feat(preprod): Add VCS parameters to snapshots upload command (#3200)

### Description
Adds vcs parameters to preprod snapshots upload command. Shares
utilities with existing build upload command.

### Issues
<!--
* resolves: #1234
* resolves: LIN-1234
-->

<!--
#### Reminders
- Add GH Issue ID _&_ Linear ID (if applicable)
- PR title should use [conventional
commit](https://develop.sentry.dev/engineering-practices/commit-messages/#type)
style (`feat:`, `fix:`, `ref:`, `meta:`)
- For external contributors:
[CONTRIBUTING.md](https://github.com/getsentry/sentry-cli/blob/master/CONTRIBUTING.md),
[Sentry SDK development docs](https://develop.sentry.dev/sdk/), [Discord
community](https://discord.gg/Ww9hbqr)
-->

### Legal Boilerplate
Look, I get it. The entity doing business as "Sentry" was incorporated
in the State of Delaware in 2015 as Functional Software, Inc. and is
gonna need some rights from me in order to utilize my contributions in
this here PR. So here's the deal: I retain all rights, title and
interest in and to my contributions, and by keeping this boilerplate
intact I confirm that Sentry can use, modify, copy, and redistribute my
contributions, under Sentry's choice of terms.

Author: rbro112

src/api/data_types/snapshots.rs

@@ -5,6 +5,8 @@ use std::collections::HashMap;
 use serde::{Deserialize, Serialize};
 use serde_json::Value;
 
+use super::VcsInfo;
+
 const IMAGE_FILE_NAME_FIELD: &str = "image_file_name";
 const WIDTH_FIELD: &str = "width";
 const HEIGHT_FIELD: &str = "height";
@@ -21,9 +23,11 @@ pub struct CreateSnapshotResponse {
 // Keep in sync with https://github.com/getsentry/sentry/blob/master/src/sentry/preprod/snapshots/manifest.py
 /// Manifest describing a set of snapshot images for an app.
 #[derive(Debug, Serialize)]
-pub struct SnapshotsManifest {
+pub struct SnapshotsManifest<'a> {
     pub app_id: String,
     pub images: HashMap<String, ImageMetadata>,
+    #[serde(flatten)]
+    pub vcs_info: VcsInfo<'a>,
 }
 
 // Keep in sync with https://github.com/getsentry/sentry/blob/master/src/sentry/preprod/snapshots/manifest.py

src/commands/build/snapshots.rs

@@ -18,6 +18,8 @@ use walkdir::WalkDir;
 use crate::api::{Api, CreateSnapshotResponse, ImageMetadata, SnapshotsManifest};
 use crate::config::{Auth, Config};
 use crate::utils::args::ArgExt as _;
+use crate::utils::build_vcs::collect_git_metadata;
+use crate::utils::ci::is_ci;
 
 const EXPERIMENTAL_WARNING: &str =
     "[EXPERIMENTAL] The \"build snapshots\" command is experimental. \
@@ -47,6 +49,7 @@ pub fn make_command(command: Command) -> Command {
                 .help("The application identifier.")
                 .required(true),
         )
+        .git_metadata_args()
 }
 
 struct ImageInfo {
@@ -80,6 +83,13 @@ pub fn execute(matches: &ArgMatches) -> Result<()> {
         anyhow::bail!("Path is not a directory: {}", dir_path.display());
     }
 
+    // Collect git metadata if running in CI, unless explicitly enabled or disabled.
+    let should_collect_git_metadata =
+        matches.get_flag("force_git_metadata") || (!matches.get_flag("no_git_metadata") && is_ci());
+
+    // Always collect git metadata, but only perform automatic inference when enabled
+    let vcs_info = collect_git_metadata(matches, &config, should_collect_git_metadata);
+
     debug!("Scanning for images in: {}", dir_path.display());
     debug!("Organization: {org}");
     debug!("Project: {project}");
@@ -114,6 +124,7 @@ pub fn execute(matches: &ArgMatches) -> Result<()> {
     let manifest = SnapshotsManifest {
         app_id: app_id.clone(),
         images: manifest_entries,
+        vcs_info,
     };
 
     // POST manifest to API

src/commands/build/upload.rs

@@ -1,4 +1,3 @@
-use std::borrow::Cow;
 use std::io::Write as _;
 use std::path::Path;
 use std::thread;
@@ -8,7 +7,6 @@ use anyhow::{anyhow, bail, Context as _, Result};
 use clap::{Arg, ArgAction, ArgMatches, Command};
 use indicatif::ProgressStyle;
 use log::{debug, info, warn};
-use sha1_smol::Digest;
 use symbolic::common::ByteView;
 use zip::write::SimpleFileOptions;
 use zip::{DateTime, ZipWriter};
@@ -22,17 +20,13 @@ use crate::utils::build::{handle_asset_catalogs, ipa_to_xcarchive, is_apple_app,
 use crate::utils::build::{
     is_aab_file, is_apk_file, is_zip_file, normalize_directory, write_version_metadata,
 };
+use crate::utils::build_vcs::collect_git_metadata;
 use crate::utils::chunks::{upload_chunks, Chunk, ASSEMBLE_POLL_INTERVAL};
 use crate::utils::ci::is_ci;
 use crate::utils::fs::get_sha1_checksums;
 use crate::utils::fs::TempDir;
 use crate::utils::fs::TempFile;
 use crate::utils::progress::ProgressBar;
-use crate::utils::vcs::{
-    self, get_github_base_ref, get_github_head_ref, get_github_pr_number, get_provider_from_remote,
-    get_repo_from_remote_preserve_case, git_repo_base_ref, git_repo_base_repo_name_preserve_case,
-    git_repo_head_ref, git_repo_remote_url,
-};
 
 pub fn make_command(command: Command) -> Command {
     #[cfg(all(target_os = "macos", target_arch = "aarch64"))]
@@ -54,51 +48,7 @@ pub fn make_command(command: Command) -> Command {
                 .action(ArgAction::Append)
                 .required(true),
         )
-        .arg(
-            Arg::new("head_sha")
-                .long("head-sha")
-                .value_parser(parse_sha_allow_empty)
-                .help("The VCS commit sha to use for the upload. If not provided, the current commit sha will be used.")
-        )
-        .arg(
-            Arg::new("base_sha")
-                .long("base-sha")
-                .value_parser(parse_sha_allow_empty)
-                .help("The VCS commit's base sha to use for the upload. If not provided, the merge-base of the current and remote branch will be used.")
-        )
-        .arg(
-            Arg::new("vcs_provider")
-                .long("vcs-provider")
-                .help("The VCS provider to use for the upload. If not provided, the current provider will be used.")
-        )
-        .arg(
-            Arg::new("head_repo_name")
-                .long("head-repo-name")
-                .help("The name of the git repository to use for the upload (e.g. organization/repository). If not provided, the current repository will be used.")
-        )
-        .arg(
-            Arg::new("base_repo_name")
-                .long("base-repo-name")
-                .help("The name of the git repository to use for the upload (e.g. organization/repository). If not provided, the current repository will be used.")
-        )
-        .arg(
-            Arg::new("head_ref")
-                .long("head-ref")
-                .help("The reference (branch) to use for the upload. If not provided, the current reference will be used.")
-        )
-        .arg(
-            Arg::new("base_ref")
-                .long("base-ref")
-                .help("The base reference (branch) to use for the upload. If not provided, the merge-base with the remote tracking branch will be used.")
-        )
-        .arg(
-            Arg::new("pr_number")
-                .long("pr-number")
-                .value_parser(clap::value_parser!(u32))
-                .help("The pull request number to use for the upload. If not provided and running \
-                    in a pull_request-triggered GitHub Actions workflow, the PR number will be automatically \
-                    detected from GitHub Actions environment variables.")
-        )
+        .git_metadata_args()
         .arg(
             Arg::new("build_configuration")
                 .long("build-configuration")
@@ -119,22 +69,6 @@ pub fn make_command(command: Command) -> Command {
                     for each other.",
                 )
         )
-        .arg(
-            Arg::new("force_git_metadata")
-                .long("force-git-metadata")
-                .action(ArgAction::SetTrue)
-                .conflicts_with("no_git_metadata")
-                .help("Force collection and sending of git metadata (branch, commit, etc.). \
-                    If neither this nor --no-git-metadata is specified, git metadata is \
-                    automatically collected when running in most CI environments.")
-        )
-        .arg(
-            Arg::new("no_git_metadata")
-                .long("no-git-metadata")
-                .action(ArgAction::SetTrue)
-                .conflicts_with("force_git_metadata")
-                .help("Disable collection and sending of git metadata.")
-        )
 }
 
 /// Parse plugin info from SENTRY_PIPELINE environment variable.
@@ -309,181 +243,6 @@ pub fn execute(matches: &ArgMatches) -> Result<()> {
     Ok(())
 }
 
-/// Collects git metadata from arguments and VCS introspection.
-///
-/// When `auto_collect` is false, only explicitly provided values are collected;
-/// automatic inference from git repository and CI environment is skipped.
-fn collect_git_metadata(
-    matches: &ArgMatches,
-    config: &Config,
-    auto_collect: bool,
-) -> VcsInfo<'static> {
-    let head_sha = matches
-        .get_one::<Option<Digest>>("head_sha")
-        .map(|d| d.as_ref().cloned())
-        .or_else(|| auto_collect.then(|| vcs::find_head_sha().ok()))
-        .flatten();
-
-    let cached_remote = config.get_cached_vcs_remote();
-    let (vcs_provider, head_repo_name, head_ref, base_ref, base_repo_name) = {
-        let repo = if auto_collect {
-            git2::Repository::open_from_env().ok()
-        } else {
-            None
-        };
-        let repo_ref = repo.as_ref();
-        let remote_url = repo_ref.and_then(|repo| git_repo_remote_url(repo, &cached_remote).ok());
-
-        let vcs_provider = matches
-            .get_one("vcs_provider")
-            .cloned()
-            .or_else(|| {
-                auto_collect
-                    .then(|| remote_url.as_ref().map(|url| get_provider_from_remote(url)))?
-            })
-            .unwrap_or_default();
-
-        let head_repo_name = matches
-            .get_one("head_repo_name")
-            .cloned()
-            .or_else(|| {
-                auto_collect.then(|| {
-                    remote_url
-                        .as_ref()
-                        .map(|url| get_repo_from_remote_preserve_case(url))
-                })?
-            })
-            .unwrap_or_default();
-
-        let head_ref = matches
-            .get_one("head_ref")
-            .cloned()
-            .or_else(|| auto_collect.then(get_github_head_ref)?)
-            .or_else(|| {
-                auto_collect.then(|| {
-                    repo_ref.and_then(|r| match git_repo_head_ref(r) {
-                        Ok(ref_name) => {
-                            debug!("Found current branch reference: {ref_name}");
-                            Some(ref_name)
-                        }
-                        Err(e) => {
-                            debug!("No valid branch reference found (likely detached HEAD): {e}");
-                            None
-                        }
-                    })
-                })?
-            })
-            .unwrap_or_default();
-
-        let base_ref = matches
-            .get_one("base_ref")
-            .cloned()
-            .or_else(|| auto_collect.then(get_github_base_ref)?)
-            .or_else(|| {
-                auto_collect.then(|| {
-                    repo_ref.and_then(|r| match git_repo_base_ref(r, &cached_remote) {
-                        Ok(base_ref_name) => {
-                            debug!("Found base reference: {base_ref_name}");
-                            Some(base_ref_name)
-                        }
-                        Err(e) => {
-                            info!("Could not detect base branch reference: {e}");
-                            None
-                        }
-                    })
-                })?
-            })
-            .unwrap_or_default();
-
-        let base_repo_name = matches
-            .get_one("base_repo_name")
-            .cloned()
-            .or_else(|| {
-                auto_collect.then(|| {
-                    repo_ref.and_then(|r| match git_repo_base_repo_name_preserve_case(r) {
-                        Ok(Some(base_repo_name)) => {
-                            debug!("Found base repository name: {base_repo_name}");
-                            Some(base_repo_name)
-                        }
-                        Ok(None) => {
-                            debug!("No base repository found - not a fork");
-                            None
-                        }
-                        Err(e) => {
-                            warn!("Could not detect base repository name: {e}");
-                            None
-                        }
-                    })
-                })?
-            })
-            .unwrap_or_default();
-
-        (
-            vcs_provider,
-            head_repo_name,
-            head_ref,
-            base_ref,
-            base_repo_name,
-        )
-    };
-
-    let base_sha_from_user = matches.get_one::<Option<Digest>>("base_sha").is_some();
-    let base_ref_from_user = matches.get_one::<String>("base_ref").is_some();
-
-    let mut base_sha = matches
-        .get_one::<Option<Digest>>("base_sha")
-        .map(|d| d.as_ref().cloned())
-        .or_else(|| {
-            if auto_collect {
-                Some(
-                    vcs::find_base_sha(&cached_remote)
-                        .inspect_err(|e| debug!("Error finding base SHA: {e}"))
-                        .ok()
-                        .flatten(),
-                )
-            } else {
-                None
-            }
-        })
-        .flatten();
-
-    let mut base_ref = base_ref;
-
-    // If base_sha equals head_sha and both were auto-inferred, skip setting base_sha and base_ref
-    if !base_sha_from_user
-        && !base_ref_from_user
-        && base_sha.is_some()
-        && head_sha.is_some()
-        && base_sha == head_sha
-    {
-        debug!(
-            "Base SHA equals head SHA ({}), and both were auto-inferred. Skipping base_sha and base_ref, but keeping head_sha.",
-            base_sha.expect("base_sha is Some at this point")
-        );
-        base_sha = None;
-        base_ref = "".into();
-    }
-
-    let pr_number = matches.get_one("pr_number").copied().or_else(|| {
-        if auto_collect {
-            get_github_pr_number()
-        } else {
-            None
-        }
-    });
-
-    VcsInfo {
-        head_sha,
-        base_sha,
-        vcs_provider: Cow::Owned(vcs_provider),
-        head_repo_name: Cow::Owned(head_repo_name),
-        base_repo_name: Cow::Owned(base_repo_name),
-        head_ref: Cow::Owned(head_ref),
-        base_ref: Cow::Owned(base_ref),
-        pr_number,
-    }
-}
-
 fn handle_file(
     path: &Path,
     byteview: &ByteView,
@@ -709,22 +468,6 @@ fn upload_file(
     }
 }
 
-/// Utility function to parse a SHA1 digest, allowing empty strings.
-///
-/// Empty strings result in Ok(None), otherwise we return the parsed digest
-/// or an error if the SHA is invalid.
-fn parse_sha_allow_empty(sha: &str) -> Result<Option<Digest>> {
-    if sha.is_empty() {
-        return Ok(None);
-    }
-
-    let digest = sha
-        .parse()
-        .with_context(|| format!("{sha} is not a valid SHA1 digest"))?;
-
-    Ok(Some(digest))
-}
-
 #[cfg(not(windows))]
 #[cfg(test)]
 mod tests {