Simplify Clumplet Reader/Writer + code analysis issues

[mrotteveel]

Uses isTagged() instead of having three or four different ways of discerning tagged vs untagged (in some cases, missing some kinds). Some code cleanup, annotating [[noreturn]]/noexcept, addressing switch code analysis warnings.

Also fixes a bug where ClumpletWriter.reset(...) shouldn't work for tagged buffers (apparently not happening in production code?)

[AlexPeshkoff]

lenSize is set only by our code, please remove this default label, use fb_assert() instead

[mrotteveel]

OK

[mrotteveel]

Do you mean fb_assert(false) here in the default branch, or adding fb_assert(lenSize >= 0 && lenSize <= 2 || lenSize == 4) earlier?

[mrotteveel]

Right now I pushed a change with the second assert I mentioned in the switch; do you think that it's OK, or do you want it moved outside the switch, or as fb_assert(false)?

[mrotteveel]

@AlexPeshkoff Do you want me to make any other changes.

BTW: I suggest squash-merging this for a cleaner history.

[AlexPeshkoff]

On 4/10/25 12:16, Mark Rotteveel wrote: @AlexPeshkoff <https://github.com/AlexPeshkoff> Do you want me to make any other changes. BTW: I suggest squash-merging this for a cleaner history. — Reply to this email directly, view it on GitHub <#8507 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA44OUPHAGAZNWHJIPWGPP32YYZHPAVCNFSM6AAAAAB2VOMJVWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOOJSGEYDAOJWGI>. You are receiving this because you were mentioned.Message ID: ***@***.***> BTW: I suggest squash-merging this for a cleaner history.

I almost always prerer that way, details of how solution was taken typically makes no interest.

[TommiPrami]

ClumpletWriter.reset(...)

From out of curiocity what/where this term "Clumpet" comes from and/or mean, in this context?

Google gave me an answers that where more confusing than informative 😄

[mrotteveel]

ClumpletWriter.reset(...)

From out of curiocity what/where this term "Clumpet" comes from and/or mean, in this context?

Google gave me an answers that where more confusing than informative 😄

It's clumplet, not clumpet; I believe it is a word made up by Jim Starkey (not 100% sure though), and I think you should read it as "[a collection of] small clumps of bytes", as in "-let" being a diminutive, and clumps of bytes (compare clumps of clay) for the contents of the buffer. That said, I might just read too much in it.

[AlexPeshkoff]

ClumpletWriter.reset(...)

From out of curiocity what/where this term "Clumpet" comes from and/or mean, in this context?
Google gave me an answers that where more confusing than informative 😄

It's clumplet, not clumpet; I believe it is a word made up by Jim Starkey (not 100% sure though), and I think you should read it as "[a collection of] small clumps of bytes", as in "-let" being a diminutive, and clumps of bytes (compare clumps of clay) for the contents of the buffer. That said, I might just read too much in it.

Not by Jim Starkey but by Nickolay Samofatov .

[dyemanov]

ClumpletWriter.reset(...)

From out of curiocity what/where this term "Clumpet" comes from and/or mean, in this context?
Google gave me an answers that where more confusing than informative 😄

It's clumplet, not clumpet; I believe it is a word made up by Jim Starkey (not 100% sure though), and I think you should read it as "[a collection of] small clumps of bytes", as in "-let" being a diminutive, and clumps of bytes (compare clumps of clay) for the contents of the buffer. That said, I might just read too much in it.

Not by Jim Starkey but by Nickolay Samofatov .

Header page clumplets exist in the InterBase 6.0 source code, so I believe credits really belong to Jim ;-) Nickolay just extended the term to various API parameter buffers.

[mrotteveel]

@AlexPeshkoff Anything else, or can I merge it?

[AlexPeshkoff]

@AlexPeshkoff Anything else, or can I merge it?

OK to merge