1# Adding a directory to catapult
2
3## Where should the code live?
4
5Catapult is intended to be a set of performance tools, mostly based on tracing,
6for developers of Chromium and other software to analyze that software’s
7performance. It has a lot of supporting libraries and tooling to make this
8happen. We’d like to create an organizational structure that makes it easy to
9find the performance tooling developers want, and hard to accidentally depend
10on something internal. Furthermore, we’d like to make it easy for code in
11catapult to eventually grow to its own repo, when possible. To that end, we
12use these guidelines to decide where code should live.
13
14  * Is it a **performance product**? Meant be used by external developers for
15    performance analysis? Some examples include telemetry, perf dashboard. If
16    so, it should be at `toplevel`.
17  * Is it used only by our buildbot or build process? Put it in
18    `catapult_build`.
19  * If it's neither of the above, it should go in `common/`
20  * If it is aspiring to be its own repo someday, that doesn't affect where it
21    goes. You should follow the above rules for directory placement. Third
22    party must only be real third party repos to conform to rules of repos
23    which include catapult.
24  * If something is experimental, then talk with the catapult admins to build
25    the best guess of where it should go.
26
27## How should directories be structured?
28We have some rules on directory structure to add consistency and avoid
29overloaded python imports.
30
31  * Toplevel directories are **not** modules. E.g. if `x` is a toplevel
32    directory, `x/__init__.py` **does not** exist. Directories in `common/`
33    do not have this restriction.
34  * Toplevel directories and directories in `common` should centralize all
35    their path computation and sys.path manipulation in their master init file
36    ([example](https://github.com/catapult-project/catapult/blob/master/telemetry/telemetry/__init__.py)).
37  * Projects using web server should provide a module which defines all the
38    search paths to their html & javascript resources in their top directory
39    ([example](https://github.com/catapult-project/catapult/blob/master/dashboard/dashboard_project.py)).
40  * Build code should be separate from production code. Build scripts for
41    projects should be in `x/x_build/`
42  * If you have a feature that has an implementation in JS and Py, then it
43    should be in the same folder.
44  * HTML search paths are arranged such that they have the same _name_ as they
45    would in python. E.g. `tracing/tracing/base/math.html` is
46    `/tracing/base/math.html` for an HTML import, and
47    `tracing/tracing/base/math.py` is `tracing.base.math` in python.
48  * Executable files (e.g. files chmodded to +x) must live in `x/bin`.
49    `bin/` must not be a module, e.g. contain `__init__.py`, as such a name
50    would create namespace conflicts. Executable files should **not** have a
51    `.py` extension.
52  * We use a single dev server, `$catapult/bin/run_dev_server`; and have
53    per-project `bin/run_dev_server_tests` scripts.
54  * All python modules should have unique names. `$catpult/catapult_build`
55    instead of `$catapult/build`.
56