Skip to content

search command is working on a local scope#4

Merged
ranyitz merged 4 commits into
masterfrom
search-command
Jan 26, 2017
Merged

search command is working on a local scope#4
ranyitz merged 4 commits into
masterfrom
search-command

Conversation

@davidfirst
Copy link
Copy Markdown
Member

No description provided.

@ranyitz ranyitz merged commit dc88c55 into master Jan 26, 2017
@ranyitz ranyitz deleted the search-command branch April 27, 2017 12:42
davidfirst added a commit that referenced this pull request Apr 27, 2020
* public-api links are created only for components in components directory

* extract bind() to a seperate public function

* create dependencies links only for components that have references in bit.json

* never delete the entire node_module/bit dir

* create public-api links for inline components
davidfirst added a commit that referenced this pull request Oct 8, 2025
Combine duplicate branches that both called addSpecificComponentConfig
into a single else clause covering cases #2, #3, #4, #5
davidfirst added a commit that referenced this pull request May 13, 2026
Audit 06 enumerates every explicit cache-invalidation call site, classifies
each by what it currently invalidates vs. what it needs to invalidate, and
identifies three patterns of over-invalidation:

- Pattern A (#10, #11): callers already have the affected ID list in scope,
  but call clearAllComponentsCache anyway. Mechanical swap to
  clearComponentsCache(ids). Tasks 8.11.
- Pattern B (#12, #4-6, #13): callers know the seed IDs but not the full
  affected set (e.g. auto-tag dependents, env consumers). Each needs a
  small reverse-index. Tasks 8.12, 8.14.
- Pattern C (#3, #8, #9): callers genuinely don't know the affected set.
  Each handled case-by-case: dead-code check, file-read narrowing, or
  accept worst-case nuke. Tasks 8.13, 8.15.

Audit also documents 5 "OK, don't regress" per-id sites and 3 FS-cache
sites, plus the two indirect cascade triggers (bitmap invalidate, configStore
invalidate) which are already minimal.

15 full-clear sites total, 7 of them are over-clears. The largest expected
wins are from narrowing snapping's pre-tag/snap clear (8.12) and the
install-time clears (8.14, deferred until Lever 1 ships and we measure).

No code changes — audit + task planning only.
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.

3 participants