Descriptions for import screen

[johnwdubois]

Background The import screen shows information to guide the user in selecting which import type is appropriate.
Providing a description for each data type will give the user useful information for choosing the import.

Options to display Show the import options in the following sequence.

Choose the data type that best matches your file: Song & Verse Prose One Word Per Line (OWPL) CoNLL-U Transcription Elan (tab-delimited) Interlinear Glossed Text (IGT)

Screenshot

1. Use the following new tab design, which is simpler and clearer than the current screen.
2. Notice that the table of features shown in the previous design (in the middle part of the display) is replaced in the new design by the bullet point descriptions.

BULLET POINTS Depending on the import type that is selected by the user, display the following bullet points in a window labeled "Description", displayed in the middle column of the import screen (as shown in the screenshot above).

1. Song & Verse • short lines of text • line breaks are meaningful • no word wrap • split words on whitespace • file type: plain text (.txt) • example: song, poem, etc.

2. Prose • long paragraphs of text • one hard return (newline) at end of paragraph • words wrap to fit on page or screen • split words on whitespace • file type: plain text (.txt) • example: news, blog, Wikipedia, novel

3. One Word Per Line (OWPL) • columns and rows • text column reads vertically • each row represents 1 word (token) • each column shows a word feature • file type: spreadsheet (.csv) • example: Santa Barbara Corpus .csv

4. CoNLL-U • columns and rows • text column reads vertically • each row represents 1 word (token) • each column shows a word feature • hashtag lines mark unit features • file type: spreadsheet (.csv) • example: Universal Dependencies corpus

5. Transcription • one unit per line • text reads normally • tab for speaker labels • tab for timestamps (optional) • split words on whitespace • file type: tab-delimited text (.txt, .csv) • example: Santa Barbara Corpus .txt

6. Elan (tab-delimited export) • one unit per block • blocks have 1+ lines • text reads normally • tab for speaker labels • tab for timestamps • split words on whitespace • file type: tab-delimited text (.txt) • example:

7. Interlinear Glossed Text (IGT) • one unit per block • blocks have 2+ lines • blocks separated by a blank line • split morphs on whitespace & hyphen • file type: plain text (.txt) • example: Nuuchanulth

Additional updates

1. Note the changes above to the text to be shown (for example, the word "tokenize" is changed to "split" wherever it is found, especially on the import screen).
2. In the examples displayed for "Transcription" and "Song & Verse", don't show the document's title. Show the title only for the "Prose" example.

[gtroiani]

@ZoeF99 can you prepare each of the bullet points as strings for localization?

[hopesu21]

Translation of bullet points in various languages will be required later.

[terrydubois]

Here is Lauren's mockup of the Pre-Import screen made with Figma

[johnwdubois]

Check to make sure localizations for the import screen are present in the next build of Rezonator.

[johnwdubois]

Translations needed for all descriptions.