Docker/mealie
Docker/mealie
1
0
Fork 0
mirror of https://github.com/mealie-recipes/mealie.git synced 2025-08-05 13:35:23 +02:00
Code Issues Releases Activity

feat: Remove OCR Support (#2838)

Browse source 
* remove ocr package

* remove tesseract

* remove OCR from app

* remove OCR from tests

* fix docs
This commit is contained in:
Michael Genson 2023-12-14 20:26:43 -06:00 • committed by GitHub
parent c48680374d
commit ca9f66ee24
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
34 changed files with 29 additions and 1570 deletions
Download patch file Download diff file Expand all files Collapse all files

BIN
tests/data/images/test-ocr.png
View file

Binary file not shown.
Side by side

Before

Width:  |  Height:  |  Size: 11 KiB

73
tests/data/text/test-ocr.tsv
View file

@ -1,73 +0,0 @@
level page_num block_num par_num line_num word_num left top width height conf text
1 1 0 0 0 0 0 0 640 480 -1
2 1 1 0 0 0 36 92 582 269 -1
3 1 1 1 0 0 36 92 582 92 -1
4 1 1 1 1 0 36 92 544 30 -1
5 1 1 1 1 1 36 92 60 24 87.137558 This
5 1 1 1 1 2 109 92 20 24 87.137558 is
5 1 1 1 1 3 141 98 15 18 87.823906 a
5 1 1 1 1 4 169 92 32 24 87.823906 lot
5 1 1 1 1 5 212 92 28 24 92.965874 of
5 1 1 1 1 6 251 92 31 24 93.247513 12
5 1 1 1 1 7 296 92 68 30 92.734741 point
5 1 1 1 1 8 374 93 53 23 92.996040 text
5 1 1 1 1 9 437 93 26 23 93.160057 to
5 1 1 1 1 10 474 93 52 23 92.312637 test
5 1 1 1 1 11 536 92 44 24 92.312637 the
4 1 1 1 2 0 36 126 582 31 -1
5 1 1 1 2 1 36 132 45 18 90.505524 ocr
5 1 1 1 2 2 91 126 69 24 90.505524 code
5 1 1 1 2 3 172 126 51 24 91.169167 and
5 1 1 1 2 4 236 132 50 18 89.765854 see
5 1 1 1 2 5 299 126 15 24 85.827324 if
5 1 1 1 2 6 325 126 14 24 93.116241 it
5 1 1 1 2 7 348 126 85 24 92.394562 works
5 1 1 1 2 8 445 132 33 18 30.119690 on
5 1 1 1 2 9 500 126 29 24 30.119690 all
5 1 1 1 2 10 541 127 77 30 92.090988 types
4 1 1 1 3 0 36 160 187 24 -1
5 1 1 1 3 1 36 160 28 24 92.476135 of
5 1 1 1 3 2 72 160 41 24 90.919365 file
5 1 1 1 3 3 123 160 100 24 91.360367 format.
3 1 1 2 0 0 36 194 561 167 -1
4 1 1 2 1 0 36 194 549 31 -1
5 1 1 2 1 1 36 194 55 24 89.098892 The
5 1 1 2 1 2 102 194 75 30 89.098892 quick
5 1 1 2 1 3 189 194 85 24 91.415680 brown
5 1 1 2 1 4 287 194 52 31 91.943085 dog
5 1 1 2 1 5 348 194 108 31 92.167969 jumped
5 1 1 2 1 6 468 200 63 18 91.970985 over
5 1 1 2 1 7 540 194 45 24 92.843704 the
4 1 1 2 2 0 37 228 548 31 -1
5 1 1 2 2 1 37 228 55 31 92.262550 lazy
5 1 1 2 2 2 103 228 50 24 92.693161 fox.
5 1 1 2 2 3 165 228 55 24 92.947639 The
5 1 1 2 2 4 232 228 75 30 90.589806 quick
5 1 1 2 2 5 319 228 85 24 91.051247 brown
5 1 1 2 2 6 417 228 51 31 91.925011 dog
5 1 1 2 2 7 478 228 107 31 91.471077 jumped
4 1 1 2 3 0 36 262 561 31 -1
5 1 1 2 3 1 36 268 63 18 90.210129 over
5 1 1 2 3 2 109 262 44 24 90.210129 the
5 1 1 2 3 3 165 262 56 31 91.178192 lazy
5 1 1 2 3 4 231 262 50 24 92.794647 fox.
5 1 1 2 3 5 294 262 55 24 91.388016 The
5 1 1 2 3 6 360 262 75 30 92.525742 quick
5 1 1 2 3 7 447 262 85 24 90.425552 brown
5 1 1 2 3 8 545 262 52 31 90.425552 dog
4 1 1 2 4 0 43 296 518 31 -1
5 1 1 2 4 1 43 296 107 31 91.759590 jumped
5 1 1 2 4 2 162 302 64 18 92.923576 over
5 1 1 2 4 3 235 296 44 24 92.017929 the
5 1 1 2 4 4 292 296 55 31 91.558884 lazy
5 1 1 2 4 5 357 296 50 24 92.687485 fox.
5 1 1 2 4 6 420 296 55 24 91.922661 The
5 1 1 2 4 7 486 296 75 30 91.870224 quick
4 1 1 2 5 0 37 330 524 31 -1
5 1 1 2 5 1 37 330 85 24 92.923935 brown
5 1 1 2 5 2 135 330 52 31 91.468765 dog
5 1 1 2 5 3 196 330 108 31 91.425491 jumped
5 1 1 2 5 4 316 336 63 18 91.489830 over
5 1 1 2 5 5 388 330 45 24 91.740379 the
5 1 1 2 5 6 445 330 55 31 92.110054 lazy
5 1 1 2 5 7 511 330 50 24 93.180054 fox.
1 level page_num block_num par_num line_num word_num left top width height conf text
2 1 1 0 0 0 0 0 0 640 480 -1
3 2 1 1 0 0 0 36 92 582 269 -1
4 3 1 1 1 0 0 36 92 582 92 -1
5 4 1 1 1 1 0 36 92 544 30 -1
6 5 1 1 1 1 1 36 92 60 24 87.137558 This
7 5 1 1 1 1 2 109 92 20 24 87.137558 is
8 5 1 1 1 1 3 141 98 15 18 87.823906 a
9 5 1 1 1 1 4 169 92 32 24 87.823906 lot
10 5 1 1 1 1 5 212 92 28 24 92.965874 of
11 5 1 1 1 1 6 251 92 31 24 93.247513 12
12 5 1 1 1 1 7 296 92 68 30 92.734741 point
13 5 1 1 1 1 8 374 93 53 23 92.996040 text
14 5 1 1 1 1 9 437 93 26 23 93.160057 to
15 5 1 1 1 1 10 474 93 52 23 92.312637 test
16 5 1 1 1 1 11 536 92 44 24 92.312637 the
17 4 1 1 1 2 0 36 126 582 31 -1
18 5 1 1 1 2 1 36 132 45 18 90.505524 ocr
19 5 1 1 1 2 2 91 126 69 24 90.505524 code
20 5 1 1 1 2 3 172 126 51 24 91.169167 and
21 5 1 1 1 2 4 236 132 50 18 89.765854 see
22 5 1 1 1 2 5 299 126 15 24 85.827324 if
23 5 1 1 1 2 6 325 126 14 24 93.116241 it
24 5 1 1 1 2 7 348 126 85 24 92.394562 works
25 5 1 1 1 2 8 445 132 33 18 30.119690 on
26 5 1 1 1 2 9 500 126 29 24 30.119690 all
27 5 1 1 1 2 10 541 127 77 30 92.090988 types
28 4 1 1 1 3 0 36 160 187 24 -1
29 5 1 1 1 3 1 36 160 28 24 92.476135 of
30 5 1 1 1 3 2 72 160 41 24 90.919365 file
31 5 1 1 1 3 3 123 160 100 24 91.360367 format.
32 3 1 1 2 0 0 36 194 561 167 -1
33 4 1 1 2 1 0 36 194 549 31 -1
34 5 1 1 2 1 1 36 194 55 24 89.098892 The
35 5 1 1 2 1 2 102 194 75 30 89.098892 quick
36 5 1 1 2 1 3 189 194 85 24 91.415680 brown
37 5 1 1 2 1 4 287 194 52 31 91.943085 dog
38 5 1 1 2 1 5 348 194 108 31 92.167969 jumped
39 5 1 1 2 1 6 468 200 63 18 91.970985 over
40 5 1 1 2 1 7 540 194 45 24 92.843704 the
41 4 1 1 2 2 0 37 228 548 31 -1
42 5 1 1 2 2 1 37 228 55 31 92.262550 lazy
43 5 1 1 2 2 2 103 228 50 24 92.693161 fox.
44 5 1 1 2 2 3 165 228 55 24 92.947639 The
45 5 1 1 2 2 4 232 228 75 30 90.589806 quick
46 5 1 1 2 2 5 319 228 85 24 91.051247 brown
47 5 1 1 2 2 6 417 228 51 31 91.925011 dog
48 5 1 1 2 2 7 478 228 107 31 91.471077 jumped
49 4 1 1 2 3 0 36 262 561 31 -1
50 5 1 1 2 3 1 36 268 63 18 90.210129 over
51 5 1 1 2 3 2 109 262 44 24 90.210129 the
52 5 1 1 2 3 3 165 262 56 31 91.178192 lazy
53 5 1 1 2 3 4 231 262 50 24 92.794647 fox.
54 5 1 1 2 3 5 294 262 55 24 91.388016 The
55 5 1 1 2 3 6 360 262 75 30 92.525742 quick
56 5 1 1 2 3 7 447 262 85 24 90.425552 brown
57 5 1 1 2 3 8 545 262 52 31 90.425552 dog
58 4 1 1 2 4 0 43 296 518 31 -1
59 5 1 1 2 4 1 43 296 107 31 91.759590 jumped
60 5 1 1 2 4 2 162 302 64 18 92.923576 over
61 5 1 1 2 4 3 235 296 44 24 92.017929 the
62 5 1 1 2 4 4 292 296 55 31 91.558884 lazy
63 5 1 1 2 4 5 357 296 50 24 92.687485 fox.
64 5 1 1 2 4 6 420 296 55 24 91.922661 The
65 5 1 1 2 4 7 486 296 75 30 91.870224 quick
66 4 1 1 2 5 0 37 330 524 31 -1
67 5 1 1 2 5 1 37 330 85 24 92.923935 brown
68 5 1 1 2 5 2 135 330 52 31 91.468765 dog
69 5 1 1 2 5 3 196 330 108 31 91.425491 jumped
70 5 1 1 2 5 4 316 336 63 18 91.489830 over
71 5 1 1 2 5 5 388 330 45 24 91.740379 the
72 5 1 1 2 5 6 445 330 55 31 92.110054 lazy
73 5 1 1 2 5 7 511 330 50 24 93.180054 fox.

9
tests/data/text/test-ocr.txt
View file

@ -1,9 +0,0 @@
This is a lot of 12 point text to test the
ocr code and see if it works on all types
of file format.
The quick brown dog jumped over the
lazy fox. The quick brown dog jumped
over the lazy fox. The quick brown dog
jumped over the lazy fox. The quick
brown dog jumped over the lazy fox.

58
tests/unit_tests/services_tests/test_ocr_service.py
View file

@ -1,58 +0,0 @@
from pathlib import Path
import pytest
from mealie.services.ocr.pytesseract import OcrService
ocr_service = OcrService()
@pytest.mark.skip("Tesseract is not reliable between environments")
def test_image_to_string():
with open(Path("tests/data/images/test-ocr.png"), "rb") as image:
result = ocr_service.image_to_string(image)
with open(Path("tests/data/text/test-ocr.txt"), encoding="utf-8") as expected_result:
assert result == expected_result.read()
@pytest.mark.skip("Tesseract is not reliable between environments")
def test_image_to_tsv():
with open(Path("tests/data/images/test-ocr.png"), "rb") as image:
result = ocr_service.image_to_tsv(image.read())
with open(Path("tests/data/text/test-ocr.tsv"), encoding="utf-8") as expected_result:
assert result == expected_result.read()
def test_format_tsv_output():
tsv = " level\tpage_num\tblock_num\tpar_num\tline_num\tword_num\tleft\ttop\twidth\theight\tconf\ttext \n1\t1\t0\t0\t0\t0\t0\t0\t640\t480\t-1\t\n5\t1\t1\t1\t1\t1\t36\t92\t60\t24\t87.137558\tThis"
expected_result = [
{
"level": 1,
"page_num": 1,
"block_num": 0,
"par_num": 0,
"line_num": 0,
"word_num": 0,
"left": 0,
"top": 0,
"width": 640,
"height": 480,
"conf": -1.0,
"text": "",
},
{
"level": 5,
"page_num": 1,
"block_num": 1,
"par_num": 1,
"line_num": 1,
"word_num": 1,
"left": 36,
"top": 92,
"width": 60,
"height": 24,
"conf": 87.137558,
"text": "This",
},
]
assert ocr_service.format_tsv_output(tsv) == expected_result

8
tests/utils/api_routes/__init__.py
View file

@ -113,12 +113,6 @@ groups_webhooks_rerun = "/api/groups/webhooks/rerun"
"""`/api/groups/webhooks/rerun`"""
media_docker_validate_txt = "/api/media/docker/validate.txt"
"""`/api/media/docker/validate.txt`"""
ocr = "/api/ocr/"
"""`/api/ocr/`"""
ocr_asset_to_tsv = "/api/ocr/asset-to-tsv"
"""`/api/ocr/asset-to-tsv`"""
ocr_file_to_tsv = "/api/ocr/file-to-tsv"
"""`/api/ocr/file-to-tsv`"""
organizers_categories = "/api/organizers/categories"
"""`/api/organizers/categories`"""
organizers_categories_empty = "/api/organizers/categories/empty"
@ -151,8 +145,6 @@ recipes_bulk_actions_tag = "/api/recipes/bulk-actions/tag"
"""`/api/recipes/bulk-actions/tag`"""
recipes_create_from_zip = "/api/recipes/create-from-zip"
"""`/api/recipes/create-from-zip`"""
recipes_create_ocr = "/api/recipes/create-ocr"
"""`/api/recipes/create-ocr`"""
recipes_create_url = "/api/recipes/create-url"
"""`/api/recipes/create-url`"""
recipes_create_url_bulk = "/api/recipes/create-url/bulk"