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
|
test-extraction-data
====================
Each test for the extractor is defined by a .expected file.
.expected file format
=====================
It is a JSON file with two toplevel key: "test" and "metadata".
The "test" key should contain the following subkeys:
* Filename (mandatory): relative path from the .expected of the file under test.
* Bugzilla (optional): reference to a bug related with the file.
* Comment (optional): brief description of why that file is in the test suite (problematic, format example...)
The "metadata" section lists the metadata we expect to see returned by
tracker-extract (when tracker-extract is run with --output-format=json-ld).
There are a few special rules:
1. If the property name is prefixed with '@' then the property is expected in the extraction, but the value
won't be checked.
2. If the property name is prefixed with '!' then the property is NOT expected in the extraction
If the negated property has a value, it forbids the exact value. Otherwise, it forbids the property at all.
E.G. The extraction is: { "@type": "nmm:Video", "nfo:duration": 50 }
{ "!@type": "nmm:Audio" } -> PASS because there is no { "@type": "nmm:Audio" }
{ "!nfo_duration": "12" } -> PASS because duration has a different value
{ "!nfo_duration": null} } -> FAIL because there shouldn't be any nfo:duration at all
3. There is (so far only) one constant defined to use in the values:
@URNUUID@ meaning an automatic generated urn:uuid:1231-123-123 URL
[This constant must not be used with multiple-valued properties. The code only check the first result.]
Example
======
Everything together should look like this:
{
"test": {
"Filename": "x.mp4",
"Bugzilla": "GB#123123",
"Comment": "Video usually clasiffied as Audio by mistake"
},
"metadata": {
"@type": "nmm:Video",
"!@type": "nmm:Audio",
"nfo:duration": 123
}
}
|
