backward_limit);
+    HasherSearchResult sr;
+    int dict_id = 0;
+    uint8_t p1 = 0;
+    uint8_t p2 = 0;
+    if (params->dictionary.contextual.context_based) {
+      p1 = position >= 1 ?
+          ringbuffer[(size_t)(position - 1) & ringbuffer_mask] : 0;
+      p2 = position >= 2 ?
+          ringbuffer[(size_t)(position - 2) & ringbuffer_mask] : 0;
+      dict_id = params->dictionary.contextual.context_map[
+          BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+    }
+    sr.len = 0;
+    sr.len_code_delta = 0;
+    sr.distance = 0;
+    sr.score = kMinScore;
+    FN(FindLongestMatch)(privat, params->dictionary.contextual.dict[dict_id],
+        ringbuffer, ringbuffer_mask, dist_cache, position, max_length,
+        max_distance, dictionary_start + gap, params->dist.max_distance, &sr);
+    if (ENABLE_COMPOUND_DICTIONARY) {
+      LookupCompoundDictionaryMatch(&params->dictionary.compound, ringbuffer,
+          ringbuffer_mask, dist_cache, position, max_length,
+          dictionary_start, params->dist.max_distance, &sr);
+    }
+    if (sr.score > kMinScore) {
+      /* Found a match. Let's look for something even better ahead. */
+      int delayed_backward_references_in_row = 0;
+      --max_length;
+      for (;; --max_length) {
+        const score_t cost_diff_lazy = 175;
+        HasherSearchResult sr2;
+        sr2.len = params->quality < MIN_QUALITY_FOR_EXTENSIVE_REFERENCE_SEARCH ?
+            BROTLI_MIN(size_t, sr.len - 1, max_length) : 0;
+        sr2.len_code_delta = 0;
+        sr2.distance = 0;
+        sr2.score = kMinScore;
+        max_distance = BROTLI_MIN(size_t, position + 1, max_backward_limit);
+        dictionary_start = BROTLI_MIN(size_t,
+            position + 1 + position_offset, max_backward_limit);
+        if (params->dictionary.contextual.context_based) {
+          p2 = p1;
+          p1 = ringbuffer[position & ringbuffer_mask];
+          dict_id = params->dictionary.contextual.context_map[
+              BROTLI_CONTEXT(p1, p2, literal_context_lut)];
+        }
+        FN(FindLongestMatch)(privat,
+            params->dictionary.contextual.dict[dict_id],
+            ringbuffer, ringbuffer_mask, dist_cache, position + 1, max_length,
+            max_distance, dictionary_start + gap, params->dist.max_distance,
+            &sr2);
+        if (ENABLE_COMPOUND_DICTIONARY) {
+          LookupCompoundDictionaryMatch(
+              &params->dictionary.compound, ringbuffer,
+              ringbuffer_mask, dist_cache, position + 1, max_length,
+              dictionary_start, params->dist.max_distance, &sr2);
+        }
+        if (sr2.score >= sr.score + cost_diff_lazy) {
+          /* Ok, let's just write one byte for now and start a match from the
+             next byte. */
+          ++position;
+          ++insert_length;
+          sr = sr2;
+          if (++delayed_backward_references_in_row < 4 &&
+              position + FN(HashTypeLength)() < pos_end) {
+            continue;
+          }
+        }
+        break;
+      }
+      apply_random_heuristics =
+          position + 2 * sr.len + random_heuristics_window_size;
+      dictionary_start = BROTLI_MIN(size_t,
+          position + position_offset, max_backward_limit);
+      {
+        /* The first 16 codes are special short-codes,
+           and the minimum offset is 1. */
+        size_t distance_code = ComputeDistanceCode(
+            sr.distance, dictionary_start + gap, dist_cache);
+        if ((sr.distance <= (dictionary_start + gap)) && distance_code > 0) {
+          dist_cache[3] = dist_cache[2];
+          dist_cache[2] = dist_cache[1];
+          dist_cache[1] = dist_cache[0];
+          dist_cache[0] = (int)sr.distance;
+          FN(PrepareDistanceCache)(privat, dist_cache);
+        }
+        InitCommand(commands++, &params->dist, insert_length,
+            sr.len, sr.len_code_delta, distance_code);
+      }
+      *num_literals += insert_length;
+      insert_length = 0;
+      /* Put the hash keys into the table, if there are enough bytes left.
+         Depending on the hasher implementation, it can push all positions
+         in the given range or only a subset of them.
+         Avoid hash poisoning with RLE data. */
+      {
+        size_t range_start = position + 2;
+        size_t range_end = BROTLI_MIN(size_t, position + sr.len, store_end);
+        if (sr.distance < (sr.len >> 2)) {
+          range_start = BROTLI_MIN(size_t, range_end, BROTLI_MAX(size_t,
+              range_start, position + sr.len - (sr.distance << 2)));
+        }
+        FN(StoreRange)(privat, ringbuffer, ringbuffer_mask, range_start,
+                       range_end);
+      }
+      position += sr.len;
+    } else {
+      ++insert_length;
+      ++position;
+      /* If we have not seen matches for a long time, we can skip some
+         match lookups. Unsuccessful match lookups are very very expensive
+         and this kind of a heuristic speeds up compression quite
+         a lot. */
+      if (position > apply_random_heuristics) {
+        /* Going through uncompressible data, jump. */
+        if (position >
+            apply_random_heuristics + 4 * random_heuristics_window_size) {
+          /* It is quite a long time since we saw a copy, so we assume
+             that this data is not compressible, and store hashes less
+             often. Hashes of non compressible data are less likely to
+             turn out to be useful in the future, too, so we store less of
+             them to not to flood out the hash table of good compressible
+             data. */
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 4);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 16, pos_end - kMargin);
+          for (; position < pos_jump; position += 4) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 4;
+          }
+        } else {
+          const size_t kMargin =
+              BROTLI_MAX(size_t, FN(StoreLookahead)() - 1, 2);
+          size_t pos_jump =
+              BROTLI_MIN(size_t, position + 8, pos_end - kMargin);
+          for (; position < pos_jump; position += 2) {
+            FN(Store)(privat, ringbuffer, ringbuffer_mask, position);
+            insert_length += 2;
+          }
+        }
+      }
+    }
+  }
+  insert_length += pos_end - position;
+  *last_insert_len = insert_length;
+  *num_commands += (size_t)(commands - orig_commands);
+}
+#undef HASHER
+
+#undef ENABLE_COMPOUND_DICTIONARY
+#undef PREFIX
+
+#undef EXPORT_FN
+#undef FN
+#undef CAT
+#undef EXPAND_CAT
+
+void duckdb_brotli::BrotliCreateBackwardReferences(size_t num_bytes,
+    size_t position, const uint8_t* ringbuffer, size_t ringbuffer_mask,
+    ContextLut literal_context_lut, const BrotliEncoderParams* params,
+    Hasher* hasher, int* dist_cache, size_t* last_insert_len,
+    Command* commands, size_t* num_commands, size_t* num_literals) {
+  if (params->dictionary.compound.num_chunks != 0) {
+    switch (params->hasher.type) {
+#define CASE_(N)                                                    \
+      case N:                                                       \
+        CreateBackwardReferencesDH ## N(num_bytes,                  \
+            position, ringbuffer, ringbuffer_mask,                  \
+            literal_context_lut, params, hasher, dist_cache,        \
+            last_insert_len, commands, num_commands, num_literals); \
+        return;
+      CASE_(5)
+      CASE_(6)
+      CASE_(40)
+      CASE_(41)
+      CASE_(42)
+      CASE_(55)
+      CASE_(65)
+#undef CASE_
+      default:
+        BROTLI_DCHECK(false);
+        break;
+    }
+  }
+
+  switch (params->hasher.type) {
+#define CASE_(N)                                                  \
+    case N:                                                       \
+      CreateBackwardReferencesNH ## N(num_bytes,                  \
+          position, ringbuffer, ringbuffer_mask,                  \
+          literal_context_lut, params, hasher, dist_cache,        \
+          last_insert_len, commands, num_commands, num_literals); \
+      return;
+    FOR_GENERIC_HASHERS(CASE_)
+#undef CASE_
+    default:
+      BROTLI_DCHECK(false);
+      break;
+  }
+}
+
+