| Age | Commit message (Collapse) | Author |
|---|
|
Collect all errors into a list. I think I'm going to return them as a
list from this function.
The runtime appears a lot slower with this change. Need to figure out
what that's about.
|
|
Use the Tokio runtime we created to run the blocking async tasks.
Trying to set this up so I can get results back from the spawned tasks,
but I'm currently having trouble working out how to extract them from
the async task and return them from `run()`. I suppose I could just
print out the errors directly in that `while let` loop, but ideally I'd
like to return all errors from `run()` rather than printing in `run()`.
|
|
To separate the actions more.
|
|
|
|
Remove the hard-coded test repositories I was using and replace them
with real ones retrieved from the GitHub API.
Enable I/O and timers on the Tokio runtime in order to enable the async
GitHub API request.
|
|
Don't panic here so we can use our own error message template.
|
|
Print the error instead of unwrapping.
|
|
|
|
Don't see any reason to do this now.
|
|
|
|
Allows a maximum repo size to be given as a command line argument.
Repos larger than this will not be mirrored. This gives us a way to save
server space by avoiding gigantic repositories.
|
|
Remove this now that we have something that I think works.
|
|
Rather that relying on the Cargo features we've enabled to define this,
create a multi-threaded runtime in code.
|
|
I think, at least. Took a lot of research and trial and error to get
this to compile, working out how to set up the multi-threading for async
code. The idea here is to be able to process each repo in potentially
multiple threads and do that processing work in parallel.
|
|
|
|
|
|
This is no longer relevant.
|
|
Add command line argument value here in preparation for when we enable
this code.
|
|
|
|
|
|
|
|
Define the options we want to take. Not using them yet.
|
|
A repository cloned with:
$ git clone --mirror REPO
doesn't have any ref files in `repo.git/refs/heads/*`. Instead, the refs
are stored in `repo.git/packed-refs`. Update the pack file if the
default branch ref file doesn't exist. CGit will look at the time on the
'packed-refs' file when that's the case.
|
|
The Angular.js repo was 51 MB, while DDHotKey is 95 K.
|
|
|
|
CGit reads the repository modification time from the following
locations, in order from top to bottom:
1. agefile
2. repo.git/refs/heads/{default_branch | "master"}
3. repo.git/packed-refs
(https://git.zx2c4.com/cgit/tree/ui-repolist.c?id=bd6f5683f6cde4212364354b3139c1d521f40f39#n35)
Update the `/refs/heads/{default_branch}` file mtime when cloning and
updating the repo to match the GitHub `updated_at` time.
This ensures that when mirroring old repositories, they don't appear at
the top of the CGit repository index list when sorting by age.
|
|
|
|
|
|
|
|
Also switch from `reqwest::blocking` to async because I was getting this
error, probably because I call `fetch_repos()` in the async 'tokio'
function `main()`:
thread 'main' panicked at 'Cannot drop a runtime in a context where
blocking is not allowed. This happens when a runtime is dropped from
within an asynchronous context.',
$HOME/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.6.1/src/runtime/blocking/shutdown.rs:51:21
|
|
We'll call this from `main()` when things are more ready.
|
|
This is all done now.
|
|
Allow user to not be required to specify a base cgitrc file for cloned
repositories.
|
|
This lets us define common cgitrc configuration for all mirrored repos.
|
|
Check the repository description that comes back from the GitHub API
against our cached description in the database. Only write the new
description if it changed so we can avoid writing to the file in that
case.
|
|
Forgot to add commas when I added the additional fields.
|
|
This is simpler, and means we don't have to check if the database file
exists and only initialise if it doesn't. Here, we can just run the code
and trust it will do the right thing in both cases.
|
|
Store the repository description so we can check whether the description
was updated and copy it in our mirror accordingly.
|
|
|
|
If the repository was updated, write the description into the
`description` file.
Add a `github::Repo.description()` method to get an empty string if the
description is `None`. This facilitates writing to the `description`
file.
|
|
Looks like I'm not going to be using this, since the functions in this
match arm that take `database::Repo`s should take the one based on the
`github::Repo` rather than the one fetched from the database.
|
|
This gets rid of the default description string:
Unnamed repository; edit this file 'description' to name the repository.
|
|
|
|
Separate source and fork repositories into different paths.
|
|
|
|
I wanted to move `clone_path` into `mirror()` so I could make a
different clone path for forks.
|
|
Copy the repository description from GitHub into the clone repo.
|
|
If we haven't encountered a repository yet, mirror it to the filesystem.
Change `From<github::Repo>` to `From<&github::Repo>` so we don't consume
the repo and can use it later in the repo list loop.
|
|
|
|
|
