buffer: fix indexOf for empty searches #13024

addaleax · 2017-05-14T18:00:45Z

Make searches for empty subsequences return 0 for buffer.indexOf()
and buffer.length for buffer.lastIndexOf(), because those
are the indices of the first and last occurrence of an empty
subsequence, respectively.

Make searches for empty subsequences do exactly what String.prototype.indexOf() does.

Fixes: #13023

Checklist

make -j4 test (UNIX), or vcbuild test (Windows) passes
tests and/or benchmarks are included
commit message follows commit guidelines

Affected core subsystem(s)

buffer

/cc @nodejs/buffer and because people might to consider this a semver-major change (I would not) @nodejs/ctc

CI: https://ci.nodejs.org/job/node-test-commit/9880/

Trott · 2017-05-14T18:49:26Z

Q: semver-major?

Trott · 2017-05-14T18:54:54Z

Documentation update to include this info? MDN mentions the behavior in its docs for String.prototype.lastIndexOf() and String.prototype.indexOf() and we should probably follow suit.

Trott · 2017-05-14T18:56:23Z

test/parallel/test-buffer-indexof.js

-assert.strictEqual(b.indexOf('', b.length + 1), -1);
-assert.strictEqual(b.indexOf('', Infinity), -1);
+assert.strictEqual(b.indexOf(''), 0);
+assert.strictEqual(b.indexOf('', 1), 0);


To match the behavior of String.prototype.indexOf(), this should return 1 rather than 0.

Trott · 2017-05-14T18:57:07Z

test/parallel/test-buffer-indexof.js

-assert.strictEqual(b.indexOf('', Infinity), -1);
+assert.strictEqual(b.indexOf(''), 0);
+assert.strictEqual(b.indexOf('', 1), 0);
+assert.strictEqual(b.indexOf('', b.length + 1), 0);


To match the behavior of String.prototype.indexOf(), this should return b.length, not 0.

Trott · 2017-05-14T18:57:42Z

test/parallel/test-buffer-indexof.js

+assert.strictEqual(b.indexOf(''), 0);
+assert.strictEqual(b.indexOf('', 1), 0);
+assert.strictEqual(b.indexOf('', b.length + 1), 0);
+assert.strictEqual(b.indexOf('', Infinity), 0);


To match the behavior of String.prototype.indexOf(), this should return b.length, not 0.

addaleax · 2017-05-14T19:44:34Z

@Trott updated!

New CI: https://ci.nodejs.org/job/node-test-commit/9882/

Trott · 2017-05-14T19:50:28Z

doc/api/buffer.md

@@ -1340,6 +1340,10 @@ console.log(b.indexOf('b', null));
 console.log(b.indexOf('b', []));
 ```

+If `value` is an empty string or `Buffer` and `byteOffset` is less


Nit: empty string or empty Buffer would avoid confusion over whether or not the Buffer must be empty.

Might be even simpler to replace is an empty string or Buffer with is empty or has a length of zero?

If value is empty and byteOffset is less

or

If value has a length of zero and byteOffset is less

Yea, makes sense. I went with the former because is empty sounds a bid odd to me (like empty/undefined variables?), and has a length of zero doesn’t make sense for integer input (and users can figure that out, but, you know), so I chose the imho explicit version.

Trott · 2017-05-14T19:51:03Z

doc/api/buffer.md

@@ -1450,6 +1454,8 @@ console.log(b.lastIndexOf('b', null));
 console.log(b.lastIndexOf('b', []));
 ```

+If `value` is an empty string or `Buffer`, `byteOffset` will be returned.


Same as above comment.

Make searches for empty subsequences do exactly what `String.prototype.indexOf()` does. Fixes: nodejs#13023

Trott

LGTM to me if CI is green and doc nits are addressed. Not sure about the semver-ness here, but maybe CITGM run results would help make the case for patch vs. major?

addaleax · 2017-05-14T19:57:10Z

That’s fair:

CITGM: https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/784/
CITGM for master: https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/783/

refack · 2017-05-15T16:09:41Z

Assuming previous Windows fails was infra. Rerunning: https://ci.nodejs.org/job/node-test-commit-windows-fanned/9084/

refack · 2017-05-15T16:23:59Z

src/node_buffer.cc

      // Valid positive offset.
      return offset_i64;
+    } else if (needle_length == 0) {
+      // Out of buffer bounds, but empty needle: point to end of buffer.


Suggestion:
Either ref - http://www.ecma-international.org/ecma-262/6.0/#sec-string.prototype.indexof
Or state Same semantics as String.prototype.indexof

@refack Ack, added comments (at the GetReturnValue() sites)

👍
I was feeling it missing here (you explain "what" but not "why")

addaleax · 2017-05-18T21:20:30Z

Landed in 28ddac2

Make searches for empty subsequences do exactly what `String.prototype.indexOf()` does. Fixes: nodejs#13023 PR-URL: nodejs#13024 Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Rich Trott <rtrott@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>

MylesBorins · 2017-07-17T14:31:17Z

@addaleax is this applicable for v6.x?

MylesBorins · 2017-08-14T18:05:50Z

I'm going to opt to not land this on lts as it does change behavior of buffer. While it might be "more correct", I'd rather not pull the rug under people

/cc @addaleax @nodejs/ctc

addaleax · 2017-08-14T18:07:36Z

I’m okay with that.

mcollina · 2017-08-16T06:36:08Z

I'm ok as well.

nodejs-github-bot added buffer Issues and PRs related to the buffer subsystem. c++ Issues and PRs that require attention from people who are familiar with C++. labels May 14, 2017

Trott reviewed May 14, 2017

View reviewed changes

addaleax force-pushed the buffer-indexof-emptystring branch 2 times, most recently from ce70e2d to 8bcee3f Compare May 14, 2017 19:38

Trott reviewed May 14, 2017

View reviewed changes

buffer: fix indexOf for empty searches

bb875a5

Make searches for empty subsequences do exactly what `String.prototype.indexOf()` does. Fixes: nodejs#13023

addaleax force-pushed the buffer-indexof-emptystring branch from 8bcee3f to bb875a5 Compare May 14, 2017 19:54

Trott approved these changes May 14, 2017

View reviewed changes

jasnell approved these changes May 14, 2017

View reviewed changes

cjihrig approved these changes May 15, 2017

View reviewed changes

refack approved these changes May 15, 2017

View reviewed changes

[squash] add comment about String#indexOf()

1393011

thefourtheye approved these changes May 16, 2017

View reviewed changes

addaleax closed this May 18, 2017

addaleax deleted the buffer-indexof-emptystring branch May 18, 2017 21:20

jasnell mentioned this pull request May 28, 2017

8.0.0 Release Proposal #12220

Closed

gibfahn mentioned this pull request Jun 15, 2017

Auditing for 6.11.1 nodejs/Release#230

Closed

3 tasks

MylesBorins added the lts-watch-v6.x label Jul 17, 2017

MylesBorins added dont-land-on-v6.x and removed lts-watch-v6.x labels Aug 14, 2017

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

buffer: fix indexOf for empty searches #13024

buffer: fix indexOf for empty searches #13024

addaleax commented May 14, 2017 •

edited

Loading

Trott commented May 14, 2017

Trott commented May 14, 2017

Trott May 14, 2017

Trott May 14, 2017

Trott May 14, 2017

addaleax commented May 14, 2017

Trott May 14, 2017

addaleax May 14, 2017

Trott May 14, 2017

Trott left a comment

addaleax commented May 14, 2017

refack commented May 15, 2017 •

edited

Loading

refack May 15, 2017

addaleax May 15, 2017

refack May 15, 2017

addaleax commented May 18, 2017

MylesBorins commented Jul 17, 2017

MylesBorins commented Aug 14, 2017

addaleax commented Aug 14, 2017

mcollina commented Aug 16, 2017

buffer: fix indexOf for empty searches #13024

buffer: fix indexOf for empty searches #13024

Conversation

addaleax commented May 14, 2017 • edited Loading

Checklist

Affected core subsystem(s)

Trott commented May 14, 2017

Trott commented May 14, 2017

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

addaleax commented May 14, 2017

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Trott left a comment

Choose a reason for hiding this comment

addaleax commented May 14, 2017

refack commented May 15, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

addaleax commented May 18, 2017

MylesBorins commented Jul 17, 2017

MylesBorins commented Aug 14, 2017

addaleax commented Aug 14, 2017

mcollina commented Aug 16, 2017

addaleax commented May 14, 2017 •

edited

Loading

refack commented May 15, 2017 •

edited

Loading