SQLite Forum

Miniscule Doc Typos
Login

Miniscule Doc Typos

(1) By SeverKetor on 2023-08-05 18:32:22 [link] [source]

While very bored at work, I found some small typos in https://sqlite.org/opcode.html.

In section 2.2, "P5 is an 16-bit unsigned integer..."
In section 2.3's list of objects, "An arbitrary length strings"
In the opcode list entry for Clear, "The table being clear is in the main [...] the table to be clear is in the auxiliary..."
In the entry for Compare, "This opcode must be immediately followed by an Jump opcode."
In the Delete entry, I think there's a missing "is"? It says "...that indicates this delete one of several associated with deleting [...] Exactly one of those deletes..."
In the Destroy entry, "...then the table to be clear is in the auxiliary..."
In the ElseEq entry, there's a missing "the", "If result of an Eq comparison..."
In the IfNoHope entry (didn't realize SQL statements could get depressed!), "...but extra works is performed." and "If it is not possible, we can skips some work."
In the ReleaseReg entry, "...will release registers before their are used..."
In the SorterData entry, "This opcode is normally use to move..."
In the ZeroOrNull entry, "If all both registers P1 and P3..."

(2.1) By Stephan Beal (stephan) on 2023-08-05 19:05:55 edited from 2.0 in reply to 1 [link] [source]

I found some small typos in https://sqlite.org/opcode.html.

i'm working on these now but at least one of them is intentional:

"This opcode must be immediately followed by an Jump opcode."

That's actually correct in the native documentation (in comment blocks in vdbe.c):

** This opcode must be immediately followed by an OP_Jump opcode.

The OP_ prefix is used to generate a link to the Jump opcode and strip that prefix from the link's text. If we correct "an" to "a" there then the original text-format doc has a new typo and someone will eventually get around to "fixing" it back to "an" all over again.

Edit:

In the ZeroOrNull entry, "If all both registers P1 and P3..."

i'm reluctant to fix that one (but will) because "all both" or is a legitimate German slang expression ;).

(3) By SeverKetor on 2023-08-05 19:05:53 in reply to 2.0 [link] [source]

Hmm, how about "This opcode must be immediately followed by the Jump/OP_Jump opcode"? Slightly less nice, but works for both places.

(4) By Stephan Beal (stephan) on 2023-08-05 19:11:10 in reply to 3 [source]

"This opcode must be immediately followed by the Jump/OP_Jump opcode"

There are at least 3 cases of "an OP_Jump" in the text-format docs, and those are the canonical docs, so i'm hesitant to go rephrase all such instances just for the sake of nitpicking the HTML-side output. (Noting that i'm not objecting to the nitpicking, but am just nitpickingly defending the opposing view on this one particular nit.)

i think i've got the rest of them covered (and "their" had already been fixed, but the web-facing docs are normally only updated around releases).

(5) By Warren Young (wyoung) on 2023-08-05 20:15:49 in reply to 4 [link] [source]

i'm not objecting to the nitpicking, but am just nitpickingly defending the opposing view on this one particular nit.

In that spirit…”minuscule”. 😉