Skip to content

Conversation

robertbastian
Copy link
Member

@robertbastian robertbastian commented Oct 2, 2025

#6787

Fixes #7034

@robertbastian robertbastian marked this pull request as ready for review October 6, 2025 06:59
Manishearth
Manishearth previously approved these changes Oct 6, 2025
Comment on lines 238 to 248
let url = if tag >= "release-78.1" {
format!(
"https://github.com/unicode-org/icu/releases/download/{tag}/icu4x-exportdata-{}.zip",
tag.replace("release-", "")
)
} else {
format!(
"https://github.com/unicode-org/icu/releases/download/{tag}/icuexportdata_{}.zip",
tag.replace('/', "-")
)
};
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggestion:

  • Keep the / to - substitution, in addition to the "release-" substitution
  • Special-case "release-77.1" and any other releases that this datagen version is compatible with

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Based on the agreement with the ICU TC, this change is no longer necessary.

Copy link
Member

@hsivonen hsivonen left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

AFAICT, this pessimizes the compressibility check that before this change was formulated to validate the length upon construction so that bound checking was elided at access time. Perhaps that's not super-important, but it still seems bad to break that property if we could avoid breaking that property.

let field = self
.last_primaries
.get(MaxVariable::Currency as usize + usize::from(b >> 4))
.unwrap_or_else(|| CollationSpecialPrimaries::HARDCODED_FALLBACK[usize::from(b >> 4)]);
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

To keep the above comment true, shouldn't the compressibility bits be referenced as a fixed-length array in the collator constructor instead of doing slice access on every call to this method?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

done

@robertbastian robertbastian requested a review from hsivonen October 8, 2025 16:19
hsivonen
hsivonen previously approved these changes Oct 9, 2025
Copy link
Member

@hsivonen hsivonen left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. Thanks.

@robertbastian robertbastian requested a review from sffc October 9, 2025 17:45
@robertbastian robertbastian removed the request for review from echeran October 10, 2025 16:27
Copy link
Member

@sffc sffc left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please wait for @hsivonen to approve the collator changes (edit: looks like he already did)

@robertbastian robertbastian merged commit 6dd8b66 into unicode-org:main Oct 13, 2025
31 checks passed
@robertbastian robertbastian deleted the 78 branch October 13, 2025 08:13
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

icuexportdata name changes in ICU 78

4 participants