Fix empty group special case

.gitignore

@@ -9,6 +9,7 @@ decondenser-wasm/src/bindings.rs
 
 # Files where logs can be stored and not committed
 *.log
+!decondenser/tests/integration/formatting/*.log
 
 # Special name pattern for dirty experimentation that should not be committed
 *dbg*

.vscode/launch.json

@@ -6,7 +6,9 @@
       "type": "extensionHost",
       "request": "launch",
       "args": [
-        "--extensionDevelopmentPath=${workspaceFolder}/decondenser-vscode"
+        "--extensionDevelopmentPath=${workspaceFolder}/decondenser-vscode",
+        "${workspaceFolder}/decondenser",
+        "${workspaceFolder}/decondenser/tests/integration/formatting/001-rust.log"
       ],
       "outFiles": ["${workspaceFolder}/decondenser-vscode/dist/**/*.js"],
       "preLaunchTask": "${defaultBuildTask}"
@@ -21,7 +23,7 @@
         },
         "args": ["build", "-p", "decondenser-cli"]
       },
-      "args": ["--input=dbg.log", "--output=dbg.out.log"]
+      "args": ["fmt", "--input=dbg.log", "--output=dbg.out.log"]
     }
   ]
 }

decondenser-wasm/src/lib.rs

@@ -28,6 +28,8 @@ impl wit::GuestDecondenser for Decondenser {
     }
 
     fn format(&self, input: String) -> String {
+        eprintln!("EPRINTLN INVOKED");
+        println!("PRINTLN INVOKED");
         // Delegate to the inherent method of the same name. No, it's not an
         // unconditional recursion, the compiler would report it otherwise  .
         self.format(&input)

decondenser/src/formatting/mod.rs

@@ -95,8 +95,6 @@ impl<'i> FormattingCtx<'_, 'i> {
     fn on_group(&mut self, leading_space: Option<&'i str>, group: &'i parsing::l2::Group<'i>) {
         let config = group.config;
 
-        self.fmt.begin(group.config.break_style.0);
-
         let is_empty_group = group
             .content
             .iter()
@@ -108,6 +106,8 @@ impl<'i> FormattingCtx<'_, 'i> {
             return;
         }
 
+        self.fmt.begin(group.config.break_style.0);
+
         let mut tokens = group.content.iter();
 
         let closing_punct_leading_blank = tokens

decondenser/tests/integration/formatting/001-rust.log

@@ -0,0 +1 @@
+Library { name: "Thornwick Archive", location: "Eldenmoor, Westreach", books: [Book { id: 1001, title: "The Hollow Lexicon", authors: [Author { name: "Elenya Thorne", pseudonym: Some("E. T. Mire"), birth_year: Some(1972) }], metadata: Metadata { isbn: Some("978-3-8267-9321-0"), pages: 412, genres: ["Mystery", "Arcane Fiction"], language: "Eldentongue", rating: Some(4.6) }, status: Available }, Book { id: 1002, title: "Cradle of Ironlight", authors: [Author { name: "Jorren Valek", pseudonym: None, birth_year: Some(1960) }, Author { name: "Silra Venn", pseudonym: None, birth_year: None }], metadata: Metadata { isbn: None, pages: 634, genres: ["Epic Fantasy"], language: "Virellic", rating: Some(4.9) }, status: CheckedOut { by_user_id: 78230944, due_date: "2025-08-21" } }, Book { id: 1003, title: "Treatise on Umbral Taxonomy", authors: [Author { name: "Professor N. W. Glaive", pseudonym: None, birth_year: None }], metadata: Metadata { isbn: Some("979-1-9472-3829-5"), pages: 271, genres: ["Non-fiction", "Shadow Studies"], language: "Imperial Common", rating: None }, status: Reserved(2198371) }], publishers: {"Marrow & Quill": Publisher { name: "Marrow & Quill", country: "Draventh", founded: 1951, imprints: [] }, "Cindergrove Press": Publisher { name: "Cindergrove Press", country: "Virelia", founded: 1894, imprints: ["Ash & Ink", "Frostmark Editions"] }} }

decondenser/tests/integration/formatting/001-rust.out.log

@@ -0,0 +1,78 @@
+Library {
+    name: "Thornwick Archive",
+    location: "Eldenmoor, Westreach",
+    books: [
+        Book {
+            id: 1001,
+            title: "The Hollow Lexicon",
+            authors: [
+                Author {
+                    name: "Elenya Thorne",
+                    pseudonym: Some("E. T. Mire"),
+                    birth_year: Some(1972)
+                }
+            ],
+            metadata: Metadata {
+                isbn: Some("978-3-8267-9321-0"),
+                pages: 412,
+                genres: ["Mystery", "Arcane Fiction"],
+                language: "Eldentongue",
+                rating: Some(4.6)
+            },
+            status: Available
+        },
+        Book {
+            id: 1002,
+            title: "Cradle of Ironlight",
+            authors: [
+                Author {
+                    name: "Jorren Valek",
+                    pseudonym: None,
+                    birth_year: Some(1960)
+                },
+                Author { name: "Silra Venn", pseudonym: None, birth_year: None }
+            ],
+            metadata: Metadata {
+                isbn: None,
+                pages: 634,
+                genres: ["Epic Fantasy"],
+                language: "Virellic",
+                rating: Some(4.9)
+            },
+            status: CheckedOut { by_user_id: 78230944, due_date: "2025-08-21" }
+        },
+        Book {
+            id: 1003,
+            title: "Treatise on Umbral Taxonomy",
+            authors: [
+                Author {
+                    name: "Professor N. W. Glaive",
+                    pseudonym: None,
+                    birth_year: None
+                }
+            ],
+            metadata: Metadata {
+                isbn: Some("979-1-9472-3829-5"),
+                pages: 271,
+                genres: ["Non-fiction", "Shadow Studies"],
+                language: "Imperial Common",
+                rating: None
+            },
+            status: Reserved(2198371)
+        }
+    ],
+    publishers: {
+        "Marrow & Quill": Publisher {
+            name: "Marrow & Quill",
+            country: "Draventh",
+            founded: 1951,
+            imprints: []
+        },
+        "Cindergrove Press": Publisher {
+            name: "Cindergrove Press",
+            country: "Virelia",
+            founded: 1894,
+            imprints: ["Ash & Ink", "Frostmark Editions"]
+        }
+    }
+}

decondenser/tests/integration/formatting/002-elixir.log

@@ -0,0 +1 @@
+no match of right hand side value: {:ok, %Req.Response{status: 400, headers: %{"content-type" => ["application/json; charset=UTF-8"], "task_resource_usage" => ["{\"action\":\"indices:data/read/search[phase/query]\",\"taskId\":1979,\"parentTaskId\":1978,\"nodeId\":\"bpv3-LSSR4KXrWJNugVRDA\",\"taskResourceUsage\":{\"cpu_time_in_nanos\":392450,\"memory_in_bytes\":7456}}"]}, body: %{"error" => %{"caused_by" => %{"caused_by" => %{"reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.", "type" => "illegal_argument_exception"}, "reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.", "type" => "illegal_argument_exception"}, "failed_shards" => [%{"index" => "tags", "node" => "bpv3-LSSR4KXrWJNugVRDA", "reason" => %{"reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.", "type" => "illegal_argument_exception"}, "shard" => 0}], "grouped" => true, "phase" => "query", "reason" => "all shards failed", "root_cause" => [%{"reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.", "type" => "illegal_argument_exception"}], "type" => "search_phase_execution_exception"}, "status" => 400}, trailers: %{}, private: %{}}}

decondenser/tests/integration/formatting/002-elixir.out.log

@@ -0,0 +1,48 @@
+no match of right hand side value: {
+    :ok,
+    %Req.Response{
+        status: 400,
+        headers: %{
+            "content-type" => ["application/json; charset=UTF-8"],
+            "task_resource_usage" => [
+                "{\"action\":\"indices:data/read/search[phase/query]\",\"taskId\":1979,\"parentTaskId\":1978,\"nodeId\":\"bpv3-LSSR4KXrWJNugVRDA\",\"taskResourceUsage\":{\"cpu_time_in_nanos\":392450,\"memory_in_bytes\":7456}}"
+            ]
+        },
+        body: %{
+            "error" => %{
+                "caused_by" => %{
+                    "caused_by" => %{
+                        "reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.",
+                        "type" => "illegal_argument_exception"
+                    },
+                    "reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.",
+                    "type" => "illegal_argument_exception"
+                },
+                "failed_shards" => [
+                    %{
+                        "index" => "tags",
+                        "node" => "bpv3-LSSR4KXrWJNugVRDA",
+                        "reason" => %{
+                            "reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.",
+                            "type" => "illegal_argument_exception"
+                        },
+                        "shard" => 0
+                    }
+                ],
+                "grouped" => true,
+                "phase" => "query",
+                "reason" => "all shards failed",
+                "root_cause" => [
+                    %{
+                        "reason" => "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [name] in order to load field data by uninverting the inverted index. Note that this can use significant memory.",
+                        "type" => "illegal_argument_exception"
+                    }
+                ],
+                "type" => "search_phase_execution_exception"
+            },
+            "status" => 400
+        },
+        trailers: %{},
+        private: %{}
+    }
+}

decondenser/tests/integration/main.rs

@@ -3,17 +3,26 @@
 use decondenser::Decondenser;
 use std::borrow::Cow;
 use std::io::Write;
-use std::path::PathBuf;
+use std::path::{Path, PathBuf};
 use std::process::Stdio;
 use std::str::FromStr;
 
-/// This test updates the `output` values in the `formatting.toml` file. It
-/// never fails, it only updates the file. However, on CI we make sure that the
-/// `formatting.toml` file is fresh by checking if it changes after the test
-/// run. This way we can observe the changes to the output values in PR diffs
-/// and have them automatically updated via `scripts/update-tests.sh`.
+fn tests_dir() -> &'static Path {
+    static CACHE: std::sync::OnceLock<PathBuf> = std::sync::OnceLock::new();
+
+    CACHE.get_or_init(|| {
+        PathBuf::from_iter([
+            std::env::var("CARGO_MANIFEST_DIR")
+                .as_deref()
+                .unwrap_or(env!("CARGO_MANIFEST_DIR")),
+            "tests",
+            "integration",
+        ])
+    })
+}
+
 #[test]
-fn formatting() {
+fn formatting_toml() {
     Snapshot::new("formatting.toml").update(|test| {
         let input = test["input"].as_str().unwrap();
 
@@ -54,7 +63,33 @@ fn formatting() {
 }
 
 #[test]
-fn unescaping() {
+fn formatting_dir() {
+    let tests = std::fs::read_dir(tests_dir().join("formatting")).unwrap();
+
+    for test in tests {
+        let test_path = test.unwrap().path();
+        let test_name = test_path.to_string_lossy();
+
+        let mut test_name_parts = test_name.split('.').collect::<Vec<_>>();
+
+        if test_name_parts.get(test_name_parts.len().saturating_sub(2)) == Some(&"out") {
+            continue;
+        }
+
+        let decondenser = Decondenser::generic();
+
+        let input = std::fs::read_to_string(&test_path).unwrap();
+        let output = decondenser.format(&input);
+
+        test_name_parts.insert(test_name_parts.len().saturating_sub(1), "out");
+        let out_path = test_path.with_file_name(test_name_parts.join("."));
+
+        std::fs::write(&out_path, output).unwrap();
+    }
+}
+
+#[test]
+fn unescaping_toml() {
     Snapshot::new("unescaping.toml").update(|test| {
         let input = test["input"].as_str().unwrap();
         let output = decondenser::unescape(input);
@@ -74,14 +109,7 @@ struct Snapshot {
 
 impl Snapshot {
     fn new(file_name: &str) -> Self {
-        let path = PathBuf::from_iter([
-            std::env::var("CARGO_MANIFEST_DIR")
-                .as_deref()
-                .unwrap_or(env!("CARGO_MANIFEST_DIR")),
-            "tests",
-            "integration",
-            file_name,
-        ]);
+        let path = tests_dir().join(file_name);
 
         let file = std::fs::read_to_string(&path).unwrap();