Cross-references broken when ref includes special characters (e.g., hyphens or underscores) about ghostwriter HOT 6 CLOSED

As a workaround, you can use {{.ref exampleptc}}.

The current code strips non-ASCII-alphanumeric characters from the evidence name when generating it's ref, but the {{.ref}} tag does not do that, hence the mismatch. We could change it to strip the name similarly, but if I'm reading the docx spec right, we could also use the full name in double quotes and just escape " and \ characters. @chrismaddalena Would appreciate your thoughts.

from ghostwriter.

@ColonelThirtyTwo Yeah, let's try using quotes. If it works, that's the better solution. Escaping quotes is easy enough, and it should be unlikely that someone uses double quotes in a name anyway. This would also avoid a situation where someone could have evidence1 and evidence-1 and have their reference to the second one end up pointing to the first if we stripped the -.

from ghostwriter.

Speaking of duplicate evidence files, there does not appear to be any checks to block that. I'm able to create two evidence items with the same name. So, quoting the evidence names wouldn't help here.

I think this indicates that there are two issues:

1. Special characters are treated differently for figure naming and cross-reference naming. One strips the special characters and the other does not, respectively.
2. There are no checks to prevent evidence naming duplicates upstream in the Ghostwriter UI or when stripping them when generating the report. Currently, a user can create multiple evidence items with the same friendly name.

from ghostwriter.

Could we concatenate the evidence friendly name and the evidence ID when generating the DOCX? This should prevent naming collisions, even if multiple evidence items have the same friendly name.

For example:

• "example-evidence", id 10 -> "example-evidence10"
• "example-evidence", id 11 -> "example-evidence11"

from ghostwriter.

Name+ID could still theoretically experience collisions but I think a user would really have to go out of their way (or, be incredibly [un]lucky) to encounter it.

For example:

• "example-evidence1", id: 10 -> "example-evidence110"
• "example-evidence", id: 110 -> "example-evidence110"

from ghostwriter.

Keep in mind that captions made with {{.caption <name>}} can be referenced with {{.ref}} as well, so it's not as simple as looking up the evidence by the name passed to {{.ref}}. You could even reference a bookmark in the template with {{.ref}} if you knew the internal reference name.

But I can see name collisions being an issue moving forward, especially with extra fields attempting to generate an evidence figure multiple times.

from ghostwriter.