fix(cli,fixtures): support .yml config extension and fix fixture document paths

- Add .yml as accepted YAML config file extension (alongside .yaml)
- Update 63 fixture JSON files with correct test_documents paths after
  directory reorganization (pdfs/ -> pdf/, web/ -> html/, etc.)
- Fix test_invalid_config_values test using nonexistent config field

Author: Goldziher

CHANGELOG.md

@@ -7,6 +7,18 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ---
 
+## [Unreleased]
+
+### Fixed
+
+#### CLI
+- **`.yml` config files rejected by `--config` flag**: The CLI only accepted `.yaml` for YAML config files, rejecting the common `.yml` extension with a "must have .toml, .yaml, or .json extension" error. The config loader now accepts both `.yml` and `.yaml` (case-insensitive).
+
+#### CI
+- **Fixture document paths broken after test_documents reorganization**: All 63 fixture JSON files under `fixtures/` still referenced old directory names (`pdfs/`, `pdfs_with_tables/`, `web/`, `documents/`, `legacy_office/`, `presentations/`, `spreadsheets/`, `data_formats/`, `office/`) after the test_documents directory was restructured into per-format directories (`pdf/`, `html/`, `docx/`, `doc/`, `ppt/`, `pptx/`, `xlsx/`, `xls/`, `json/`, `yaml/`). This caused all Deno E2E tests and PDFium system tests to fail with "No such file or directory" errors.
+
+---
+
 ## [4.2.12] - 2026-02-06
 
 ### Fixed

crates/kreuzberg-cli/src/commands/config.rs

@@ -18,13 +18,13 @@ use std::path::PathBuf;
 ///
 /// Supports three formats, determined by file extension:
 /// - `.toml`: TOML format (recommended for humans)
-/// - `.yaml`: YAML format
+/// - `.yaml` / `.yml`: YAML format
 /// - `.json`: JSON format
 ///
 /// # Errors
 ///
 /// Returns an error if:
-/// - Explicit config file has unsupported extension (must be .toml, .yaml, or .json)
+/// - Explicit config file has unsupported extension (must be .toml, .yaml, .yml, or .json)
 /// - Config file cannot be read or parsed
 /// - Config file contains invalid extraction settings
 pub fn load_config(config_path: Option<PathBuf>) -> Result<ExtractionConfig> {
@@ -33,12 +33,12 @@ pub fn load_config(config_path: Option<PathBuf>) -> Result<ExtractionConfig> {
         let path_lower = path_str.to_lowercase();
         let config = if path_lower.ends_with(".toml") {
             ExtractionConfig::from_toml_file(&path)
-        } else if path_lower.ends_with(".yaml") {
+        } else if path_lower.ends_with(".yaml") || path_lower.ends_with(".yml") {
             ExtractionConfig::from_yaml_file(&path)
         } else if path_lower.ends_with(".json") {
             ExtractionConfig::from_json_file(&path)
         } else {
-            anyhow::bail!("Config file must have .toml, .yaml, or .json extension (case-insensitive)");
+            anyhow::bail!("Config file must have .toml, .yaml, .yml, or .json extension (case-insensitive)");
         };
         config.with_context(|| format!("Failed to load configuration from '{}'. Ensure the file exists, is readable, and contains valid configuration.", path.display()))
     } else {

crates/kreuzberg-cli/tests/config_discovery_test.rs

@@ -595,7 +595,7 @@ fn test_invalid_config_values() {
     fs::write(
         &config_path,
         r#"
-max_pages = -1
+use_cache = "not_a_bool"
 "#,
     )
     .unwrap();

fixtures/contract/api_batch_bytes_async.json

@@ -3,7 +3,7 @@
 	"description": "Tests async batch bytes extraction API (batch_extract_bytes)",
 	"tags": ["contract", "api", "batch"],
 	"document": {
-		"path": "pdfs/fake_memo.pdf"
+		"path": "pdf/fake_memo.pdf"
 	},
 	"extraction": {
 		"method": "batch_async",

fixtures/contract/api_batch_bytes_sync.json

@@ -3,7 +3,7 @@
 	"description": "Tests sync batch bytes extraction API (batch_extract_bytes_sync)",
 	"tags": ["contract", "api", "batch"],
 	"document": {
-		"path": "pdfs/fake_memo.pdf"
+		"path": "pdf/fake_memo.pdf"
 	},
 	"extraction": {
 		"method": "batch_sync",

fixtures/contract/api_batch_file_async.json

@@ -3,7 +3,7 @@
 	"description": "Tests async batch file extraction API (batch_extract_file)",
 	"tags": ["contract", "api", "batch"],
 	"document": {
-		"path": "pdfs/fake_memo.pdf"
+		"path": "pdf/fake_memo.pdf"
 	},
 	"extraction": {
 		"method": "batch_async",

fixtures/contract/api_batch_file_sync.json

@@ -3,7 +3,7 @@
 	"description": "Tests sync batch file extraction API (batch_extract_file_sync)",
 	"tags": ["contract", "api", "batch"],
 	"document": {
-		"path": "pdfs/fake_memo.pdf"
+		"path": "pdf/fake_memo.pdf"
 	},
 	"extraction": {
 		"method": "batch_sync",

fixtures/contract/api_extract_bytes_async.json

@@ -3,7 +3,7 @@
 	"description": "Tests async bytes extraction API (extract_bytes)",
 	"tags": ["contract", "api"],
 	"document": {
-		"path": "pdfs/fake_memo.pdf"
+		"path": "pdf/fake_memo.pdf"
 	},
 	"extraction": {
 		"method": "async",

fixtures/contract/api_extract_bytes_sync.json

@@ -3,7 +3,7 @@
 	"description": "Tests sync bytes extraction API (extract_bytes_sync)",
 	"tags": ["contract", "api"],
 	"document": {
-		"path": "pdfs/fake_memo.pdf"
+		"path": "pdf/fake_memo.pdf"
 	},
 	"extraction": {
 		"method": "sync",

fixtures/contract/api_extract_file_async.json

@@ -3,7 +3,7 @@
 	"description": "Tests async file extraction API (extract_file)",
 	"tags": ["contract", "api"],
 	"document": {
-		"path": "pdfs/fake_memo.pdf"
+		"path": "pdf/fake_memo.pdf"
 	},
 	"extraction": {
 		"method": "async",