ketr.photos/scanner/scanner.c
James Ketrenos c234abbcc2 Update migration
Signed-off-by: James Ketrenos <james_git@ketrenos.com>
2020-01-05 21:50:48 -08:00

199 lines
4.9 KiB
C

#include <stdio.h>
#include <stdlib.h>
#include <dirent.h>
#include <sys/types.h>
#include <string.h>
#include <math.h>
#include <sqlite3.h>
typedef struct Face {
long double descriptor[128];
int faceId;
long double *distances;
struct Face *next;
struct Face *prev;
} Face;
char fileBuf[5000];
char pathBuf[1028];
Face *readFaceDescriptor(int id, char *path) {
FILE *f;
Face *pFace = (Face *)malloc(sizeof(Face));
memset(pFace, 0, sizeof(Face));
f = fopen(path, "r");
if (!f) {
free(pFace);
return NULL;
}
size_t s = fread(fileBuf, 1, sizeof(fileBuf), f);
fclose(f);
char *p = fileBuf;
fileBuf[s] = 0;
while (*p && *p != '-' && *p != '+' && (*p < '0' || *p > '9')) {
p++;
}
for (int i = 0; i < 128; i++) {
char *start = p;
while (*p && *p != ',' && *p != ']' && *p != ' ' && *p != '\n') {
p++;
}
if (!*p) {
break;
}
*p++ = 0;
sscanf(start, "%Lf", &pFace->descriptor[i]);
}
pFace->faceId = id;
pFace->next = pFace->prev = NULL;
return pFace;
}
long double euclideanDistance(long double *a, long double *b) {
long double sum = 0.0;
for (int i = 0; i < 128; i++) {
long double delta = a[i] - b[i];
sum += delta * delta;
}
return sqrtl(sum);
}
int main(int argc, char *argv[]) {
int maxId = 0;
int len = 0;
int i;
Face *pChain = NULL;
for (i = 0; i < 100; i++) {
sprintf(pathBuf, "%s/face-data/%d", argv[1], i);
DIR *faceDir = opendir(pathBuf);
fprintf(stderr, "Reading %s...\n", pathBuf);
if (!faceDir) {
fprintf(stderr, "Can not open %s\n", pathBuf);
continue;
}
struct dirent *ent;
while ((ent = readdir(faceDir)) != NULL) {
if (strstr(ent->d_name, ".json") == NULL) {
continue;
}
int id = 0;
char *p = ent->d_name;
while (*p && *p != '-') {
id *= 10;
id += *p - '0';
p++;
}
char path[1028*2];
sprintf(path, "%s/%s", pathBuf, ent->d_name);
maxId = maxId > id ? maxId : id;
Face *pFace = readFaceDescriptor(id, path);
if (!pFace) {
continue;
}
len++;
if (len % 1000 == 0) {
fprintf(stderr, "...read %d...\n", len);
}
if (pChain) {
pFace->next = pChain;
}
pChain = pFace;
}
closedir(faceDir);
}
fprintf(stderr, "Read %d face descriptors...\n", len);
/* Allocate storage for all distances */
Face *pLink = pChain;
while (pLink) {
pLink->distances = (long double *)malloc(sizeof(long double) * len);
pLink = pLink->next;
}
sqlite3 *db;
int rc = sqlite3_open("db/photos.db", &db);
if (rc != SQLITE_OK) {
fprintf(stderr, "Cannot open database: %s\n", sqlite3_errmsg(db));
sqlite3_close(db);
return 1;
}
fprintf(stderr, "DB opened.");
char *err_msg = NULL;
char *sql =
"DELETE FROM facedistances;"
"BEGIN TRANSACTION;";
rc = sqlite3_exec(db, sql, 0, 0, &err_msg);
if (rc != SQLITE_OK ) {
fprintf(stderr, "SQL error: %s\n", err_msg);
sqlite3_free(err_msg);
sqlite3_close(db);
return 1;
}
fprintf(stderr, "facedistances deleted and transaction started.\n");
char sqlBuf[1024];
pLink = pChain;
int sourceIndex = 0, lines = 0;
while (pLink) {
int targetIndex = 0;
Face *pTarget = pChain;
while (pTarget) {
if (targetIndex == sourceIndex) {
pLink->distances[targetIndex] = 0.0;
pTarget->distances[sourceIndex] = 0.0;
} else {
if (pLink->distances[targetIndex] == 0.0) {
pLink->distances[targetIndex] =
pTarget->distances[sourceIndex] = euclideanDistance(pLink->descriptor, pTarget->descriptor);
if (pLink->distances[targetIndex] < 0.5) {
sprintf(sqlBuf, "INSERT INTO facedistances (face1Id,face2Id,distance) VALUES (%d,%d,%Lf);",
((pLink->faceId < pTarget->faceId) ? pLink->faceId : pTarget->faceId),
((pLink->faceId < pTarget->faceId) ? pTarget->faceId : pLink->faceId),
pLink->distances[targetIndex]);
rc = sqlite3_exec(db, sqlBuf, 0, 0, &err_msg);
if (rc != SQLITE_OK ) {
fprintf(stderr, "SQL error: %s\n", err_msg);
sqlite3_free(err_msg);
sqlite3_close(db);
return 1;
}
lines++;
if (lines % 1000 == 0) {
fprintf(stderr, "...output %d DB lines (%0.2f complete)...\n", lines, (float)(1. * sourceIndex / (1. * len)));
}
}
}
}
pTarget = pTarget->next;
targetIndex++;
}
pLink = pLink->next;
sourceIndex++;
}
sprintf(sqlBuf, "UPDATE faces SET lastComparedId=%d;", maxId);
rc = sqlite3_exec(db, "COMMIT;", 0, 0, &err_msg);
if (rc != SQLITE_OK ) {
fprintf(stderr, "SQL error: %s\n", err_msg);
sqlite3_free(err_msg);
sqlite3_close(db);
return 1;
}
sqlite3_close(db);
return 0;
}