Split merge segments

Splitting and merging is available to Superusers and Administrators only. It is the primary new power of the Superuser, intended for those who are the primary translator of a root text.

If you’re a Superuser and do not know basic HTML, you need to learn. The tags you will deal with are typically paragraphs (<p>), and occasionally lists (<ol>, <ul>, <dl>, <li>), and headings (<h1>, <h2>, etc.).

The split/merge function affects all JSON files associated with the relevant segments, not just the ones you see.

numbering

It will not only affect the specific split/merged segments, but will automatically adjust the numbers of the following segments. This will permanently change the segment number of those segments. It will affect the numbers, not of the whole text, but as far as the next minor segment number. In this way the effects of any change are kept local and minimal.

To illustrate, suppose the following:

mn123.4.5: Doe, a deer,
mn123.4.6: a female deer.
mn123.4.7: Ray, a drop of golden sun.
mn123.4.8: Me, a name I call myself.
mn123.5.1: Far, a long long way to run. 
mn123.5.2: So, a needle pulling thread.

We merge the first two segments, and the numbers are adjusted as follows.

mn123.4.5: Doe, a deer, a female deer.
mn123.4.6: Ray, a drop of golden sun.
mn123.4.7: Me, a name I call myself.
mn123.5.1: Far, a long long way to run. 
mn123.5.2: So, a needle pulling thread.

And the opposite with split.

mn123.4.5: Doe, a deer,
mn123.4.6: a female deer.
mn123.4.7: Ray, a drop 
mn123.4.8: of golden sun.
mn123.4.9: Me, a name I call myself.
mn123.5.1: Far, a long long way to run. 
mn123.5.2: So, a needle pulling thread.

Again, these new numbers propagate automatically to all related files, you do not have to do anything.

text

The text is adjusted so far as possible so the split/merge can be automatic.

In splits, a new empty segment is created. Since there is no way of knowing what part of the text should appear in the new segment, no text is shifted automatically. The Superuser should cut and paste the root text and their own translation as appropriate. For other translators, their text will remain attached to the first segment. This means, at worst, their text is not divided exactly the same as the new root.

In merges, two segments become one, and we simply add the two segments one after the other with a space in between. This means that merges can almost always be fully automatic.

If translators have subscribed to the Superuser (which is recommended), they will get a notification, and can adjust their own translation if need be.

html

HTML is a little tricky, as splitting and merging can easily create missing or invalid HTML. Thus, whenever a split or merge is performed, Bilara forces open the HTML column (if it is not already there) and asks you to check.

In Bilara data, every segment must have one and only one {} in the HTML column; blank columns are forbidden. {} tells the system, “insert text here”. A segment can have just a {}, or else {} and markup too. The {} must be inside the markup.

• ✅ {}</p>
• 🚫 </p>{}
• ✅ <p>{}

By default, Bilara will add {} to the new segment when splitting. You must check this and adjust if necessary.

For merging, the merged HTML will include HTML of both segments, dropping unnecessary {}. Again, you must check it.

Be aware that each segment should have HTML that is not only valid, but also makes sense.

• <p></p><p>{} is valid, but makes no sense, as it adds an empty paragraph.

HTML rules

Generally speaking the following rules apply to our HTML:

• headings must be one segment only
• also list items should be one segment (I'm not sure if this is always observed.)