Enforce starting newlines#563
Merged
max-sixty merged 38 commits intomitsuhiko:masterfrom Sep 10, 2025
Merged
Conversation
max-sixty
added a commit
to max-sixty/insta
that referenced
this pull request
Aug 12, 2024
This will make diffs in mitsuhiko#563 more reasonable (Though it was good to have some snapshots that _weren't_ updated, as meta-tests that un-updated snapshots still worked -- possibly we try and add an integration test in that direction...)
max-sixty
added a commit
that referenced
this pull request
Aug 12, 2024
This will make diffs in #563 more reasonable (Though it was good to have some snapshots that _weren't_ updated, as meta-tests that un-updated snapshots still worked -- possibly we try and add an integration test in that direction...)
max-sixty
commented
Aug 12, 2024
max-sixty
added a commit
to max-sixty/insta
that referenced
this pull request
Aug 17, 2024
We previously excluded these. But it's somewhat required for mitsuhiko#563, and I don't think there's a big downside -- we only generate pending files when a snapshot doesn't match of `--force-update-snapshots` is passed.
max-sixty
added a commit
to max-sixty/insta
that referenced
this pull request
Sep 17, 2024
Moving some unambiguous code from mitsuhiko#563 into a new PR so we can keep moving
max-sixty
added a commit
to max-sixty/insta
that referenced
this pull request
Sep 17, 2024
Another PR taking some unambiguous code from mitsuhiko#563
max-sixty
added a commit
that referenced
this pull request
Sep 17, 2024
Moving some unambiguous code from #563 into a new PR so we can keep moving
max-sixty
added a commit
that referenced
this pull request
Sep 17, 2024
Another PR taking some unambiguous code from #563
max-sixty
added a commit
that referenced
this pull request
Sep 19, 2024
#581) This solves the issue in #573 for the moment: - When `--force-update-snapshots` in passed, we only update inline snapshots when there's some difference _within_ the string, such as an additional linebreak at the start or the end - We no longer update the surrounding hashes. In the existing code, this happened because we were writing too many inline snapshots, not because we were actually checking the number of hashes - Checking the number of hashes isn't easy to do — reason outlined at #573 (comment) - The main change in the code is that we store the unnormalized snapshot and then normalize when we need to. The existing code stored the normalized version, which meant we couldn't evaluate whether the unnormalized versions were different It's a decent number of changes, but will integrate nicely with #563. ~(FYI we currently don't look at the indentation, but we could adjust this)~ Now indentation works too
Collaborator
Author
|
One downside (but balanced against lots of upsides!) is that because we insert a newline at the start of a multiline string, a literal string can look a bit odd: assert_snapshot!("
foo
bar
", @r"
foo
bar
");Notice that because the original literal has a newline after its This is mostly only noticeable in examples: very few generated values start with a newline! |
Closed
Collaborator
Author
|
I'm tempted to merge this as there's no downside to enforce starting newlines; and it simplifies the code overall lmk any objections... |
- Test common indentation removal with 4 and 8 spaces - Test relative indentation preservation - Test tab indentation handling - Test edge cases with empty lines and whitespace - Test single line trimming behavior - Test force-update aggressive trimming
- Replace problematic trimming_indentation.rs with test_trimming_behavior.rs - New tests create snapshots with excess indentation and verify trimming works - Tests now correctly assert the expected trimmed output - All tests passing locally
- Fix module ordering in main.rs - Fix line breaks in assertions - Remove trailing whitespace
- Remove redundant test_trimming_behavior.rs test file - Keep comprehensive trimming.rs with precise diff assertions - All trimming tests now in one well-structured file - Tests cover: basic trimming, force-update, file snapshots, complex indentation
Based on code review feedback: - Add test_edge_cases covering mixed tabs/spaces, no common indent, whitespace-only - Improve file snapshot test assertions to verify structure - Tests now cover all important edge cases identified in review
- Test that old format snapshots still pass (backwards compat) - Test that --accept migrates old format to new trimmed format - Test multiline warning behavior - Test that properly formatted snapshots remain unchanged - Test force-update on old format snapshots These tests ensure: 1. No breaking changes for existing users 2. Clean migration path via --accept 3. Proper warning messages 4. Correct behavior preservation
- Merged trimming.rs and backwards_compat.rs into inline_snapshot_trimming.rs - 9 comprehensive tests covering: - Basic trimming functionality (4 tests) - Backwards compatibility (5 tests) - Tests cover all critical scenarios: - Old format compatibility - Migration path via --accept - Force-update aggressive trimming - Warning behavior - File snapshots - Single file is easier to maintain and understand
Co-authored-by: Claude <no-reply@anthropic.com>
- Simpler, clearer description - Focuses on what users care about: readable snapshots - Avoids implementation details
The feature was announced in 1.41.0 but this PR implements it properly
Copy the existing 1.41.0 entry about --force-update-snapshots updating inline snapshots' delimiters and indentation to Unreleased section, since this PR properly implements that feature
The changelog now reflects that leading newlines are no longer trimmed during assertions. Co-authored-by: Claude <no-reply@anthropic.com>
Move the indentation trimming entry from 1.41.0 (where it was incorrectly announced but not implemented) to Unreleased (where it's actually being implemented in this PR)
This reverts commit 70a962c.
The indentation trimming functionality already existed in master. The actual NEW feature is the leading newline requirement for multiline snapshots and the associated warning. Changes: - Reduced from 12 tests (777 lines) to 6 focused tests (344 lines) - Tests now specifically verify the leading newline requirement - Each test case is properly isolated (no test munging) - Removed redundant tests for pre-existing trimming functionality - Fixed test assertions to check exact behavior (no || conditions) - Improved test documentation to clarify NEW vs existing features
- Remove empty line between doc comment and function - Fix format string to use inline variables per clippy suggestion
Collaborator
Author
|
merging but one click to revert if anyone disagrees |
Merged
max-sixty
added a commit
that referenced
this pull request
Nov 20, 2025
## Summary Prepare for the 1.44.0 release: - Bump version to 1.44.0 in `insta/Cargo.toml` and `cargo-insta/Cargo.toml` - Update CHANGELOG.md with all changes since 1.43.2 ## Changes in 1.44.0 - Added non-interactive snapshot review and reject modes for use in non-TTY environments (LLMs, CI pipelines, scripts) #815 - Add `--disable-nextest-doctest` flag with deprecation warning #803 - Add ergonomic `--test-runner-fallback` / `--no-test-runner-fallback` flags #811 - Apply redactions to snapshot metadata #813 - Remove confusing 'previously unseen snapshot' message #812 - Speed up JSON float rendering #806 (@nyurik) - Allow globset version up to 0.4.16 #810 (@g0hl1n) - Improve documentation #814 (@tshepang) - Enforce starting newlines in assertions #563 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude <noreply@anthropic.com>
max-sixty
added a commit
to max-sixty/insta
that referenced
this pull request
Nov 27, 2025
This fixes a regression introduced in 1.44.0 where existing inline snapshots using the legacy multiline format would fail to match. The problem: In 1.44.0, we changed how inline snapshots handle leading newlines (PR mitsuhiko#563). This caused snapshots like: insta::assert_snapshot!(value, @r" Unconflicted Mode(FILE) 0839b2e9412b "); to fail when the actual value is "Unconflicted Mode(FILE) 0839b2e9412b". The legacy format stored single-line content in a multiline raw string with source code indentation, which after processing becomes " Unconflicted...\n " - the leading spaces from indentation weren't being stripped in the comparison. The fix: In `matches_legacy`, detect the legacy single-line-in-multiline pattern by checking if the raw contents contain a newline but collapse to a single line after trimming. When this pattern is detected, apply `trim_start()` to strip the source code indentation artifacts. This ensures: - Legacy snapshots continue to pass (with a warning to update) - Modern single-line snapshots with intentional spaces are preserved - True multiline snapshots are unaffected We take backward compatibility seriously - upgrading insta should not cause existing valid tests to fail. Reported-by: jj-vcs/jj See: mitsuhiko#819 (comment) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
3 tasks
max-sixty
added a commit
that referenced
this pull request
Nov 27, 2025
## Summary This fixes a regression introduced in 1.44.0 where existing inline snapshots using the legacy multiline format would fail to match. **The problem:** PR #563 changed how inline snapshots handle leading newlines. This caused snapshots like: ```rust insta::assert_snapshot!(value, @r" Unconflicted Mode(FILE) 0839b2e9412b "); ``` to fail when the actual value is `"Unconflicted Mode(FILE) 0839b2e9412b"`. The legacy format stored single-line content in a multiline raw string with source code indentation. After `from_inline_literal` processing, this becomes `" Unconflicted...\n "` — the leading spaces from code indentation weren't being stripped in the legacy comparison path. **The fix:** In `matches_legacy`, detect the legacy single-line-in-multiline pattern: ```rust let is_legacy_single_line_in_multiline = sc.contents.contains('\n') && sc.contents.trim_end().lines().count() <= 1; ``` When this pattern is detected, apply `trim_start()` to strip the source code indentation artifacts. This ensures: - ✅ Legacy snapshots continue to pass (with a warning to update) - ✅ Modern single-line snapshots with intentional leading spaces are preserved - ✅ True multiline snapshots are unaffected ## Test plan - [x] Added 8 comprehensive functional tests covering: - Single-line content in multiline format (the jj bug) - Raw string multiline variant - Force update reformatting works - True multiline content unaffected - Multiline with relative indentation preserved - Intentional leading spaces correctly fail (no false positives) - Trailing whitespace line edge case - Empty snapshot edge case - [x] All 72 functional tests pass - [x] Full test suite passes Reported-by: jj-vcs/jj See: #819 (comment) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude <noreply@anthropic.com>
max-sixty
added a commit
to max-sixty/insta
that referenced
this pull request
Nov 27, 2025
…ko#830) ## Summary This fixes a regression introduced in 1.44.0 where existing inline snapshots using the legacy multiline format would fail to match. **The problem:** PR mitsuhiko#563 changed how inline snapshots handle leading newlines. This caused snapshots like: ```rust insta::assert_snapshot!(value, @r" Unconflicted Mode(FILE) 0839b2e9412b "); ``` to fail when the actual value is `"Unconflicted Mode(FILE) 0839b2e9412b"`. The legacy format stored single-line content in a multiline raw string with source code indentation. After `from_inline_literal` processing, this becomes `" Unconflicted...\n "` — the leading spaces from code indentation weren't being stripped in the legacy comparison path. **The fix:** In `matches_legacy`, detect the legacy single-line-in-multiline pattern: ```rust let is_legacy_single_line_in_multiline = sc.contents.contains('\n') && sc.contents.trim_end().lines().count() <= 1; ``` When this pattern is detected, apply `trim_start()` to strip the source code indentation artifacts. This ensures: - ✅ Legacy snapshots continue to pass (with a warning to update) - ✅ Modern single-line snapshots with intentional leading spaces are preserved - ✅ True multiline snapshots are unaffected ## Test plan - [x] Added 8 comprehensive functional tests covering: - Single-line content in multiline format (the jj bug) - Raw string multiline variant - Force update reformatting works - True multiline content unaffected - Multiline with relative indentation preserved - Intentional leading spaces correctly fail (no false positives) - Trailing whitespace line edge case - Empty snapshot edge case - [x] All 72 functional tests pass - [x] Full test suite passes Reported-by: jj-vcs/jj See: mitsuhiko#819 (comment) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This is an example of what I've been suggesting at #506 (+ some linked issues):
Stacked on Unify handling of file & inline snapshots (compat) #528 so that would need to merge firstLinesdeals ambiguously with trailing new lines)matches_legacymethod on snapshot contents — this encapsulates the older formats we still support and warns about themIt's not ready to merge yet, because we don't seem to actually support--force-update-snapshotson inline values, which is somewhat required. But I'm publishing for feedback & to put some weight on #528 in case we don't want to pursue.