profiler-md 0.9.0 → 0.9.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/cli/index.mjs +431 -298
- package/dist/index.d.mts +31 -9
- package/dist/index.js +2 -2
- package/dist/man/profiler-md.1 +1 -1
- package/dist/node/index.mjs +2 -2
- package/docs/formats/v8-heap-profile.md +2 -1
- package/package.json +2 -1
- package/docs/CLAUDE.md +0 -22
|
@@ -8,7 +8,8 @@ It stores an allocation call tree whose nodes each state the bytes allocated at
|
|
|
8
8
|
that call site, and a sequence of samples referencing those nodes.
|
|
9
9
|
|
|
10
10
|
Its samples are allocations, so it converts to a call stack profile of where the
|
|
11
|
-
program allocated
|
|
11
|
+
program allocated. By default it contains only the samples whose objects were
|
|
12
|
+
still alive when it was written.
|
|
12
13
|
|
|
13
14
|
It supports a single byte metric per profile and a single profile per file.
|
|
14
15
|
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "profiler-md",
|
|
3
|
-
"version": "0.9.
|
|
3
|
+
"version": "0.9.1",
|
|
4
4
|
"author": {
|
|
5
5
|
"name": "Tomer Aberbach",
|
|
6
6
|
"email": "tomer@aberba.ch",
|
|
@@ -45,6 +45,7 @@
|
|
|
45
45
|
"files": [
|
|
46
46
|
"dist",
|
|
47
47
|
"docs",
|
|
48
|
+
"!docs/CLAUDE.md",
|
|
48
49
|
"skills"
|
|
49
50
|
],
|
|
50
51
|
"type": "module",
|
package/docs/CLAUDE.md
DELETED
|
@@ -1,22 +0,0 @@
|
|
|
1
|
-
# Docs
|
|
2
|
-
|
|
3
|
-
- NEVER state what the tool doesn't support (e.g. a rejected flag, an unread
|
|
4
|
-
record type, a skipped frame). File an issue for it instead. The error or
|
|
5
|
-
warning reports it at runtime, and a doc statement goes stale once the tool
|
|
6
|
-
supports it
|
|
7
|
-
|
|
8
|
-
## Language docs
|
|
9
|
-
|
|
10
|
-
- A top-level section is what the profile measures (e.g. `## CPU profiling`,
|
|
11
|
-
`## Memory profiling`), opening with what it records and what it is useful
|
|
12
|
-
for. A tool is never a top-level section: `## perf profiling` states what
|
|
13
|
-
wrote the profile, and `## CPU profiling` what it measured. The test: two
|
|
14
|
-
tools that sample the same thing by different mechanisms (e.g. a kernel
|
|
15
|
-
counter and an interval timer) share a section
|
|
16
|
-
- Where a section covers one tool, nest its `### CLI` and `### Programmatic API`
|
|
17
|
-
directly. Where it covers several, give each a `### <tool>` subsection and
|
|
18
|
-
nest `#### CLI` and `#### Programmatic API` under it
|
|
19
|
-
- Put a `#### Flags` or `#### Environment variables` table under the `### CLI`
|
|
20
|
-
that uses it. Where the tool has a `### <tool>` subsection instead, move its
|
|
21
|
-
table to a `## <tool> CLI flags` section after the last section that uses the
|
|
22
|
-
tool
|