Summary:
- rename the checker "Uninitialized Variable" to "Uninitialized Value"
as this is the name of the issue type
- delete timestamp XML comment from the man pages to avoid future git
churn when updating the website
- counting is hard
Reviewed By: martintrojer
Differential Revision: D24219165
fbshipit-source-id: cf3057373
Summary:
This is required for versioning to work correctly when creating a new
version of the website.
Reviewed By: martintrojer
Differential Revision: D24217125
fbshipit-source-id: 8f7a4bc44
Summary:
We forgot to add this when adding the new subcommand. These should
really be auto-generated but let's save that for another diff.
Reviewed By: martintrojer
Differential Revision: D24217111
fbshipit-source-id: 860860533
Summary:
- freshen up /docs/next/absint-framework to give sensible advice, and
delete outdated bits that are now in the API docs so they remain fresh
- delete SimpleChecker.ml as it's just a source of bitrot
- delete the "adding checkers" page as it's completely outdated and
subsumed by the "AI framework" page + the labs.
Reviewed By: jberdine
Differential Revision: D23597271
fbshipit-source-id: 78b541746
Summary:
Better API for creating issue types:
- distinguish hidden/normal/dynamic issue types
- normal issue types should always be documented
- add "TODO" to missing documentation
- dynamic issue types are the only ones that can be created outside of
IssueType.ml
I had to document the new CCBM and the resource leak lab exercise to
keep Help.ml happy, did `make doc-publish`.
Reviewed By: ngorogiannis
Differential Revision: D22118766
fbshipit-source-id: 3d0194518
Summary:
A bug in docusaurus makes relative URLs fail depending on how the page
was accessed, because the URL of a page in docs/ will end in / if
accessed directly or via hyperlink, but that / will be omitted when
clicking on the sidebar. The final / makes all the difference when
interpreting relative URLs so relative URLs are essentially broken.
See https://github.com/facebook/docusaurus/issues/2832 for more details.
This changes URL generation to generate URLs /docs/next/..., and
manually substitute relative URLs that had been written by hand.
Also fix a few other things about outdated links/comments.
Finally, `make doc-publish`.
Reviewed By: dulmarod
Differential Revision: D22117187
fbshipit-source-id: 32e2ba7e1
Summary:
Needed to remove user_documentation for the new
CONFIG_CHECK_BETWEEN_MARKERS issue type otherwise it violated the
invariant that the corresponding checker should be documented too but
its development has just started.
Reviewed By: skcho
Differential Revision: D22065820
fbshipit-source-id: 4b3a58850
Summary:
alicelieutier graciously contributed this React component to savagely
include the HTML from the man pages into docusaurus. Now they render
like the rest of the documentation!
Reviewed By: cristianoc
Differential Revision: D22050580
fbshipit-source-id: e8e314109
Summary:
Hi,
I have found a couple of small issues while reading the docs. The pull request contains the fixes.
BTW, the link to PLDI slices is also broken here: https://fbinfer.com/docs/absint-framework. I haven't found the slides anywhere, so not sure how to fix it.
Pull Request resolved: https://github.com/facebook/infer/pull/1278
Reviewed By: dulmarod
Differential Revision: D22018947
Pulled By: jvillard
fbshipit-source-id: 13f6d27b1
Summary:
The page itself was already there but not linked from the rest of the
website.
Reviewed By: ezgicicek
Differential Revision: D21997950
fbshipit-source-id: 30380da65
Summary:
Delete the following outdated pages. Some of them have better
replacements now but some of them are just too out of date to easily
salvage.
- 01-adding-models.md: this is biabduction-specific and hard to make
that distinction. Most checkers have no way to add models. Even
suggesting that people add their own biabduction models was a bit
awkward as it requires recompiling infer to use.
- 01-checkers.md: outdated document that describes "new" checkers like
Quandary :)
- 01-experimental-checkers.md: very outdated!
- 02-limitations.md: also outdated: it roughly says "we don't do concurrency or
arithmetic, the former is basically impossible but the latter should
be easy"... well, 5 years later this hasn't aged too well ;)
- 01-eradicate.md: content has moved to auto-generated documentation
- 01-linters.md: content has moved to auto-generated documentation
- 01-racerd.md: content has moved to auto-generated documentation
Reviewed By: mityal
Differential Revision: D21934748
fbshipit-source-id: a933e5022
Summary:
New website! Now with one page per checker, and only one page containing
all issue types for easy tooling. Each checker page also lists the
issue types it can report.
For now only issue types with documentation appear in either page,
filling these up is still TODO.
Reviewed By: mityal
Differential Revision: D21934465
fbshipit-source-id: 82ae1e417
Summary:
Take all the issue type documentation in the website and add it to infer
itself so it can generate the website (see further diffs).
I mostly generated this diff by writing a script to call from `utop`
with various file lists to do most of the heavy lifting for me and make
sure I didn't forget any issue types: P132800781
Reviewed By: ngorogiannis
Differential Revision: D21934372
fbshipit-source-id: f3ea8c566
Summary:
Add a new section in the docs, creatively named "Versions", with links
to the 0.17.0 and "next" version. Staying vague on what "next" points to
because I don't know how often we'll be able to update it yet.
Reviewed By: ngorogiannis
Differential Revision: D21573699
fbshipit-source-id: fc527cf61
Summary:
Run `make doc-publish` and adjust links in
website/docs/04-internal-API.md to point to odoc/next/infer/infer.html as the
starting point of the docs has changed (was: odoc/index.html but now
this also contains dead links to sledge and ppx_trace... also infer.mld
is a me-written exposé of the various libraries and index.mld is an
auto-generated list of all the modules now).
Reviewed By: ngorogiannis
Differential Revision: D21552864
fbshipit-source-id: c3977f645
Summary:
The plan is to keep only the man pages and OCaml doc for:
- the latest published version (0.17.0 currently)
- master-ish
Next commit will introduce versioning of the documentation.
Reviewed By: dulmarod
Differential Revision: D21526342
fbshipit-source-id: 6a49d168e
Summary:
This didn't work in prod, testing with `yarn build` also reveals the
problem locally.
Reviewed By: skcho
Differential Revision: D21522971
fbshipit-source-id: 6b6bd322f
Summary:
- Specify a "position" for the docs/support/blog entries in the top bar.
They had disappeared! This puts them back for some reason.
- Add "Facebook" to the "Social" list in the footer (it's a social
network, right?)
- remove executable permission on some text files (hmmmm)
- rename .md.md file to just .md (this also fixes a dead link)
- update yarn.lock (happened automatically when I tested the website
locally)
Reviewed By: ezgicicek
Differential Revision: D21504931
fbshipit-source-id: 5f89935e2
Summary:
Signed-off-by: Andrew V. Jones <andrew.jones@vector.com>
## Issue
Commit [068622c](https://github.com/facebook/infer/commit/068622c) renamed the option `test` to be `only-cheap-debug` and also flipped the default value. Therefore, what used to be `--no-test` (i.e., you _do_ want expensive debug information) is now `--no-only-cheap-debug`.
However, [`01-advanced-features.md`](https://github.com/facebook/infer/blob/master/website/docs/01-advanced-features.md) listed the option as `no_test` (with an underscore rather than a hyphen), which I guess got missed when renaming the option/changing the default.
The suggestion to use `--no_test` was recently suggested in [https://github.com/facebook/infer/issues/1252](https://github.com/facebook/infer/issues/1252#issuecomment-618922148), so we should probably remove this, otherwise you get (even if you use the "correct" syntax for an older Infer):
```
$ infer --no-test
/path/to/infer/bin/infer: unknown option '--no-test'.
```
## Description of changes
Given that `only-cheap-debug` now has a default of _false_ (i.e., if you request `-g`/`--debug`, you automatically get `--no-only-cheap-debug`), making the suggestion to users to turn _off_ "no expensive debugging" is not required.
Consequently, this PR removes the recommendation to pass `--no-test`/`--no-only-cheap-debug` in the documentation.
Pull Request resolved: https://github.com/facebook/infer/pull/1253
Reviewed By: ezgicicek
Differential Revision: D21424108
Pulled By: jvillard
fbshipit-source-id: 2b4623380
Summary:
Seems like a more sensible name. Most tooling should read report.json so
won't notice.
Still output a bugs.txt file with a message to point to report.txt while
people migrate.
Reviewed By: mityal, artempyanykh
Differential Revision: D20626111
fbshipit-source-id: efb84d098
Summary:
Fixes#1234
Turns out I didn't catch this in local testing because docusaurus falls back on
exploring files in website/ locally for some reason.
Reviewed By: skcho
Differential Revision: D20342513
fbshipit-source-id: 87587ccc2
Summary:
It's a lot of code to maintain for something that no one ever uses
anymore.
Reviewed By: ngorogiannis
Differential Revision: D20282794
fbshipit-source-id: 28422c415
Jules Villard
Dulma Churchill
Marat Khabibullin
Andrew V. Jones
BSKY
