feat: ignore glob in sourcemap commands (#36366)

cli/Cargo.lock

@@ -194,6 +194,16 @@ dependencies = [
  "generic-array",
 ]
 
+[[package]]
+name = "bstr"
+version = "1.12.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "234113d19d0d7d613b40e86fb654acf958910802bcceab913a4f9e7cda03b1a4"
+dependencies = [
+ "memchr",
+ "serde",
+]
+
 [[package]]
 name = "bumpalo"
 version = "3.19.0"
@@ -719,6 +729,19 @@ version = "0.31.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "07e28edb80900c19c28f1072f2e8aeca7fa06b23cd4169cefe1af5aa3260783f"
 
+[[package]]
+name = "globset"
+version = "0.4.16"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "54a1028dfc5f5df5da8a56a73e6c153c9a9708ec57232470703592a3f18e49f5"
+dependencies = [
+ "aho-corasick",
+ "bstr",
+ "log",
+ "regex-automata 0.4.9",
+ "regex-syntax 0.8.5",
+]
+
 [[package]]
 name = "h2"
 version = "0.3.27"
@@ -1480,6 +1503,7 @@ dependencies = [
  "clap",
  "crossterm 0.28.1",
  "dirs",
+ "globset",
  "inquire",
  "magic_string",
  "miette",

cli/Cargo.toml

@@ -24,6 +24,7 @@ dirs = "6.0.0"
 inquire = "0.7.5"
 posthog-symbol-data = "0.1.0"
 walkdir = "2.5.0"
+globset = "0.4"
 ratatui = "0.29.0"
 crossterm = "0.28.1"
 tui-textarea = "0.7.0"

cli/src/commands/mod.rs

@@ -45,6 +45,10 @@ pub enum SourcemapCommand {
         /// The directory containing the bundled chunks
         #[arg(short, long)]
         directory: PathBuf,
+
+        /// One or more directory glob patterns to ignore
+        #[arg(short, long)]
+        ignore: Vec<String>,
     },
     /// Upload the bundled chunks to PostHog
     Upload {
@@ -52,6 +56,10 @@ pub enum SourcemapCommand {
         #[arg(short, long)]
         directory: PathBuf,
 
+        /// One or more directory glob patterns to ignore
+        #[arg(short, long)]
+        ignore: Vec<String>,
+
         /// The project name associated with the uploaded chunks. Required to have the uploaded chunks associated with
         /// a specific release, auto-discovered from git information on disk if not provided.
         #[arg(long)]
@@ -99,11 +107,12 @@ impl Cli {
                 login::login()?;
             }
             Commands::Sourcemap { cmd } => match cmd {
-                SourcemapCommand::Inject { directory } => {
-                    sourcemap::inject::inject(directory)?;
+                SourcemapCommand::Inject { directory, ignore } => {
+                    sourcemap::inject::inject(directory, ignore)?;
                 }
                 SourcemapCommand::Upload {
                     directory,
+                    ignore,
                     project,
                     version,
                     delete_after,
@@ -111,6 +120,7 @@ impl Cli {
                     sourcemap::upload::upload(
                         command.host,
                         directory,
+                        ignore,
                         project.clone(),
                         version.clone(),
                         *delete_after,

cli/src/commands/sourcemap/inject.rs

@@ -5,7 +5,7 @@ use uuid;
 
 use crate::utils::{posthog::capture_command_invoked, sourcemaps::read_pairs};
 
-pub fn inject(directory: &Path) -> Result<()> {
+pub fn inject(directory: &Path, ignore_globs: &[String]) -> Result<()> {
     let capture_handle = capture_command_invoked("sourcemap_inject", None::<&str>);
     let directory = directory.canonicalize().map_err(|e| {
         anyhow!(
@@ -15,7 +15,7 @@ pub fn inject(directory: &Path) -> Result<()> {
         )
     })?;
     info!("Processing directory: {}", directory.display());
-    let mut pairs = read_pairs(&directory)?;
+    let mut pairs = read_pairs(&directory, ignore_globs)?;
     if pairs.is_empty() {
         bail!("No source files found");
     }

cli/src/commands/sourcemap/upload.rs

@@ -47,6 +47,7 @@ struct BulkUploadFinishRequest {
 pub fn upload(
     host: Option<String>,
     directory: &PathBuf,
+    ignore_globs: &[String],
     project: Option<String>,
     version: Option<String>,
     delete_after: bool,
@@ -61,7 +62,7 @@ pub fn upload(
         host, token.env_id
     );
 
-    let pairs = read_pairs(directory)?;
+    let pairs = read_pairs(directory, ignore_globs)?;
     let sourcemap_paths = pairs
         .iter()
         .map(|pair| pair.sourcemap.path.clone())

cli/src/utils/sourcemaps.rs

@@ -1,5 +1,6 @@
 use anyhow::{anyhow, bail, Ok, Result};
 use core::str;
+use globset::{Glob, GlobSetBuilder};
 use magic_string::{GenerateDecodedMapOptions, MagicString};
 use posthog_symbol_data::{write_symbol_data, SourceAndMap};
 use serde::{Deserialize, Serialize};
@@ -152,16 +153,28 @@ impl SourcePair {
     }
 }
 
-pub fn read_pairs(directory: &PathBuf) -> Result<Vec<SourcePair>> {
+pub fn read_pairs(directory: &PathBuf, ignore_globs: &[String]) -> Result<Vec<SourcePair>> {
     // Make sure the directory exists
     if !directory.exists() {
         bail!("Directory does not exist");
     }
 
+    let mut builder = GlobSetBuilder::new();
+    for glob in ignore_globs {
+        builder.add(Glob::new(&glob)?);
+    }
+    let set: globset::GlobSet = builder.build()?;
+
     let mut pairs = Vec::new();
     for entry in WalkDir::new(directory).into_iter().filter_map(|e| e.ok()) {
         let entry_path = entry.path().canonicalize()?;
-        if is_javascript_file(&entry_path) {
+
+        if set.is_match(&entry_path) {
+            info!(
+                "Skipping because it matches an ignored glob: {}",
+                entry_path.display()
+            );
+        } else if is_javascript_file(&entry_path) {
             info!("Processing file: {}", entry_path.display());
             let source = SourceFile::load(&entry_path)?;
             let sourcemap_path = get_sourcemap_path(&source)?;

cli/tests/sourcemap.rs

@@ -19,14 +19,24 @@ fn assert_file_eq(base_path: &Path, path: &str, actual: &str) {
 
 #[test]
 fn test_search() {
-    let pairs = read_pairs(&get_case_path("search")).expect("Failed to read pairs");
+    let pairs = read_pairs(&get_case_path("search"), &Vec::new()).expect("Failed to read pairs");
+    assert_eq!(pairs.len(), 2);
+}
+
+#[test]
+fn test_ignore() {
+    let pairs = read_pairs(&get_case_path(""), &Vec::new()).expect("Failed to read pairs");
+    assert_eq!(pairs.len(), 4);
+
+    let pairs = read_pairs(&get_case_path(""), &vec!["**/search/**".to_string()])
+        .expect("Failed to read pairs");
     assert_eq!(pairs.len(), 2);
 }
 
 #[test]
 fn test_pair_inject() {
     let case_path = get_case_path("inject");
-    let mut pairs = read_pairs(&case_path).expect("Failed to read pairs");
+    let mut pairs = read_pairs(&case_path, &Vec::new()).expect("Failed to read pairs");
     assert_eq!(pairs.len(), 1);
     let current_pair = pairs.first_mut().expect("Failed to get first pair");
     let chunk_id = "00000-00000-00000";
@@ -49,7 +59,7 @@ fn test_pair_inject() {
 #[test]
 fn test_index_inject() {
     let case_path = get_case_path("index_map");
-    let mut pairs = read_pairs(&case_path).expect("Failed to read pairs");
+    let mut pairs = read_pairs(&case_path, &Vec::new()).expect("Failed to read pairs");
     let current_pair = pairs.first_mut().expect("Failed to get first pair");
     let chunk_id = "00000-00000-00000";
     current_pair