Record and Tuple object model spec #132

rricard · 2020-06-10T09:13:48Z

No description provided.

littledan

Some initial comments

spec/overview.html

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

spec/data-types-and-values.html

rricard · 2020-06-11T17:28:35Z

Thanks dan, I'll make another pass on this tomorrow

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

spec/data-types-and-values.html

spec/overview.html

spec/ordinary-and-exotic-object-behaviours.html

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

littledan

Some editorial comments on the Record side, but things seem to be getting in good shape; I mostly just have editorial comments. The Tuple side seems to be in progress so I'll leave that for a later review.

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

ljharb · 2020-06-24T22:00:18Z

spec/ordinary-and-exotic-object-behaviours.html

+        <h1>RecordCreate ( _value_ )</h1>
+        <p>This abstract operation RecordCreate takes a _value_ argument. It is used to specify the creation of new Record exotic objects. It performs the following steps when called:</p>
+        <emu-alg>
+            1. Assert: Type(_value_) is a Record.


why would RecordCreate take a Record? This seems like more the domain of https://tc39.es/ecma262/#sec-toobject than of its own abstract op

I was a little mystified by the naming convention at first, but see analogous operations like https://tc39.es/ecma262/#sec-stringcreate and https://tc39.es/ecma262/#sec-modulenamespacecreate

hmm, strange that String has it but Number, Boolean, etc don't (objects having a FooCreate makes sense to me; but primitives don't). It feels like the precedent is more for primitives not to have one.

I guess what distinguishes String is that its wrapper is exotic, like Record. Maybe we should continue this discussion in an editorial issue in ecma262? I want to make sure we have clear text somewhere or other to explain how the wrapper is set up, and I think it should match String, but I am fine with moving to some other way of organizing the text.

Please open an issue if you think this should still be changed @ljharb , we can discuss the choices in more detail.

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

rricard · 2020-06-26T08:18:02Z

Thanks Dan, I will try to get around it soon!

Co-authored-by: Jordan Harband <[email protected]>

…cord-tuple into rb/ops-and-algs

spec/data-types-and-values.html

ljharb · 2020-06-30T16:18:29Z

spec/ordinary-and-exotic-object-behaviours.html

+          1. Assert: IsPropertyKey(_P_) is *true*.
+          1. If Type(_P_) is a Symbol, return *undefined*.
+          1. Let _getResult_ be ! RecordGet(_R_, _P_).
+          1. If _getResult_ is ~empty~, return *undefined*.


does this mean i can't set an own expando string property on a boxed Record? var o = Object('a'); o.foo = true; o.foo; works, for example.

That is the intent here, if that is an issue for you we should probably discuss that in an issue. If that is a blocker I'd be happy to enable additional properties in the spec and wait for Dan's return to discuss it more in depth

Creating an issue, this is worth discussing either way. I think I'll do a first draft as is and we can update that once that issue progresses.

spec/data-types-and-values.html

Co-authored-by: Jordan Harband <[email protected]>

spec/data-types-and-values.html

ljharb · 2020-06-30T16:48:22Z

spec/data-types-and-values.html

+              1. Assert: Type(_kv_.[[Key]]) is String.
+              1. Assert: Type(_kv_.[[Value]]) is not Object.
+              1. If there is no entry _existing_ in _uniqueEntries_ such that _existing_.[[Key]] is _kv_.[[Key]], append _kv_ to _uniqueEntries_.
+            1. Let _sortedEntries_ be _uniqueEntries_ sorted such that, for any entry two entries _a_ preceding _b_, then the result of performing Abstract Relational Comparison _a_.[[Key]] < _b_.[[Key]] is *true*.


"for any entry two entries a preceding b" is something i can't get my head to understand, can you help me make sense of it?

also, by using <, it will return a different value than localeCompare would. is that intentional?

I think using < would be more accurate than localeCompare as localeCompare would have different results in different locales, what I am trying to express is char code ordering, I will attempt another change to the wording but if that is not good enough to you, is it possible
to suggest something that would express that?

How about:

Let sortedEntries be uniqueEntries sorted such that, for any two integers i, j less than the length of list uniqueEntries, If i < j then the result of performing Abstract Relational Comparison uniqueEntries[i].[[Key]] < uniqueEntries[j].[[Key]] is true.

spec/data-types-and-values.html

spec/ordinary-and-exotic-object-behaviours.html

Co-authored-by: Jordan Harband <[email protected]>

Record and Tuple abstract ops

rickbutton

couple of minor copy-paste fixes

spec/ordinary-and-exotic-object-behaviours.html

Co-authored-by: Rick Button <[email protected]>

…/proposal-record-tuple into spec/record-object-model

…model

rickbutton

approving this and will merge it. we can continue to iterate on this text in subsequent PRs, there will definitely be many more changes after this PR :)

* inital pass of R&T abstract ops * Update spec/data-types-and-values.html Co-authored-by: Jordan Harband <[email protected]> * fix SameValue usage inside of SameValueNonNumeric * reduce verbosity of samevalue checks * added missing reduce/reduceRight * simplify Tuple toString and ToLocaleString * switch to dot notiation for internal slots on records and tuples * equality/comparison operators * Update spec/data-types-and-values.html Co-authored-by: Jordan Harband <[email protected]> * don't create new lists for Record::equal * SameValueNonNumeric -> SameValueNonGeneric * started drafting record exotic object * Define records more in depth * Add RecordCreate * revamp definitions * add RecordTypeCreate * Redo record's define own prop * Apply suggestions from code review Co-authored-by: Jordan Harband <[email protected]> * feedback from Dan * redefine sorting * rework recordcreate * nit: add dots * additional changes * add more internal properties * get through RecordGetOwnProperty * reword Record Exotic Objects intro * Add prevent extensions * refactor internal properties a bit more * fix ! * MakeRecordValue redone * Start properly defining tuple * start structuring tuple section * More structure * advancing tuple * dan's feddback * is/is a * ! SameValue * remove an * Apply suggestions from code review Co-authored-by: Jordan Harband <[email protected]> * first attempt at tuple * adjustments * Fallback to prototype * Remove references to array.prototype * Reword MakeRecordValue again * Apply suggestions from code review Co-authored-by: Jordan Harband <[email protected]> * Apply suggestions from code review Co-authored-by: Jordan Harband <[email protected]> * Rework loop * Use directly Immutable Prototype Exotic Objects * refactor ordering line * Allow symbols to be set on exotics * minor editorial changes * explicit new list * use abstract closure instead of builtin function * Apply suggestions from code review Co-authored-by: Rick Button <[email protected]> * Fix previous bad merge Co-authored-by: Richard button <[email protected]> Co-authored-by: Rick Button <[email protected]> Co-authored-by: Jordan Harband <[email protected]>

littledan reviewed Jun 11, 2020

View reviewed changes

ljharb reviewed Jun 12, 2020

View reviewed changes

spec/data-types-and-values.html Outdated Show resolved Hide resolved

spec/data-types-and-values.html Outdated Show resolved Hide resolved

littledan reviewed Jun 12, 2020

View reviewed changes

ljharb reviewed Jun 16, 2020

View reviewed changes

spec/ordinary-and-exotic-object-behaviours.html Outdated Show resolved Hide resolved

littledan reviewed Jun 16, 2020

View reviewed changes

littledan reviewed Jun 17, 2020

View reviewed changes

spec/ordinary-and-exotic-object-behaviours.html Show resolved Hide resolved

littledan reviewed Jun 24, 2020

View reviewed changes

ljharb reviewed Jun 24, 2020

View reviewed changes

littledan reviewed Jun 25, 2020

View reviewed changes

inital pass of R&T abstract ops

15e7069

littledan mentioned this pull request Jun 26, 2020

Record and Tuple abstract ops #135

Merged

rickbutton and others added 11 commits June 29, 2020 12:21

Update spec/data-types-and-values.html

eea9090

Co-authored-by: Jordan Harband <[email protected]>

fix SameValue usage inside of SameValueNonNumeric

89de16c

reduce verbosity of samevalue checks

a22faef

added missing reduce/reduceRight

66ca962

simplify Tuple toString and ToLocaleString

19f7d5c

switch to dot notiation for internal slots on records and tuples

5ab6db6

equality/comparison operators

25d5631

Update spec/data-types-and-values.html

01fb860

Co-authored-by: Jordan Harband <[email protected]>

don't create new lists for Record::equal

e987cb4

Merge branch 'rb/ops-and-algs' of https://github.com/tc39/proposal-re…

3693f94

…cord-tuple into rb/ops-and-algs

SameValueNonNumeric -> SameValueNonGeneric

d3f12f7

ljharb reviewed Jun 30, 2020

View reviewed changes

spec/data-types-and-values.html Outdated Show resolved Hide resolved

rricard marked this pull request as ready for review June 30, 2020 16:38

rricard changed the title ~~DRAFT: record object model spec~~ Record and Tuple object model spec Jun 30, 2020

Robin Ricard added 2 commits June 30, 2020 17:40

started drafting record exotic object

c021408

Define records more in depth

dd8559c

Robin Ricard and others added 8 commits June 30, 2020 17:40

! SameValue

3f6dd44

remove an

1981ebb

Apply suggestions from code review

c08af58

Co-authored-by: Jordan Harband <[email protected]>

first attempt at tuple

7444d3b

adjustments

dfe02b5

Fallback to prototype

7be58a8

Remove references to array.prototype

584da98

Reword MakeRecordValue again

7b91d2e

rricard force-pushed the spec/record-object-model branch from 91067bb to 7b91d2e Compare June 30, 2020 16:40

ljharb reviewed Jun 30, 2020

View reviewed changes

Robin Ricard and others added 7 commits July 1, 2020 08:52

Apply suggestions from code review

528d8ef

Co-authored-by: Jordan Harband <[email protected]>

Apply suggestions from code review

d7bb485

Co-authored-by: Jordan Harband <[email protected]>

Rework loop

562cec1

Use directly Immutable Prototype Exotic Objects

6ddf9eb

refactor ordering line

f2c1e03

Allow symbols to be set on exotics

472115e

minor editorial changes

497f737

rricard mentioned this pull request Jul 1, 2020

Attach algorithms to grammar and fix previous bad merge #138

Merged

Robin Ricard and others added 3 commits July 1, 2020 16:02

explicit new list

fd15d8a

use abstract closure instead of builtin function

4e560a0

Merge pull request #135 from tc39/rb/ops-and-algs

a7c84c8

Record and Tuple abstract ops

rickbutton reviewed Jul 1, 2020

View reviewed changes

spec/ordinary-and-exotic-object-behaviours.html Outdated Show resolved Hide resolved

spec/ordinary-and-exotic-object-behaviours.html Outdated Show resolved Hide resolved

Robin Ricard and others added 4 commits July 1, 2020 16:48

Apply suggestions from code review

82d9e5c

Co-authored-by: Rick Button <[email protected]>

Fix previous bad merge

60df76e

Merge branch 'spec/record-object-model' of https://github.com/rricard…

29915d5

…/proposal-record-tuple into spec/record-object-model

Merge remote-tracking branch 'origin/master' into spec/record-object-…

2cb2411

…model

rickbutton approved these changes Jul 1, 2020

View reviewed changes

rricard merged commit 664576d into master Jul 1, 2020

rricard deleted the spec/record-object-model branch July 1, 2020 16:19

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Record and Tuple object model spec #132

Record and Tuple object model spec #132

rricard commented Jun 10, 2020

littledan left a comment

rricard commented Jun 11, 2020

littledan left a comment

ljharb Jun 24, 2020

littledan Jun 24, 2020

ljharb Jun 24, 2020

littledan Jun 25, 2020

rickbutton Jul 1, 2020

rricard commented Jun 26, 2020

ljharb Jun 30, 2020

rricard Jun 30, 2020

rricard Jun 30, 2020

rricard Jun 30, 2020 •

edited

Loading

ljharb Jun 30, 2020

rricard Jul 1, 2020

rickbutton Jul 1, 2020

rickbutton left a comment

rickbutton left a comment

Record and Tuple object model spec #132

Record and Tuple object model spec #132

Conversation

rricard commented Jun 10, 2020

littledan left a comment

Choose a reason for hiding this comment

rricard commented Jun 11, 2020

littledan left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rricard commented Jun 26, 2020

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rricard Jun 30, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rickbutton left a comment

Choose a reason for hiding this comment

rickbutton left a comment

Choose a reason for hiding this comment

rricard Jun 30, 2020 •

edited

Loading