1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
|
import logging
import os
import sqlite3
import imagehash
import magic
import matplotlib.pyplot as plt
from PIL import Image
from pybktree import BKTree
logging.basicConfig(level=logging.INFO)
DB_FILENAME = os.path.expanduser("~/Documents/pixiv.image.hash.db")
DB_NAME = "hash"
CREATE_SQL = """CREATE TABLE hash (
id INTEGER NOT NULL PRIMARY KEY AUTOINCREMENT,
filename TEXT NOT NULL,
hash TEXT NOT NULL
)
"""
IMG_DIR = "/mnt/hdd43/imgs/pixiv"
PREVIEW_DIR = os.path.expanduser("~/Pictures/dups")
file_magic = magic.Magic(mime=True)
sqlite3_conn = sqlite3.connect(DB_FILENAME)
def ensure_table() -> None:
cur = sqlite3_conn.cursor()
if (
cur.execute(
"SELECT name FROM sqlite_master WHERE name = ?", (DB_NAME,)
).fetchone()
is not None
):
logging.info(f"table {DB_NAME} exists.")
return
cur.execute(CREATE_SQL)
logging.info(f"create table {DB_NAME}.")
def is_image(filename: str) -> bool:
return file_magic.from_file(filename=filename).startswith("image")
def in_db(filename: str) -> bool:
cur = sqlite3_conn.cursor()
if (
cur.execute("SELECT id FROM hash WHERE filename = ?", (filename,)).fetchone()
is None
):
return False
return True
def cal_hash_recursively(directory: str = ".") -> None:
for dirpath, _, filenames in os.walk(directory):
for filename in filenames:
f = os.path.normpath(os.path.join(dirpath, filename))
logging.info(f"processing {f}...")
if not is_image(f):
logging.info(f"{f}: not an image, skipping ...")
continue
if in_db(f):
logging.info(f"{f}: already in db, skipping ...")
continue
cur = sqlite3_conn.cursor()
cur.execute(
"INSERT INTO hash VALUES (?, ?, ?)",
(
None,
f,
str(imagehash.phash(Image.open(f), hash_size=16)),
),
)
logging.info(f"{f}: hash saved")
sqlite3_conn.commit()
def img_hamming(a: tuple, b: tuple):
return a[1] - b[1]
def construct_bktree() -> BKTree:
cur = sqlite3_conn.cursor()
cur.execute("SELECT filename, hash FROM hash")
row = cur.fetchone()
hashes = []
while row is not None:
hashes.append((row[0], imagehash.hex_to_hash(row[1])))
row = cur.fetchone()
return BKTree(img_hamming, hashes)
def find_dups(tree: BKTree) -> list[list]:
cur = sqlite3_conn.cursor()
cur.execute("SELECT filename, hash FROM hash")
row = cur.fetchone()
dups = []
while row is not None:
dup = [i[1][0] for i in tree.find((row[0], imagehash.hex_to_hash(row[1])), 25)]
if len(dup) > 1:
dups.append(dup)
row = cur.fetchone()
dups_sorted = [list(t) for t in set(tuple(sorted(l)) for l in dups)]
return dups_sorted
def generate_dup_previews(dups: list[list]) -> None:
for i, dup in enumerate(dups):
images = dup
titles = [os.path.basename(img) for img in images]
logging.info(f"processing dup #{i}: {images}")
fig, axes = plt.subplots(
nrows=1, ncols=len(titles), figsize=(3 * len(titles), 3)
)
axes = axes.flatten()
for j, ax in enumerate(axes):
with Image.open(images[j]) as img:
img.thumbnail((1024, 1024))
ax.imshow(img)
ax.set_title(titles[j], fontsize=8)
ax.axis("off")
fig.tight_layout()
fig.savefig(os.path.join(PREVIEW_DIR, f"dup.{i}.png"), dpi=300)
plt.close(fig)
ensure_table()
cal_hash_recursively(IMG_DIR)
tree = construct_bktree()
dups = find_dups(tree)
generate_dup_previews(dups)
|
