#include "db.h"
-#include "vp4.h"
+#include "turbopfor-encode.h"
#include <algorithm>
-#include <arpa/inet.h>
#include <assert.h>
#include <chrono>
-#include <endian.h>
-#include <fcntl.h>
+#include <iosfwd>
#include <math.h>
#include <memory>
+#include <stdint.h>
#include <stdio.h>
+#include <stdlib.h>
#include <string.h>
#include <string>
+#include <string_view>
#include <sys/stat.h>
-#include <sys/types.h>
-#include <unistd.h>
-#include <unordered_map>
+#include <utility>
#include <vector>
#include <zstd.h>
void write_header(uint32_t docid);
void append_block();
- vector<uint32_t> pending_docids;
+ vector<uint32_t> pending_deltas;
uint32_t last_block_end, last_docid = -1;
};
return;
}
+ pending_deltas.push_back(docid - last_docid - 1);
last_docid = docid;
- pending_docids.push_back(docid);
- if (pending_docids.size() == 128) {
+ if (pending_deltas.size() == 128) {
append_block();
- pending_docids.clear();
+ pending_deltas.clear();
last_block_end = docid;
}
++num_docids;
void PostingListBuilder::finish()
{
- if (pending_docids.empty()) {
+ if (pending_deltas.empty()) {
return;
}
// No interleaving for partial blocks.
unsigned char buf[P4NENC_BOUND(128)];
- unsigned char *end = p4d1enc32(pending_docids.data(), pending_docids.size(), buf, last_block_end);
+ unsigned char *end = encode_pfor_single_block<128>(pending_deltas.data(), pending_deltas.size(), /*interleaved=*/false, buf);
encoded.append(reinterpret_cast<char *>(buf), reinterpret_cast<char *>(end));
}
void PostingListBuilder::append_block()
{
unsigned char buf[P4NENC_BOUND(128)];
- assert(pending_docids.size() == 128);
- unsigned char *end = p4d1enc128v32(pending_docids.data(), 128, buf, last_block_end);
+ assert(pending_deltas.size() == 128);
+ unsigned char *end = encode_pfor_single_block<128>(pending_deltas.data(), 128, /*interleaved=*/true, buf);
encoded.append(reinterpret_cast<char *>(buf), reinterpret_cast<char *>(end));
}
void PostingListBuilder::write_header(uint32_t docid)
{
unsigned char buf[P4NENC_BOUND(1)];
- size_t bytes = p4nd1enc128v32(&docid, 1, buf);
- encoded.append(reinterpret_cast<char *>(buf), bytes);
+ unsigned char *end = write_baseval(docid, buf);
+ encoded.append(reinterpret_cast<char *>(buf), end - buf);
}
class Corpus {
public:
Corpus(FILE *outfp, size_t block_size)
- : invindex(new PostingListBuilder *[NUM_TRIGRAMS]), outfp(outfp), block_size(block_size) {}
+ : invindex(new PostingListBuilder *[NUM_TRIGRAMS]), outfp(outfp), block_size(block_size)
+ {
+ fill(invindex.get(), invindex.get() + NUM_TRIGRAMS, nullptr);
+ }
+ ~Corpus()
+ {
+ for (unsigned i = 0; i < NUM_TRIGRAMS; ++i) {
+ delete invindex[i];
+ }
+ }
+
void add_file(string filename);
void flush_block();
hdr.version = -1; // Mark as broken.
hdr.hashtable_size = 0; // Not known yet.
hdr.extra_ht_slots = num_overflow_slots;
+ hdr.num_docids = 0;
hdr.hash_table_offset_bytes = -1; // We don't know these offsets yet.
hdr.filename_index_offset_bytes = -1;
fwrite(&hdr, sizeof(hdr), 1, outfp);