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
|
# README
This program provides a fast local dictionary, with the option to add any word &
its definition to Anki.
You'll need a working Go runtime in order to build this program. Usually you can
get this from your distribution's package manager, or from
https://go.dev/doc/install.
## Getting started
You need a copy of
https://kaikki.org/frwiktionary/raw-wiktextract-data.jsonl.gz. Uncompress it to
raw-wiktextract-data.jsonl. This is a structured dump of French wiktionary.
To be able to add Anki cards, you need to install the `anki-connect` extension.
You can find instructions to do so here: https://git.sr.ht/~foosoft/anki-connect
When you're running this app, you'll also need to have Anki open (with the
`anki-connect` plugin installed) in order to add new cards.
Also from Anki, take note of the name of the Deck where you want to add new
cards, and the `Type` of these cards (appears in the upper-left of the 'add new
cards' dialog in Anki).
(Note: currently this application assumes that your card type has two fields:
"Front" and "Back". Eventually, this may be configurable. If the fields are
called something else in your card type, you'll currently have to update the
json tags for the `fields` struct in `add.go`.)
With all this done, you can now run the app! The first time you run it, the app
will build a SQLite database of words and definitions from the Wiktionary data,
which will take a minute or two. From the root of this repository, you can use a
command like this to start the app:
```
go run . -rawDictionary=raw-wiktextract-data.jsonl -deck="Français" -model="Basic-830ae"
```
In addition to starting the program and then looking up a word interactively,
it's also possible to provide the first word to look up as a command-line
argument. Then, for example, you could add a shell function to invoke the
program and immediately go to a definition.
For example, if you've compiled the program and placed it in your path, you
could add a shell function to your `.zshrc` or `.bashrc` with all the arguments
you need, like this:
```
wk() {
french-wiktionary-flashcards dictionary=/path/to/dictionary.sqlite3 -deck="Français" -model="Basic-830ae" -initialWord="$*"
}
```
and then, to start the program and go immediately to a definition (e.g.
"poisson"), you could invoke it like this:
```
$ wk poisson
```
## Usage
```
Usage of french-wiktionary-flashcards:
-apiURL string
Base URL to access the anki-connect plugin API. (default
"http://localhost:8765")
-deck string
Name of the deck where new Anki cards will be created.
-dictionary string
Path to the parsed dictionary data. This will be generated
from rawDictionary. (default "dictionary.sqlite3")
-model string
Name of the card type ('model') for new Anki cards.
-initialWord string
Optional: first word to look up on program launch.
-rawDictionary string
Path to the raw wiktionary data. You can get this by
downloading and unzipping
https://kaikki.org/frwiktionary/raw-wiktextract-data.jsonl.gz
(for French). (default "raw-wiktextract-data.jsonl")
```
# TODO
- Include context hints (e.g. see entry for 'panais', which has "Agriculture",
"Par métonomie", etc) in definitions (tricky because the tags are in english
and I would want to display them in the target language).
- suggest closest word when there's no match (help with spelling errors). Not
sure if this is too much complexity for my concept for this application (e.g.
how would the interface work?)
- allow setting the language for initial processing, so that we could support
languages other than French
- italicise part-of-speech in the TUI
- maybe we could create the Anki and the TUI definitions at the same time
during initial parsing? Then we'd have all the information required to do
a really good job of it.
- better gender tags -- jq through the source data to see what's common
- some kind of tests?
# Ideas for future improvements
- if the deck and model aren't provided, figure out the defaults and use them
- the first run could download the dictionary database and do all the setup,
with interactive options to choose the language and such.
- provide history (scrollback) of words looked up, to make it easier to switch
between definitions
- definitions for different parts of speech could be more clearly separated in
the preview window. I think it's best to have all the different senses
combined onto one card, but it could still be interesting to allow the user to
scroll between them.
- store data in sqlite with more structure to make it easier to customize
the format of definitions on cards
- figure out some way to customize the names of the fields on the flashcards
|
