fallow_engine/

session.rs

//! Engine-owned analysis session orchestration.

use std::path::{Path, PathBuf};
use std::sync::atomic::{AtomicBool, Ordering};
use std::sync::{Arc, Mutex};
use std::time::Instant;

use fallow_config::{DuplicatesConfig, ResolvedConfig, WorkspaceInfo};
use fallow_types::cache_rejection::CacheRejection;
use fallow_types::discover::DiscoveredFile;
use fallow_types::extract::{ModuleInfo, SourceParseDegradation, SourceReadFailure};
#[cfg(test)]
use fallow_types::results::AnalysisResults;
use fallow_types::source_fingerprint::SourceFingerprint;
use fallow_types::workspace::{WorkspaceDiagnostic, merge_workspace_diagnostics};
use rustc_hash::{FxHashMap, FxHashSet};

pub use crate::session_reuse::SessionParseCounts;
use crate::session_reuse::{
    MAX_INCREMENTAL_REPARSE_FILES, ParseCountCells, changed_file_indices, merge_reparsed_modules,
};
use crate::{
    EngineResult, core_backend, duplicates,
    project_analysis::{
        ProjectAnalysisArtifactOptions, ProjectAnalysisArtifacts, ProjectAnalysisOutput,
    },
    project_config::{ProjectConfig, config_for_project, default_project_config},
    results::{
        DeadCodeAnalysis, DeadCodeAnalysisArtifacts, DeadCodeAnalysisOutput, DuplicationAnalysis,
        SharedDeadCodeAnalysisArtifacts,
    },
    warm_parse::{WarmParse, WarmParseKey, WarmParseStore},
};

/// Reusable engine session for one resolved project.
///
/// The session owns the resolved config and discovered file set so future
/// consumers can share graph-sensitive inputs without each surface recreating
/// its own partial orchestration.
#[derive(Debug)]
pub struct AnalysisSession {
    config: ResolvedConfig,
    config_path: Option<PathBuf>,
    config_inputs: fallow_config::ConfigInputs,
    config_inputs_before_resolve: fallow_config::ConfigInputsSnapshot,
    discovery: crate::discover::AnalysisDiscovery,
    workspaces: Vec<WorkspaceInfo>,
    workspace_diagnostics: Vec<WorkspaceDiagnostic>,
    parsed_cache: Mutex<Option<ParsedModuleCache>>,
    styling_cache: Mutex<Option<Arc<crate::health::StylingAnalysisArtifacts>>>,
    cancellation: Option<Arc<AtomicBool>>,
    warm_parse: Option<Arc<WarmParseStore>>,
    /// Config load already resolved the workspaces, so a discovery refresh
    /// keeps them.
    preloaded_workspaces: bool,
    parse_counts: ParseCountCells,
    /// An incremental parse changed modules that the persisted parse cache
    /// does not hold yet. [`AnalysisSession::flush_parse_cache`] writes them.
    disk_cache_stale: AtomicBool,
}

#[derive(Debug)]
struct ParsedModuleCache {
    need_complexity: bool,
    fingerprints: Vec<SourceFingerprint>,
    modules: Arc<[ModuleInfo]>,
    /// The read failures of the parse, kept so that an incremental parse can
    /// record the full set of the project again.
    read_failures: Vec<SourceReadFailure>,
    /// The parse degradations of the parse, kept for the same reason.
    parse_degradations: Vec<SourceParseDegradation>,
}

/// Owned session parts for runners that need to continue an existing pipeline.
#[derive(Debug)]
pub struct AnalysisSessionParts {
    /// Resolved project config the session was created with.
    pub config: ResolvedConfig,
    /// Path of the loaded config file; `None` when defaults were used.
    pub config_path: Option<PathBuf>,
    /// Files discovered under the session root.
    pub files: Vec<DiscoveredFile>,
    /// Workspace metadata discovered during config resolution.
    pub workspaces: Vec<WorkspaceInfo>,
    /// Diagnostics from workspace discovery (undeclared or invalid members).
    pub workspace_diagnostics: Vec<WorkspaceDiagnostic>,
}

/// Owned session parts after parsing the discovered files.
#[derive(Debug)]
pub struct ParsedAnalysisSessionParts {
    /// Resolved project config the session was created with.
    pub config: ResolvedConfig,
    /// Path of the loaded config file; `None` when defaults were used.
    pub config_path: Option<PathBuf>,
    /// Files discovered under the session root.
    pub files: Vec<DiscoveredFile>,
    /// Parsed modules, one per discovered file.
    pub modules: Vec<ModuleInfo>,
    /// Workspace metadata discovered during config resolution.
    pub workspaces: Vec<WorkspaceInfo>,
    /// Diagnostics from workspace discovery (undeclared or invalid members).
    pub workspace_diagnostics: Vec<WorkspaceDiagnostic>,
    /// Parse wall time in milliseconds.
    pub parse_ms: f64,
    /// Parse-cache write-back wall time in milliseconds.
    pub cache_update_ms: f64,
    /// Files served from the parse cache.
    pub cache_hits: usize,
    /// Files that had to be parsed fresh.
    pub cache_misses: usize,
    /// Summed parse CPU time across rayon workers in milliseconds.
    pub parse_cpu_ms: f64,
}

#[derive(Debug)]
pub(crate) struct SharedParsedAnalysisSessionParts {
    pub(crate) config: ResolvedConfig,
    pub(crate) files: Vec<DiscoveredFile>,
    pub(crate) modules: Arc<[ModuleInfo]>,
    pub workspaces: Vec<WorkspaceInfo>,
    pub workspace_diagnostics: Vec<WorkspaceDiagnostic>,
    pub parse_ms: f64,
    pub parse_cpu_ms: f64,
}

/// Reusable artifacts produced by one session-owned dead-code run.
#[derive(Debug)]
pub struct AnalysisSessionArtifacts {
    /// Retained dead-code analysis output (results, graph, timings).
    pub analysis: DeadCodeAnalysisArtifacts,
    /// Diff scope the run was limited to, when one was resolved.
    pub changed_files: Option<FxHashSet<PathBuf>>,
    /// Per-file source fingerprints for downstream cache invalidation.
    pub source_fingerprints: FxHashMap<PathBuf, SourceFingerprint>,
}

impl AnalysisSession {
    /// Load config and discover files for a project root.
    ///
    /// # Errors
    ///
    /// Returns an error when config loading fails.
    pub fn load(root: &Path, config_path: Option<&Path>) -> EngineResult<Self> {
        let project_config = config_for_project(root, config_path)?;
        Ok(Self::from_config(project_config))
    }

    /// Load config, apply one caller-supplied config adjustment, then discover
    /// files for a project root.
    ///
    /// # Errors
    ///
    /// Returns an error when config loading fails.
    pub fn load_with_config(
        root: &Path,
        config_path: Option<&Path>,
        configure: impl FnOnce(&mut ResolvedConfig),
    ) -> EngineResult<Self> {
        Self::load_with_config_options(
            root,
            config_path,
            fallow_config::ConfigLoadOptions::default(),
            configure,
        )
    }

    /// Load config with an explicit inheritance trust policy, apply one
    /// caller-supplied adjustment, then discover project files.
    ///
    /// # Errors
    ///
    /// Returns an error when config loading fails.
    pub fn load_with_config_options(
        root: &Path,
        config_path: Option<&Path>,
        load_options: fallow_config::ConfigLoadOptions,
        configure: impl FnOnce(&mut ResolvedConfig),
    ) -> EngineResult<Self> {
        let mut project_config = crate::project_config::config_for_project_with_load_options(
            root,
            config_path,
            load_options,
        )?;
        configure(&mut project_config.config);
        project_config.workspaces.clear();
        project_config.workspace_diagnostics.clear();
        project_config.workspace_discovery_ms = None;
        Ok(Self::from_config(project_config))
    }

    /// Build a session from built-in defaults, ignoring project config files.
    ///
    /// This is intended for editor fallback paths that have already reported a
    /// config-load warning but should still surface best-effort diagnostics.
    #[must_use]
    pub fn load_default(root: &Path) -> Self {
        Self::load_default_with_config(root, |_| {})
    }

    /// Build a session from built-in defaults, apply one caller-supplied
    /// adjustment, then discover project files.
    #[must_use]
    pub fn load_default_with_config(
        root: &Path,
        configure: impl FnOnce(&mut ResolvedConfig),
    ) -> Self {
        let mut project_config = default_project_config(root);
        configure(&mut project_config.config);
        Self::from_config(project_config)
    }

    /// Build a session from a previously resolved config.
    #[must_use]
    pub fn from_config(project_config: ProjectConfig) -> Self {
        let uses_preloaded_workspaces = project_config.workspace_discovery_ms.is_some();
        let discovery = if let Some(workspace_discovery_ms) = project_config.workspace_discovery_ms
        {
            crate::discover::prepare_analysis_discovery_with_workspaces(
                &project_config.config,
                &project_config.workspaces,
                workspace_discovery_ms,
            )
        } else {
            crate::discover::prepare_analysis_discovery(&project_config.config)
        };
        let workspaces = if uses_preloaded_workspaces {
            project_config.workspaces
        } else {
            discovery.workspaces().to_vec()
        };
        // Analysis-stage diagnostics are owned by the analyze pass, which
        // refreshes the registry on every run; pinning them in the session
        // snapshot would keep a stale entry alive after the cause is fixed
        // (issue #2366). `current_workspace_diagnostics` reads them live.
        //
        // Source-discovery entries come from THIS walk's return value, not from
        // the registry: combined mode runs the dead-code and duplication walks
        // concurrently whenever a per-analysis `production` split stops them
        // from sharing a file list, and each walk replaces the registry's
        // source-discovery set for the root, so a registry read here would
        // report whichever walk happened to write last (issue #2366).
        let workspace_diagnostics = merge_workspace_diagnostics(
            merge_workspace_diagnostics(
                project_config.workspace_diagnostics,
                fallow_config::workspace_diagnostics_for(&project_config.config.root)
                    .into_iter()
                    .filter(|diagnostic| {
                        !diagnostic.kind.is_analysis_stage()
                            && !diagnostic.kind.is_source_discovery()
                    })
                    .collect(),
            ),
            discovery.source_diagnostics().to_vec(),
        );
        Self {
            config: project_config.config,
            config_path: project_config.path,
            config_inputs: project_config.inputs,
            config_inputs_before_resolve: project_config.inputs_before_resolve,
            discovery,
            workspaces,
            workspace_diagnostics,
            parsed_cache: Mutex::new(None),
            styling_cache: Mutex::new(None),
            cancellation: None,
            warm_parse: crate::warm_parse::installed(),
            preloaded_workspaces: uses_preloaded_workspaces,
            parse_counts: ParseCountCells::default(),
            disk_cache_stale: AtomicBool::new(false),
        }
    }

    /// Attach a caller-owned cancellation token to this session.
    ///
    /// Analyses that run through the session check the token at each pipeline
    /// stage boundary and inside the per-file parse loop, and return
    /// [`crate::EngineError::cancelled`] instead of a partial result once it is
    /// set. A session without a token can never be cancelled, so existing
    /// callers keep their current behavior.
    ///
    /// The stop is cooperative, and how long it takes is bounded by the
    /// longest stage that holds no check, not by any promised latency. Only
    /// the parse loop stops per file. Duplication detection (tokenization plus
    /// suffix-array matching) and the dead-code detectors run to the end of the
    /// stage once entered, and on a large repository either is seconds of work.
    /// A caller that needs a bounded stop has to kill the process instead.
    #[must_use]
    pub fn with_cancellation(mut self, cancellation: Arc<AtomicBool>) -> Self {
        self.cancellation = Some(cancellation);
        self
    }

    /// Use `store` for parsed modules across sessions, in place of the store
    /// that the process installed with [`crate::warm_parse::install`].
    /// `None` makes the session parse through the persisted cache only.
    #[must_use]
    pub fn with_warm_parse(mut self, store: Option<Arc<WarmParseStore>>) -> Self {
        self.warm_parse = store;
        self
    }

    /// Replace the cancellation token of a session that serves several runs.
    ///
    /// A long-lived session gets the token of each new run, so a set token of
    /// an earlier run does not stop the next one.
    pub fn set_cancellation(&mut self, cancellation: Arc<AtomicBool>) {
        self.cancellation = Some(cancellation);
    }

    /// The parse work of this session since it was created.
    #[must_use]
    pub fn parse_counts(&self) -> SessionParseCounts {
        self.parse_counts.snapshot()
    }

    /// Walk the project again and keep the parsed modules when the file set
    /// did not change.
    ///
    /// A session that serves several runs calls this before each run, so a
    /// created or deleted file reaches the analysis without a config reload.
    /// When the file set changed, the file ids move, so the session writes its
    /// parsed modules to the persisted parse cache and drops them. The next
    /// parse then reads that cache. Returns whether the file set changed.
    ///
    /// The session also drops its modules when a fingerprint of the cached
    /// parse cannot stand in for the file content, as on a platform without
    /// ctime. A same-size edit with a restored mtime keeps such a
    /// fingerprint, so only the persisted cache, which then compares content
    /// hashes, can tell whether the module is current.
    pub fn refresh_discovery(&mut self) -> bool {
        let discovery = if self.preloaded_workspaces {
            crate::discover::prepare_analysis_discovery_with_workspaces(
                &self.config,
                &self.workspaces,
                0.0,
            )
        } else {
            crate::discover::prepare_analysis_discovery(&self.config)
        };
        let same_files = discovery.files().len() == self.files().len()
            && discovery
                .files()
                .iter()
                .zip(self.files())
                .all(|(fresh, known)| fresh.path == known.path);
        if !same_files || !self.cached_fingerprints_are_trustworthy() {
            self.flush_parse_cache();
            *self
                .parsed_cache
                .get_mut()
                .unwrap_or_else(std::sync::PoisonError::into_inner) = None;
            self.clear_styling_cache();
        }
        if !self.preloaded_workspaces {
            self.workspaces = discovery.workspaces().to_vec();
        }
        self.discovery = discovery;
        !same_files
    }

    /// Whether each fingerprint of the cached parse can stand in for the file
    /// content. No cached parse counts as trustworthy.
    fn cached_fingerprints_are_trustworthy(&mut self) -> bool {
        self.parsed_cache
            .get_mut()
            .unwrap_or_else(std::sync::PoisonError::into_inner)
            .as_ref()
            .is_none_or(|cache| {
                cache
                    .fingerprints
                    .iter()
                    .all(|fingerprint| fingerprint.is_trustworthy_without_content())
            })
    }

    /// Write the modules of incremental parses to the persisted parse cache.
    ///
    /// Each module is stored with the fingerprint that the session read before
    /// it parsed the file. A file that changed after that parse then misses
    /// the cache instead of serving the older module. Does nothing when no
    /// incremental parse happened since the last write, or when the cache is
    /// off.
    ///
    /// The session does not write these modules when it is dropped. A caller
    /// that parses again after files changed, such as a session that lives
    /// across editor runs, calls this method before it drops the session. A
    /// session that parses once needs no call: the full parse writes the
    /// persisted cache itself. The editor server is the only caller now.
    pub fn flush_parse_cache(&self) {
        if self.config.no_cache || !self.disk_cache_stale.swap(false, Ordering::SeqCst) {
            return;
        }
        let Ok(cache) = self.parsed_cache.lock() else {
            return;
        };
        let Some(cache) = cache.as_ref() else {
            return;
        };
        let cache_max_size_bytes =
            crate::project_config::resolve_cache_max_size_bytes(&self.config);
        let mut store = fallow_extract::cache::CacheStore::load(
            &self.config.cache_dir,
            &self.config.root,
            self.config.cache_config_hash,
            cache_max_size_bytes,
        )
        .ok();
        let files = self.files();
        write_parse_cache(
            &self.config,
            &mut store,
            &ParseCacheWrite {
                modules: &cache.modules,
                files,
                need_complexity: cache.need_complexity,
                fingerprint_of: &|file: &DiscoveredFile| {
                    cache
                        .fingerprints
                        .get(file.id.0 as usize)
                        .copied()
                        .unwrap_or_else(|| SourceFingerprint::new(0, 0))
                },
            },
        );
    }

    /// Parse the discovered files into the module cache of the session,
    /// without analysis. A later run of this session then starts from warm
    /// modules.
    ///
    /// # Errors
    ///
    /// Returns [`crate::EngineError::cancelled`] when the token of the session
    /// is set.
    pub fn prewarm_parsed_modules(&self, need_complexity: bool) -> EngineResult<()> {
        self.shared_parsed_modules_cancellable(need_complexity, "prewarm")
            .map(drop)
    }

    fn clear_styling_cache(&self) {
        if let Ok(mut cache) = self.styling_cache.lock() {
            *cache = None;
        }
    }

    /// Whether this session's caller has requested cancellation.
    #[must_use]
    pub fn is_cancelled(&self) -> bool {
        self.cancellation
            .as_ref()
            .is_some_and(|cancelled| cancelled.load(Ordering::SeqCst))
    }

    fn ensure_not_cancelled(&self, stage: &str) -> EngineResult<()> {
        if self.is_cancelled() {
            return Err(crate::EngineError::cancelled(stage));
        }
        Ok(())
    }

    /// Build a session from a resolved config when the caller already owns
    /// command-specific config loading.
    ///
    /// # Errors
    ///
    /// Returns an engine error when root manifest loading fails during
    /// workspace discovery, matching `ProjectConfig::load`.
    pub fn from_resolved_config(config: ResolvedConfig) -> EngineResult<Self> {
        let (workspaces, workspace_diagnostics, workspace_discovery_ms) =
            crate::project_config::collect_workspace_metadata(&config)?;
        Ok(Self::from_config(ProjectConfig {
            inputs: fallow_config::ConfigInputs::default(),
            inputs_before_resolve: fallow_config::ConfigInputsSnapshot::default(),
            config,
            path: None,
            workspaces,
            workspace_diagnostics,
            workspace_discovery_ms: Some(workspace_discovery_ms),
        }))
    }

    /// Resolved project root.
    #[must_use]
    pub fn root(&self) -> &Path {
        &self.config.root
    }

    /// Resolved project config.
    #[must_use]
    pub fn config(&self) -> &ResolvedConfig {
        &self.config
    }

    /// Config file path when one was loaded.
    #[must_use]
    pub fn config_path(&self) -> Option<&Path> {
        self.config_path.as_deref()
    }

    /// The plugin files, rule packs and `autoDiscover` directories that
    /// config resolution read. A session built from a resolved config
    /// lists none.
    #[must_use]
    pub const fn config_inputs(&self) -> &fallow_config::ConfigInputs {
        &self.config_inputs
    }

    /// The content of [`Self::config_inputs`] just before config resolution
    /// read them. A snapshot after the load that differs from this one tells
    /// that an input changed during the load.
    #[must_use]
    pub const fn config_inputs_before_resolve(&self) -> &fallow_config::ConfigInputsSnapshot {
        &self.config_inputs_before_resolve
    }

    /// The estimated heap memory of the parsed modules that the session
    /// keeps between calls, with the estimate of
    /// [`crate::warm_parse::estimated_retained_bytes`]. Zero before the first
    /// parse.
    #[must_use]
    pub fn retained_bytes_estimate(&self) -> u64 {
        self.parsed_cache
            .lock()
            .ok()
            .and_then(|cache| {
                cache
                    .as_ref()
                    .map(|cache| crate::warm_parse::estimated_retained_bytes(&cache.fingerprints))
            })
            .unwrap_or(0)
    }

    /// Discovered files for this session.
    #[must_use]
    pub fn files(&self) -> &[DiscoveredFile] {
        self.discovery.files()
    }

    /// Workspace packages discovered during config/session setup.
    #[must_use]
    pub fn workspaces(&self) -> &[WorkspaceInfo] {
        &self.workspaces
    }

    /// Source metadata fingerprints for every discovered source file.
    #[must_use]
    fn source_fingerprints(&self) -> FxHashMap<PathBuf, SourceFingerprint> {
        self.discovery
            .files()
            .iter()
            .map(|file| {
                let fingerprint = std::fs::metadata(&file.path).map_or_else(
                    |_| SourceFingerprint::new(0, file.size_bytes),
                    |metadata| SourceFingerprint::from_metadata(&metadata),
                );
                (file.path.clone(), fingerprint)
            })
            .collect()
    }

    /// Resolve files changed since a git ref against this session root.
    ///
    /// # Errors
    ///
    /// Returns an error when the ref is invalid, git is unavailable, or the
    /// root is not part of a repository.
    pub(crate) fn changed_files_since(
        &self,
        git_ref: &str,
    ) -> Result<FxHashSet<PathBuf>, crate::changed_files::ChangedFilesError> {
        crate::changed_files::changed_files(&self.config.root, git_ref)
    }

    /// The discovery this session walked.
    pub(crate) const fn discovery(&self) -> &crate::discover::AnalysisDiscovery {
        &self.discovery
    }

    /// Workspace and source-discovery diagnostics captured for this session.
    #[must_use]
    pub fn workspace_diagnostics(&self) -> &[WorkspaceDiagnostic] {
        &self.workspace_diagnostics
    }

    /// Current diagnostics, including the source read failures the parse stage
    /// discovers and the analysis-stage entries the analyze pass records, both
    /// of which land in the registry after the session was created.
    ///
    /// The live read goes through
    /// [`fallow_config::registry_diagnostics_to_fold`], which drops
    /// walk-recorded entries for the same reason the constructor does: a
    /// concurrent walk on the same root replaces that set, so importing it here
    /// would make this session's list depend on which walk wrote last, and the
    /// combined root's union would come out in a different ORDER between runs
    /// of the same command (issue #2366). This session's own walk-recorded
    /// entries are already in the snapshot, by value, from its own walk.
    #[must_use]
    pub fn current_workspace_diagnostics(&self) -> Vec<WorkspaceDiagnostic> {
        merge_workspace_diagnostics(
            self.workspace_diagnostics.clone(),
            fallow_config::registry_diagnostics_to_fold(&self.config.root),
        )
    }

    pub(crate) fn styling_analysis_artifacts(
        &self,
    ) -> Arc<crate::health::StylingAnalysisArtifacts> {
        if let Ok(cache) = self.styling_cache.lock()
            && let Some(artifacts) = cache.as_ref()
        {
            return Arc::clone(artifacts);
        }

        let modules = self.shared_parsed_modules(false);
        let artifacts = Arc::new(crate::health::build_styling_analysis_artifacts(
            self.files(),
            &modules,
            self.config(),
        ));
        if let Ok(mut cache) = self.styling_cache.lock() {
            *cache = Some(Arc::clone(&artifacts));
        }
        artifacts
    }

    /// Consume the session and return the resolved config plus discovery data.
    #[must_use]
    pub fn into_parts(self) -> AnalysisSessionParts {
        let workspace_diagnostics = self.current_workspace_diagnostics();
        AnalysisSessionParts {
            config: self.config,
            config_path: self.config_path,
            files: self.discovery.into_files(),
            workspaces: self.workspaces,
            workspace_diagnostics,
        }
    }

    /// Consume the session, load the parser cache, and parse discovered files.
    #[must_use]
    pub fn into_parsed_parts(self, need_complexity: bool) -> ParsedAnalysisSessionParts {
        let AnalysisSessionParts {
            config,
            config_path,
            files,
            workspaces,
            workspace_diagnostics,
        } = self.into_parts();
        let ParsedModules {
            modules,
            metrics,
            source_diagnostics,
            ..
        } = parse_files_with_config(&config, &files, need_complexity, None);
        ParsedAnalysisSessionParts {
            config,
            config_path,
            files,
            modules,
            workspaces,
            workspace_diagnostics: merge_workspace_diagnostics(
                workspace_diagnostics,
                source_diagnostics,
            ),
            parse_ms: metrics.parse_ms,
            cache_update_ms: metrics.cache_ms,
            cache_hits: metrics.cache_hits,
            cache_misses: metrics.cache_misses,
            parse_cpu_ms: metrics.parse_cpu_ms,
        }
    }

    /// Parse discovered files without consuming the session.
    #[must_use]
    pub fn parsed_parts(&self, need_complexity: bool) -> ParsedAnalysisSessionParts {
        let SharedParsedModules { modules, metrics } = self.parse_modules(need_complexity, None);
        self.parsed_parts_from_modules(modules.to_vec(), metrics)
    }

    /// Parse discovered files while retaining shared immutable module storage.
    #[must_use]
    pub(crate) fn shared_parsed_parts(
        &self,
        need_complexity: bool,
    ) -> SharedParsedAnalysisSessionParts {
        let SharedParsedModules { modules, metrics } = self.parse_modules(need_complexity, None);
        SharedParsedAnalysisSessionParts {
            config: self.config.clone(),
            files: self.discovery.files().to_vec(),
            modules,
            workspaces: self.workspaces.clone(),
            workspace_diagnostics: self.current_workspace_diagnostics(),
            parse_ms: metrics.parse_ms,
            parse_cpu_ms: metrics.parse_cpu_ms,
        }
    }

    /// Return immutable parsed modules backed by the reusable session cache.
    ///
    /// Workspace-owned consumers use this additive path when they only need
    /// parsed modules and can borrow discovery and config directly from the
    /// session. Stable owned callers can continue using [`Self::parsed_parts`].
    #[doc(hidden)]
    #[must_use]
    pub(crate) fn shared_parsed_modules(&self, need_complexity: bool) -> Arc<[ModuleInfo]> {
        self.parse_modules(need_complexity, None).modules
    }

    /// Parse the discovered files, stopping if the session's caller cancelled.
    ///
    /// The token is checked on both sides of the parse loop, so the truncated
    /// module set a cancelled parse produces is never returned as a smaller
    /// project. `stage` names the work that would have followed.
    ///
    /// # Errors
    ///
    /// Returns [`crate::EngineError::cancelled`] when the token is set. A
    /// session without a token can never return it.
    pub(crate) fn shared_parsed_modules_cancellable(
        &self,
        need_complexity: bool,
        stage: &str,
    ) -> EngineResult<Arc<[ModuleInfo]>> {
        self.ensure_not_cancelled("parsing")?;
        let modules = self
            .parse_modules(need_complexity, self.cancellation.as_deref())
            .modules;
        self.ensure_not_cancelled(stage)?;
        Ok(modules)
    }

    /// Parse discovered files without consuming the session or retaining parser
    /// output in the session cache.
    #[must_use]
    pub fn parsed_parts_uncached(&self, need_complexity: bool) -> ParsedAnalysisSessionParts {
        let fingerprints = self
            .warm_parse
            .as_ref()
            .and_then(|_| source_fingerprints_for_files(self.files()));
        if let Some(warm) = self.warm_parse(need_complexity, fingerprints.as_deref(), None) {
            return self.parsed_parts_from_modules(warm.modules.to_vec(), warm.metrics);
        }
        let ParsedModules {
            modules, metrics, ..
        } = parse_files_with_config(&self.config, self.files(), need_complexity, None);
        self.parsed_parts_from_modules(modules, metrics)
    }

    fn parsed_parts_from_modules(
        &self,
        modules: Vec<ModuleInfo>,
        metrics: core_backend::ParseMetrics,
    ) -> ParsedAnalysisSessionParts {
        ParsedAnalysisSessionParts {
            config: self.config.clone(),
            config_path: self.config_path.clone(),
            files: self.discovery.files().to_vec(),
            modules,
            workspaces: self.workspaces.clone(),
            workspace_diagnostics: self.current_workspace_diagnostics(),
            parse_ms: metrics.parse_ms,
            cache_update_ms: metrics.cache_ms,
            cache_hits: metrics.cache_hits,
            cache_misses: metrics.cache_misses,
            parse_cpu_ms: metrics.parse_cpu_ms,
        }
    }

    /// Run dead-code analysis for this session.
    ///
    /// # Errors
    ///
    /// Returns an error if parsing or analysis fails.
    pub fn analyze_dead_code(&self) -> EngineResult<DeadCodeAnalysis> {
        self.analyze_dead_code_with_artifacts(false, false)
            .map(|output| DeadCodeAnalysis {
                results: output.results,
            })
    }

    /// Run dead-code analysis with retained complexity artifacts.
    ///
    /// # Errors
    ///
    /// Returns an error if parsing or analysis fails.
    pub fn analyze_dead_code_with_complexity(&self) -> EngineResult<DeadCodeAnalysisOutput> {
        self.analyze_dead_code_with_artifacts(true, false)
            .map(|output| DeadCodeAnalysisOutput {
                results: output.results,
                modules: output.modules,
                files: output.files,
            })
    }

    /// Run dead-code analysis with retained modules, discovered files and graph.
    ///
    /// # Errors
    ///
    /// Returns an error if parsing or analysis fails.
    pub fn analyze_dead_code_with_artifacts(
        &self,
        need_complexity: bool,
        retain_graph: bool,
    ) -> EngineResult<DeadCodeAnalysisArtifacts> {
        self.analyze_dead_code_with_shared_artifacts(need_complexity, retain_graph)
            .map(SharedDeadCodeAnalysisArtifacts::into_owned)
    }

    /// Run dead-code analysis with shared immutable parser artifacts.
    ///
    /// Workspace-owned consumers use this additive path to retain warm parser
    /// modules without deep-cloning the session cache. External callers can
    /// continue using [`Self::analyze_dead_code_with_artifacts`].
    ///
    /// # Errors
    ///
    /// Returns an error if parsing or analysis fails.
    #[doc(hidden)]
    pub fn analyze_dead_code_with_shared_artifacts(
        &self,
        need_complexity: bool,
        retain_graph: bool,
    ) -> EngineResult<SharedDeadCodeAnalysisArtifacts> {
        self.analyze_dead_code_with_reuse_artifacts(need_complexity, retain_graph, need_complexity)
    }

    /// Run dead-code analysis while retaining discovered files for downstream
    /// command stages that reuse discovery but do not need parser modules.
    ///
    /// # Errors
    ///
    /// Returns an error if parsing or analysis fails.
    pub fn analyze_dead_code_retaining_files(
        &self,
        need_complexity: bool,
        retain_graph: bool,
    ) -> EngineResult<DeadCodeAnalysisArtifacts> {
        self.analyze_dead_code_with_reuse_artifacts(need_complexity, retain_graph, true)
            .map(SharedDeadCodeAnalysisArtifacts::into_owned)
    }

    /// Run dead-code analysis from modules already parsed through this session.
    ///
    /// This preserves the session's resolved config and discovered file set for
    /// follow-up analyses that reuse parser output without redoing discovery.
    ///
    /// # Errors
    ///
    /// Returns an error if graph construction or analysis fails.
    pub fn analyze_dead_code_with_parsed_modules(
        &self,
        modules: &[ModuleInfo],
    ) -> EngineResult<DeadCodeAnalysisArtifacts> {
        self.analyze_dead_code_with_shared_modules(Arc::from(modules))
    }

    /// Run dead-code analysis from shared immutable parser modules.
    ///
    /// # Errors
    ///
    /// Returns an error if graph construction or analysis fails.
    #[doc(hidden)]
    pub(crate) fn analyze_dead_code_with_shared_modules(
        &self,
        modules: Arc<[ModuleInfo]>,
    ) -> EngineResult<DeadCodeAnalysisArtifacts> {
        run_engine_owned_dead_code_pipeline(EngineDeadCodePipelineInput {
            config: &self.config,
            discovery: &self.discovery,
            modules,
            metrics: reused_parse_metrics(),
            collect_usages: true,
            retain_graph: true,
            retain_modules: false,
            retain_files: false,
            cancellation: self.cancellation.as_deref(),
        })
        .map(SharedDeadCodeAnalysisArtifacts::into_owned)
    }

    fn analyze_dead_code_with_reuse_artifacts(
        &self,
        need_complexity: bool,
        retain_graph: bool,
        retain_files: bool,
    ) -> EngineResult<SharedDeadCodeAnalysisArtifacts> {
        self.ensure_not_cancelled("parsing")?;
        let SharedParsedModules { modules, metrics } =
            self.parse_modules(need_complexity, self.cancellation.as_deref());
        // The parse loop no-ops the files it has not reached yet, so a token
        // set during parsing leaves `modules` truncated. It must never reach
        // the graph as if it were the whole project.
        self.ensure_not_cancelled("the dead-code pipeline")?;
        run_engine_owned_dead_code_pipeline(EngineDeadCodePipelineInput {
            config: &self.config,
            discovery: &self.discovery,
            modules,
            metrics,
            collect_usages: true,
            retain_graph,
            retain_modules: need_complexity,
            retain_files,
            cancellation: self.cancellation.as_deref(),
        })
    }

    /// Run dead-code analysis and return the session-scoped reuse artifacts.
    ///
    /// Callers pass a changed-file set they have already resolved for the
    /// command. The returned value keeps that set beside parser, graph, and
    /// source-fingerprint data so downstream runners do not have to rebuild or
    /// rediscover the same inputs.
    ///
    /// # Errors
    ///
    /// Returns an error if parsing or analysis fails.
    pub fn analyze_dead_code_with_session_artifacts(
        &self,
        need_complexity: bool,
        retain_graph: bool,
        changed_files: Option<FxHashSet<PathBuf>>,
    ) -> EngineResult<AnalysisSessionArtifacts> {
        Ok(AnalysisSessionArtifacts {
            analysis: self.analyze_dead_code_with_artifacts(need_complexity, retain_graph)?,
            changed_files,
            source_fingerprints: self.source_fingerprints(),
        })
    }

    /// Run dead-code and duplication analysis for this session.
    ///
    /// When `retain_complexity_artifacts` is true, the dead-code result keeps
    /// parser artifacts needed by editor overlays such as inline complexity.
    ///
    /// # Errors
    ///
    /// Returns an error if dead-code parsing or analysis fails.
    pub fn analyze_project_with(
        &self,
        duplicates_config: &DuplicatesConfig,
        retain_complexity_artifacts: bool,
    ) -> EngineResult<ProjectAnalysisOutput> {
        self.analyze_project_with_artifacts(
            duplicates_config,
            ProjectAnalysisArtifactOptions {
                retain_complexity_artifacts,
                ..ProjectAnalysisArtifactOptions::default()
            },
        )
        .map(ProjectAnalysisArtifacts::into_output)
    }

    /// Run dead-code and duplication analysis with retained session reuse data.
    ///
    /// This is the engine-owned project artifact boundary for callers that need
    /// to hand one analysis result across audit, decision, editor, or follow-up
    /// analysis surfaces without rediscovering session metadata.
    ///
    /// # Errors
    ///
    /// Returns an error if dead-code parsing or analysis fails.
    pub fn analyze_project_with_artifacts(
        &self,
        duplicates_config: &DuplicatesConfig,
        options: ProjectAnalysisArtifactOptions,
    ) -> EngineResult<ProjectAnalysisArtifacts> {
        self.ensure_not_cancelled("duplication detection")?;
        let cache_dir = (!self.config.no_cache).then_some(self.config.cache_dir.as_path());
        let duplication = if let Some(changed_files) = options.changed_files.as_ref() {
            let changed_files = changed_files.iter().cloned().collect::<Vec<_>>();
            self.find_duplicates_touching_files_with_defaults(
                duplicates_config,
                &changed_files,
                cache_dir,
            )
            .report
        } else {
            self.find_duplicates_with_defaults(duplicates_config, cache_dir)
                .report
        };
        // Duplication detection is infallible, so a token set while it ran can
        // only be reported here, before its report is handed on as a complete
        // one.
        self.ensure_not_cancelled("the dead-code half of project analysis")?;
        let source_fingerprints = options
            .collect_source_fingerprints
            .then(|| self.source_fingerprints());
        Ok(ProjectAnalysisArtifacts {
            dead_code: self.analyze_dead_code_with_artifacts(
                options.retain_complexity_artifacts,
                options.retain_graph,
            )?,
            duplication,
            changed_files: options.changed_files,
            source_fingerprints,
        })
    }

    /// Run duplication detection and return report sidecar metadata.
    #[must_use]
    pub fn find_duplicates_with_defaults(
        &self,
        config: &DuplicatesConfig,
        cache_dir: Option<&Path>,
    ) -> DuplicationAnalysis {
        duplicates::find_duplicates_with_defaults(
            &self.config.root,
            self.files(),
            config,
            cache_dir,
        )
    }

    /// Run focused duplication detection for a changed-file set.
    #[must_use]
    pub fn find_duplicates_touching_files_with_defaults(
        &self,
        config: &DuplicatesConfig,
        changed_files: &[PathBuf],
        cache_dir: Option<&Path>,
    ) -> DuplicationAnalysis {
        duplicates::find_duplicates_touching_files_with_defaults(
            &self.config.root,
            self.files(),
            config,
            changed_files,
            cache_dir,
        )
    }

    /// Parse the discovered files, reusing the session's warm module cache.
    ///
    /// `cancellation` is threaded through to the per-file parse loop, so a set
    /// token truncates the returned modules. Only callers that convert a set
    /// token into an error may pass one.
    fn parse_modules(
        &self,
        need_complexity: bool,
        cancellation: Option<&AtomicBool>,
    ) -> SharedParsedModules {
        let fingerprints = source_fingerprints_for_files(self.files());
        if let Some(fingerprints) = fingerprints.as_ref() {
            if let Some(modules) = self.cached_modules(need_complexity, fingerprints) {
                self.parse_counts.record(SessionParseCounts {
                    modules_reused: modules.len(),
                    ..SessionParseCounts::default()
                });
                return SharedParsedModules {
                    modules,
                    metrics: reused_parse_metrics(),
                };
            }
            if let Some(parsed) =
                self.reparse_changed_modules(need_complexity, fingerprints, cancellation)
            {
                return parsed;
            }
        }

        let (modules, metrics, has_complexity, modules_reused, problems) = if let Some(warm) =
            self.warm_parse(need_complexity, fingerprints.as_deref(), cancellation)
        {
            let reused = if warm.reused { warm.modules.len() } else { 0 };
            (warm.modules, warm.metrics, true, reused, warm.problems)
        } else {
            let ParsedModules {
                modules,
                metrics,
                read_failures,
                parse_degradations,
                ..
            } = parse_files_with_config(&self.config, self.files(), need_complexity, cancellation);
            let problems = SourceProblems {
                read_failures,
                parse_degradations,
            };
            (modules.into(), metrics, need_complexity, 0, problems)
        };
        self.parse_counts.record(SessionParseCounts {
            modules_parsed: metrics.cache_misses,
            disk_cache_hits: metrics.cache_hits,
            modules_reused,
        });
        // A cancelled parse returns a truncated module set. Storing it would
        // serve that truncation to the next call as a warm cache hit, long
        // after the cancellation itself is forgotten.
        if !token_is_set(cancellation)
            && let Some(fingerprints) = fingerprints
            && let Ok(mut cache) = self.parsed_cache.lock()
        {
            *cache = Some(ParsedModuleCache {
                need_complexity: has_complexity,
                fingerprints,
                modules: Arc::clone(&modules),
                read_failures: problems.read_failures,
                parse_degradations: problems.parse_degradations,
            });
            // The full parse wrote the persisted cache for these modules.
            self.disk_cache_stale.store(false, Ordering::SeqCst);
            self.clear_styling_cache();
        }
        SharedParsedModules { modules, metrics }
    }

    /// Parse through the store of parsed modules across sessions.
    ///
    /// Returns `None` when the session has no store, when caching is off, or
    /// when a fingerprint cannot stand in for the file content. The caller
    /// then parses through the persisted cache only.
    ///
    /// The parse always computes complexity. The cost is
    /// about the same, and a later `health` session can then use the modules.
    /// The persisted cache and the in-session cache already serve modules with
    /// complexity to callers that need none. One case costs more: a persisted
    /// cache without complexity, from a CLI `dead-code` run, gives no hits, so
    /// the first parse through the store parses each file from source.
    fn warm_parse(
        &self,
        need_complexity: bool,
        fingerprints: Option<&[SourceFingerprint]>,
        cancellation: Option<&AtomicBool>,
    ) -> Option<WarmParsedModules> {
        let store = self.warm_parse.as_deref()?;
        if self.config.no_cache {
            return None;
        }
        let key = WarmParseKey {
            root: &self.config.root,
            cache_config_hash: self.config.cache_config_hash,
            files: self.files(),
            fingerprints: fingerprints?,
        };
        if !key.is_reusable() {
            return None;
        }
        if let Some(parse) = store.get(&key, need_complexity) {
            // A parse records its read failures and parse degradations for the
            // project, and the session reads them back as workspace
            // diagnostics. Record the kept ones again, so the diagnostics of
            // this run are the same as after a parse.
            record_source_diagnostics(
                &self.config.root,
                &parse.read_failures,
                &parse.parse_degradations,
            );
            return Some(WarmParsedModules {
                modules: parse.modules,
                metrics: reused_parse_metrics(),
                reused: true,
                problems: SourceProblems {
                    read_failures: parse.read_failures.to_vec(),
                    parse_degradations: parse.parse_degradations.to_vec(),
                },
            });
        }

        let parsed = parse_files_with_config(&self.config, self.files(), true, cancellation);
        store.record_parse(parsed.metrics.cache_misses, parsed.metrics.cache_hits);
        let modules: Arc<[ModuleInfo]> = parsed.modules.into();
        if !token_is_set(cancellation) {
            store.put(
                &key,
                true,
                WarmParse {
                    modules: Arc::clone(&modules),
                    read_failures: parsed.read_failures.clone().into(),
                    parse_degradations: parsed.parse_degradations.clone().into(),
                },
            );
        }
        Some(WarmParsedModules {
            modules,
            metrics: parsed.metrics,
            reused: false,
            problems: SourceProblems {
                read_failures: parsed.read_failures,
                parse_degradations: parsed.parse_degradations,
            },
        })
    }

    /// Parse only the files whose fingerprint changed since the cached parse,
    /// and keep the other cached modules.
    ///
    /// Returns `None` when the cache cannot serve the request: no cached
    /// parse, a cache without the requested complexity, a different file set,
    /// or more changed files than [`MAX_INCREMENTAL_REPARSE_FILES`]. The
    /// caller then parses every file through the persisted cache.
    fn reparse_changed_modules(
        &self,
        need_complexity: bool,
        fingerprints: &[SourceFingerprint],
        cancellation: Option<&AtomicBool>,
    ) -> Option<SharedParsedModules> {
        let mut guard = self.parsed_cache.lock().ok()?;
        let cache = guard.as_mut()?;
        if need_complexity && !cache.need_complexity {
            return None;
        }
        let changed = changed_file_indices(&cache.fingerprints, fingerprints)?;
        if changed.is_empty() || changed.len() > MAX_INCREMENTAL_REPARSE_FILES {
            return None;
        }
        let files: Vec<DiscoveredFile> = changed
            .iter()
            .filter_map(|&index| self.files().get(index).cloned())
            .collect();
        let parse_start = Instant::now();
        let parsed = crate::source::parse_all_files(
            &files,
            None,
            cache.need_complexity,
            cancellation,
            &self.config.flags.patterns(),
        );
        // The caller turns a set token into an error. The cache keeps the
        // earlier complete modules, because the parse above may be truncated.
        if token_is_set(cancellation) {
            return Some(SharedParsedModules {
                modules: Arc::clone(&cache.modules),
                metrics: reused_parse_metrics(),
            });
        }
        let mut fresh = parsed.modules;
        for module in &mut fresh {
            module.prepare_analysis_facts();
        }
        let fresh_count = fresh.len();
        let reparsed: Vec<_> = files.iter().map(|file| file.id).collect();
        merge_reparsed_modules(&mut cache.modules, &reparsed, fresh);
        cache.fingerprints = fingerprints.to_vec();
        cache
            .read_failures
            .retain(|failure| !reparsed.contains(&failure.file_id));
        cache.read_failures.extend(parsed.read_failures);
        cache
            .parse_degradations
            .retain(|degradation| !reparsed.contains(&degradation.file_id));
        cache.parse_degradations.extend(parsed.parse_degradations);
        record_source_diagnostics(
            &self.config.root,
            &cache.read_failures,
            &cache.parse_degradations,
        );
        self.parse_counts.record(SessionParseCounts {
            modules_parsed: parsed.cache_misses,
            disk_cache_hits: 0,
            modules_reused: cache.modules.len().saturating_sub(fresh_count),
        });
        let modules = Arc::clone(&cache.modules);
        drop(guard);
        self.disk_cache_stale.store(true, Ordering::SeqCst);
        self.clear_styling_cache();
        Some(SharedParsedModules {
            modules,
            metrics: core_backend::ParseMetrics {
                parse_ms: parse_start.elapsed().as_secs_f64() * 1000.0,
                cache_ms: 0.0,
                cache_hits: 0,
                cache_misses: parsed.cache_misses,
                parse_cpu_ms: parsed.parse_cpu_ms,
                cache_rejection: None,
                files_read: parsed.files_read,
                source_bytes_read: parsed.source_bytes_read,
                parse_cache_bytes_read: 0,
                css_masked_bytes: parsed.css_masked_bytes,
                parse_cache_load_ms: 0.0,
            },
        })
    }

    fn cached_modules(
        &self,
        need_complexity: bool,
        fingerprints: &[SourceFingerprint],
    ) -> Option<Arc<[ModuleInfo]>> {
        let Ok(cache) = self.parsed_cache.lock() else {
            return None;
        };
        let cache = cache.as_ref()?;
        let complexity_mode_satisfies_request = cache.need_complexity || !need_complexity;
        if complexity_mode_satisfies_request && cache.fingerprints == fingerprints {
            return Some(Arc::clone(&cache.modules));
        }
        None
    }
}

struct ParsedModules {
    modules: Vec<ModuleInfo>,
    metrics: core_backend::ParseMetrics,
    source_diagnostics: Vec<WorkspaceDiagnostic>,
    read_failures: Vec<SourceReadFailure>,
    parse_degradations: Vec<SourceParseDegradation>,
}

struct SharedParsedModules {
    modules: Arc<[ModuleInfo]>,
    metrics: core_backend::ParseMetrics,
}

/// Modules from [`AnalysisSession::warm_parse`]. They always have complexity.
struct WarmParsedModules {
    modules: Arc<[ModuleInfo]>,
    metrics: core_backend::ParseMetrics,
    /// The modules came from the store without parse work.
    reused: bool,
    problems: SourceProblems,
}

/// The files of a parse that did not read or that parsed with errors.
struct SourceProblems {
    read_failures: Vec<SourceReadFailure>,
    parse_degradations: Vec<SourceParseDegradation>,
}

fn token_is_set(cancellation: Option<&AtomicBool>) -> bool {
    cancellation.is_some_and(|cancelled| cancelled.load(Ordering::SeqCst))
}

fn parse_files_with_config(
    config: &ResolvedConfig,
    files: &[DiscoveredFile],
    need_complexity: bool,
    cancellation: Option<&AtomicBool>,
) -> ParsedModules {
    let parse_start = Instant::now();
    let cache_max_size_bytes = crate::project_config::resolve_cache_max_size_bytes(config);
    let mut cache_rejection = None;
    let mut parse_cache_bytes_read = 0;
    let mut parse_cache_load_ms = 0.0;
    let mut cache = if config.no_cache {
        None
    } else {
        let load_start = Instant::now();
        let (loaded, bytes_read) = fallow_extract::cache::CacheStore::load_counting_bytes(
            &config.cache_dir,
            &config.root,
            config.cache_config_hash,
            cache_max_size_bytes,
        );
        parse_cache_bytes_read = bytes_read;
        parse_cache_load_ms = load_start.elapsed().as_secs_f64() * 1000.0;
        match loaded {
            Ok(store) => Some(store),
            Err(rejection) => {
                cache_rejection = Some(rejection);
                None
            }
        }
    };
    let parse_result = crate::source::parse_all_files(
        files,
        cache.as_ref(),
        need_complexity,
        cancellation,
        &config.flags.patterns(),
    );
    let source_diagnostics = record_source_diagnostics(
        &config.root,
        &parse_result.read_failures,
        &parse_result.parse_degradations,
    );
    let mut modules = parse_result.modules;
    for module in &mut modules {
        module.prepare_analysis_facts();
    }
    let parse_ms = parse_start.elapsed().as_secs_f64() * 1000.0;
    let cache_ms = if token_is_set(cancellation) {
        0.0
    } else {
        update_parse_cache_if_enabled(config, &mut cache, &modules, files, need_complexity)
    };
    let metrics = core_backend::ParseMetrics {
        parse_ms,
        cache_ms,
        cache_hits: parse_result.cache_hits,
        cache_misses: parse_result.cache_misses,
        parse_cpu_ms: parse_result.parse_cpu_ms,
        cache_rejection,
        files_read: parse_result.files_read,
        source_bytes_read: parse_result.source_bytes_read,
        parse_cache_bytes_read,
        css_masked_bytes: parse_result.css_masked_bytes,
        parse_cache_load_ms,
    };
    ParsedModules {
        modules,
        metrics,
        source_diagnostics,
        read_failures: parse_result.read_failures,
        parse_degradations: parse_result.parse_degradations,
    }
}

/// Record the read failures and parse degradations of a parse for the
/// project, and return them as workspace diagnostics.
fn record_source_diagnostics(
    root: &Path,
    read_failures: &[SourceReadFailure],
    parse_degradations: &[SourceParseDegradation],
) -> Vec<WorkspaceDiagnostic> {
    let mut diagnostics = fallow_config::record_source_read_failures(root, read_failures);
    diagnostics.extend(fallow_config::record_source_parse_degradations(
        root,
        parse_degradations,
    ));
    diagnostics
}

fn reused_parse_metrics() -> core_backend::ParseMetrics {
    core_backend::ParseMetrics {
        parse_ms: 0.0,
        cache_ms: 0.0,
        cache_hits: 0,
        cache_misses: 0,
        parse_cpu_ms: 0.0,
        cache_rejection: None,
        files_read: 0,
        source_bytes_read: 0,
        parse_cache_bytes_read: 0,
        css_masked_bytes: 0,
        parse_cache_load_ms: 0.0,
    }
}

fn source_fingerprints_for_files(files: &[DiscoveredFile]) -> Option<Vec<SourceFingerprint>> {
    files
        .iter()
        .map(|file| {
            std::fs::metadata(&file.path)
                .ok()
                .map(|metadata| SourceFingerprint::from_metadata(&metadata))
                .filter(|fingerprint| fingerprint.has_known_mtime())
        })
        .collect()
}

fn update_parse_cache_if_enabled(
    config: &ResolvedConfig,
    cache: &mut Option<fallow_extract::cache::CacheStore>,
    modules: &[ModuleInfo],
    files: &[DiscoveredFile],
    need_complexity: bool,
) -> f64 {
    let start = Instant::now();
    write_parse_cache(
        config,
        cache,
        &ParseCacheWrite {
            modules,
            files,
            need_complexity,
            fingerprint_of: &|file: &DiscoveredFile| source_fingerprint(&file.path),
        },
    );
    start.elapsed().as_secs_f64() * 1000.0
}

/// Modules to store in the persisted parse cache, with the fingerprint that
/// each file gets in the cache.
struct ParseCacheWrite<'a> {
    modules: &'a [ModuleInfo],
    files: &'a [DiscoveredFile],
    need_complexity: bool,
    fingerprint_of: &'a dyn Fn(&DiscoveredFile) -> SourceFingerprint,
}

fn write_parse_cache(
    config: &ResolvedConfig,
    cache: &mut Option<fallow_extract::cache::CacheStore>,
    write: &ParseCacheWrite<'_>,
) {
    if config.no_cache {
        return;
    }

    let cache_max_size_bytes = crate::project_config::resolve_cache_max_size_bytes(config);
    let store = cache.get_or_insert_with(|| fallow_extract::cache::CacheStore::new(&config.root));
    if update_parse_cache(store, write)
        && let Err(error) = store.save(
            &config.cache_dir,
            config.cache_config_hash,
            cache_max_size_bytes,
        )
    {
        tracing::warn!("Failed to save cache: {error}");
    }
}

/// Mirror of `fallow_core`'s `update_cache` for session-owned parsing: rewrite
/// an unchanged entry only when its metadata moved or when this run can add
/// complexity the entry lacks, and never let a complexity-blind run strip the
/// complexity a `health` run stored.
fn update_parse_cache(
    store: &mut fallow_extract::cache::CacheStore,
    write: &ParseCacheWrite<'_>,
) -> bool {
    let ParseCacheWrite {
        modules,
        files,
        need_complexity,
        fingerprint_of,
    } = *write;
    let mut dirty = false;
    for module in modules {
        if let Some(file) = files.get(module.file_id.0 as usize) {
            let fingerprint = fingerprint_of(file);
            if let Some(cached) = store.get_by_path_only(&file.path)
                && cached.content_hash == module.content_hash
            {
                let stale_metadata = cached.source_fingerprint() != fingerprint;
                let adds_complexity = need_complexity && !cached.complexity_extracted;
                if stale_metadata || adds_complexity {
                    let preserved_last_access = cached.last_access_secs;
                    let preserved_complexity = (!need_complexity && cached.complexity_extracted)
                        .then(|| cached.complexity.clone());
                    let mut refreshed = fallow_extract::cache::module_to_cached(
                        module,
                        fingerprint,
                        need_complexity,
                    );
                    refreshed.last_access_secs = preserved_last_access;
                    if let Some(complexity) = preserved_complexity {
                        refreshed.complexity = complexity;
                        refreshed.complexity_extracted = true;
                    }
                    store.insert(&file.path, refreshed);
                    dirty = true;
                }
                continue;
            }
            store.insert(
                &file.path,
                fallow_extract::cache::module_to_cached(module, fingerprint, need_complexity),
            );
            dirty = true;
        }
    }
    store.retain_paths(files) || dirty
}

fn source_fingerprint(path: &Path) -> SourceFingerprint {
    std::fs::metadata(path).map_or_else(
        |_| SourceFingerprint::new(0, 0),
        |metadata| SourceFingerprint::from_metadata(&metadata),
    )
}

struct EngineDeadCodePipelineInput<'a> {
    config: &'a ResolvedConfig,
    discovery: &'a crate::discover::AnalysisDiscovery,
    modules: Arc<[ModuleInfo]>,
    metrics: core_backend::ParseMetrics,
    collect_usages: bool,
    retain_graph: bool,
    retain_modules: bool,
    retain_files: bool,
    cancellation: Option<&'a AtomicBool>,
}

fn run_engine_owned_dead_code_pipeline(
    input: EngineDeadCodePipelineInput<'_>,
) -> EngineResult<SharedDeadCodeAnalysisArtifacts> {
    let EngineDeadCodePipelineInput {
        config,
        discovery,
        modules,
        metrics,
        collect_usages,
        retain_graph,
        retain_modules,
        retain_files,
        cancellation,
    } = input;
    let stopped = |stage: &str| -> EngineResult<()> {
        if token_is_set(cancellation) {
            return Err(crate::EngineError::cancelled(stage));
        }
        Ok(())
    };
    stopped("the dead-code prelude")?;
    let prelude = core_backend::prepare_dead_code_backend_prelude(config, discovery)?;
    let prelude_timings = prelude.timings();
    stopped("dead-code entry-point discovery")?;
    let entry_points = core_backend::discover_dead_code_entry_points(&prelude);
    stopped("import resolution and graph construction")?;
    let (resolved, graph, graph_cache_rejection) =
        resolve_or_build_dead_code_graph(&prelude, &entry_points, &modules);
    stopped("the dead-code detectors")?;

    let mut detector = core_backend::run_dead_code_detectors(
        &prelude,
        &graph.graph,
        &resolved.project.modules,
        &modules,
        collect_usages,
        &entry_points,
    );
    // Stamp ids on the full set, before any filter removes a finding. The
    // tiebreak suffix depends on the other findings with the same identity.
    crate::dead_code::stamp_finding_ids(&mut detector.results, &config.root);
    // Security ids too, so every surface reads the id this pipeline set.
    crate::security::stamp_security_finding_ids(
        &mut detector.results.security_findings,
        &config.root,
    );
    crate::dead_code::filter_configured_ignored_findings(&mut detector.results, config);
    // The detectors are the longest uninterruptible stage. Without this a
    // token set inside them yields a complete report, so the same request
    // would be answered with results or with `cancelled` depending only on
    // which side of the stage the flip landed.
    stopped("assembling the dead-code report")?;
    let profile =
        core_backend::dead_code_pipeline_profile(core_backend::DeadCodePipelineProfileInput {
            retain_timings: retain_graph,
            prelude: &prelude,
            prelude_timings,
            parse_metrics: metrics,
            module_count: modules.len(),
            entry_points: &entry_points,
            resolved: &resolved,
            graph: &graph,
            detector: &detector,
            file_count: discovery.files().len(),
            workspace_count: discovery.workspaces().len(),
            graph_cache_rejection,
        });
    let script_used_packages = prelude.script_used_packages();
    let trace_provenance = prelude.trace_provenance(&modules);
    prelude.finish();
    let file_hashes = collect_file_hashes(&modules, discovery.files());

    Ok(SharedDeadCodeAnalysisArtifacts {
        results: detector.results,
        timings: profile.timings,
        graph: retain_graph.then_some(graph.graph),
        modules: retain_modules.then_some(modules),
        files: retain_files.then(|| discovery.files().to_vec()),
        script_used_packages,
        trace_provenance,
        file_hashes,
    })
}

/// Reuse the persisted module graph, or rebuild it and carry the reason the
/// persisted one was refused.
///
/// The reason is the third element rather than a discarded `Option`: a warm run
/// that paid for a multi-megabyte decode and reused none of it is the case the
/// perf table exists to explain, and every engine-backed command reaches the
/// pipeline through here.
fn resolve_or_build_dead_code_graph(
    prelude: &core_backend::DeadCodeBackendPrelude,
    entry_points: &core_backend::DeadCodeEntryPoints,
    modules: &[ModuleInfo],
) -> (
    core_backend::DeadCodeResolvedModules,
    core_backend::DeadCodeGraphRun,
    Option<CacheRejection>,
) {
    let rejection =
        match core_backend::try_load_dead_code_graph_cache(prelude, entry_points, modules) {
            Ok((resolved, graph)) => return (resolved, graph, None),
            Err(rejection) => rejection,
        };

    let resolved = core_backend::resolve_dead_code_imports(prelude, modules);
    let graph =
        core_backend::build_dead_code_graph(prelude, &resolved.project, entry_points, modules);
    (resolved, graph, rejection)
}

fn collect_file_hashes(
    modules: &[ModuleInfo],
    files: &[DiscoveredFile],
) -> FxHashMap<PathBuf, u64> {
    modules
        .iter()
        .filter_map(|module| {
            files
                .get(module.file_id.0 as usize)
                .map(|file| (file.path.clone(), module.content_hash))
        })
        .collect()
}

pub(crate) fn analyze_dead_code_with_parse_result_from_config(
    config: &ResolvedConfig,
    modules: &[ModuleInfo],
) -> EngineResult<DeadCodeAnalysisArtifacts> {
    let (workspaces, _diagnostics, workspaces_ms) =
        crate::project_config::collect_workspace_metadata(config)?;
    let discovery = crate::discover::prepare_analysis_discovery_with_workspaces(
        config,
        &workspaces,
        workspaces_ms,
    );
    run_engine_owned_dead_code_pipeline(EngineDeadCodePipelineInput {
        config,
        discovery: &discovery,
        modules: Arc::from(modules),
        metrics: reused_parse_metrics(),
        collect_usages: true,
        retain_graph: true,
        retain_modules: false,
        retain_files: false,
        cancellation: None,
    })
    .map(SharedDeadCodeAnalysisArtifacts::into_owned)
}

#[cfg(test)]
mod tests {
    use std::fmt::Write as _;
    use std::time::Duration;

    use super::*;

    fn session_with_source(source: &str) -> (tempfile::TempDir, AnalysisSession) {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(root.join("src/index.ts"), source).expect("write source");
        let session = AnalysisSession::load_default(root);
        (project, session)
    }

    /// A cancelled session reports the failure and names the boundary it
    /// stopped at, so a caller can tell a stopped run from a broken one.
    #[test]
    fn a_cancelled_session_returns_a_cancellation_error_not_an_empty_result() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(root.join("src/index.ts"), "export const entry = 1;\n").expect("entry");
        std::fs::write(root.join("src/orphan.ts"), "export const orphan = 1;\n").expect("orphan");

        let baseline = AnalysisSession::load_default(root)
            .analyze_dead_code()
            .expect("an uncancelled session analyzes");
        assert!(
            !baseline.results.unused_files.is_empty(),
            "the fixture must have findings, so an empty result would be a plausible wrong answer"
        );

        let error = AnalysisSession::load_default(root)
            .with_cancellation(Arc::new(AtomicBool::new(true)))
            .analyze_dead_code()
            .expect_err("a cancelled session must not return results");
        assert!(error.is_cancelled(), "unexpected error: {error}");
        assert!(error.message().contains("cancelled"));
    }

    /// The engine pipeline is what every CLI command runs, and it reported no
    /// graph-cache reason at all: the loader produced one, the boundary threw
    /// it away, and the profile hardcoded `None`. A warm run that decoded a
    /// multi-megabyte graph and then rebuilt from scratch looked exactly like a
    /// first run, so the row that explains it could never print.
    #[test]
    fn a_refused_graph_cache_names_its_reason_in_the_engine_timings() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(root.join("src/index.ts"), "export const entry = 1;\n").expect("entry");

        let cold = AnalysisSession::load_default(root)
            .analyze_dead_code_with_artifacts(false, true)
            .expect("cold run analyzes");
        assert_eq!(
            cold.timings
                .expect("cold timings retained")
                .graph_cache_rejection,
            Some(CacheRejection::Absent),
            "a first run has no persisted graph to refuse"
        );

        std::fs::write(
            root.join("src/index.ts"),
            "export const entry = 1;\nexport const added = 2;\n",
        )
        .expect("edit the entry");

        let warm = AnalysisSession::load_default(root)
            .analyze_dead_code_with_artifacts(false, true)
            .expect("warm run analyzes");
        assert_eq!(
            warm.timings
                .expect("warm timings retained")
                .graph_cache_rejection,
            Some(CacheRejection::FingerprintChanged),
            "the decoded graph was refused because a file changed, and the run must say so"
        );
    }

    /// A session that is not given a token can never be cancelled, so every
    /// existing caller keeps its current behavior.
    #[test]
    fn a_session_without_a_token_is_never_cancelled() {
        let (_project, session) = session_with_source("export const unused = 1;\n");
        assert!(!session.is_cancelled());
        session
            .analyze_dead_code()
            .expect("a session without a token analyzes");
    }

    /// Files per fixture for the parse-loop tests. Large enough that a parse
    /// takes long enough to be cancelled part way through, small enough to
    /// stay a unit test.
    const PARSE_LOOP_FILES: usize = 400;

    fn parse_loop_project() -> tempfile::TempDir {
        let project = tempfile::tempdir().expect("project");
        let src = project.path().join("src");
        std::fs::create_dir_all(&src).expect("src dir");
        for module in 0..PARSE_LOOP_FILES {
            let mut source = String::new();
            for symbol in 0..20 {
                let _ = writeln!(
                    source,
                    "export const helper{symbol} = (input: number): number => {{\n  \
                     if (input > {symbol}) {{\n    return input * {symbol};\n  }}\n  \
                     return input - {symbol};\n}};"
                );
            }
            std::fs::write(src.join(format!("mod{module}.ts")), source).expect("module");
        }
        project
    }

    /// A platform without ctime, such as Windows, gives fingerprints that
    /// can stand in for the content only after a content check. A kept
    /// session must not serve its modules to the next run on such a
    /// fingerprint, because a same-size edit with a restored mtime keeps it.
    // Unix only: other platforms expose no inode change time, so no
    // fingerprint is trustworthy and a refresh always drops the modules.
    #[cfg(unix)]
    #[test]
    fn a_refresh_drops_modules_whose_fingerprints_need_a_content_check() {
        let (_project, mut session) = session_with_source("export const kept = 1;\n");
        drop(session.parse_modules(false, None));

        assert!(!session.refresh_discovery(), "the file set is the same");
        assert!(
            session.parsed_cache.lock().expect("parse cache").is_some(),
            "fingerprints with a known ctime keep the modules for the next run"
        );

        if let Some(cache) = session
            .parsed_cache
            .get_mut()
            .expect("parse cache")
            .as_mut()
        {
            for fingerprint in &mut cache.fingerprints {
                fingerprint.ctime_ns = 0;
            }
        }
        assert!(!session.refresh_discovery(), "the file set is the same");
        assert!(
            session.parsed_cache.lock().expect("parse cache").is_none(),
            "the next run parses through the persisted cache, which checks the content"
        );
    }

    /// A session over `root` that never reads or writes the on-disk parse
    /// cache, so repeated parses in one test all do the same work.
    fn uncached_session(root: &Path) -> AnalysisSession {
        let mut project_config = crate::project_config::default_project_config(root);
        project_config.config.no_cache = true;
        AnalysisSession::from_config(project_config)
    }

    /// The parse loop is the one place cancellation stops work per item rather
    /// than at a stage boundary, so it is the one place the stop can be
    /// asserted from what was parsed instead of from how long the call took.
    ///
    /// A truncated parse is also the most dangerous thing cancellation
    /// produces: served from a cache it would read as a project with fewer
    /// modules long after the cancellation was forgotten. The second half
    /// asserts it is not retained.
    #[test]
    fn a_cancelled_parse_stops_partway_and_leaves_no_truncated_cache_behind() {
        let project = parse_loop_project();
        let root = project.path();

        // Warm the page cache, so the timed run measures parsing.
        drop(uncached_session(root).parse_modules(false, None));

        let started = Instant::now();
        let full = uncached_session(root).parse_modules(false, None);
        let full_parse = started.elapsed();
        let full_count = full.modules.len();
        assert_eq!(
            full_count, PARSE_LOOP_FILES,
            "the fixture must parse every generated module"
        );
        assert!(
            full_parse >= Duration::from_millis(20),
            "the fixture is too small to cancel part way through: {full_parse:?}"
        );

        // Where the flip lands is a scheduling outcome, so this retries until
        // one attempt lands strictly inside the loop. Every attempt asserts the
        // property; the retry only chooses an attempt that measures the stop
        // part way through rather than at the entry guard.
        let mut partial = None;
        let mut cancelled_session = None;
        for attempt in 1..=6_u32 {
            let token = Arc::new(AtomicBool::new(false));
            // Build the session before arming the watchdog. Discovery runs in
            // the constructor, and a delay spent there would leave the token
            // already set when the loop starts.
            let session = uncached_session(root).with_cancellation(Arc::clone(&token));
            let watchdog = {
                let token = Arc::clone(&token);
                let delay = full_parse * attempt / 6;
                std::thread::spawn(move || {
                    std::thread::sleep(delay);
                    token.store(true, Ordering::SeqCst);
                })
            };
            let cancelled = session.parse_modules(false, Some(&token));
            watchdog.join().expect("watchdog thread");

            let parsed = cancelled.modules.len();
            assert!(
                parsed < full_count,
                "the parse returned all {full_count} modules, so the loop never read the token"
            );
            cancelled_session = Some(session);
            if parsed > 0 {
                partial = Some(parsed);
                break;
            }
        }
        let parsed = partial.expect(
            "no attempt flipped the token while the loop was running, so this never measured a \
             stop part way through",
        );
        assert!(parsed < full_count);

        assert!(
            cancelled_session
                .expect("a cancelled session")
                .parsed_cache
                .lock()
                .expect("parse cache")
                .is_none(),
            "a truncated parse must not be retained as this session's warm cache"
        );
        let recovered = uncached_session(root).parse_modules(false, None);
        assert_eq!(
            recovered.modules.len(),
            full_count,
            "a later uncancelled parse must still see the whole project"
        );
    }

    #[test]
    fn session_retains_workspace_metadata_from_config_load() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::write(
            root.join("package.json"),
            r#"{"name":"root","workspaces":["packages/*"]}"#,
        )
        .expect("write root package");
        std::fs::create_dir_all(root.join("packages/a")).expect("create workspace");
        std::fs::write(
            root.join("packages/a/package.json"),
            r#"{"name":"pkg-a","type":"module"}"#,
        )
        .expect("write workspace package");

        let session = AnalysisSession::load(root, None).expect("session loads");

        assert!(
            session
                .workspaces()
                .iter()
                .any(|workspace| workspace.name == "pkg-a"),
            "session must retain workspace metadata discovered during config load"
        );
    }

    #[test]
    fn finding_ignore_filters_results_without_removing_graph_inputs() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(
            root.join("package.json"),
            r#"{"name":"finding-ignore","devDependencies":{"vitest":"latest"}}"#,
        )
        .expect("write package manifest");
        std::fs::write(
            root.join("vitest.config.ts"),
            "import './src/feature';\nexport default {};\n",
        )
        .expect("write vitest config");
        std::fs::write(
            root.join("src/feature.ts"),
            "export const feature = true;\n",
        )
        .expect("write reachable source");
        std::fs::write(root.join("src/hidden.ts"), "export const hidden = true;\n")
            .expect("write hidden source");

        let unfiltered = AnalysisSession::load(root, None)
            .expect("unfiltered session loads")
            .analyze_dead_code()
            .expect("unfiltered analysis succeeds");
        assert!(
            unfiltered
                .results
                .unused_files
                .iter()
                .any(|finding| finding.file.path.ends_with("src/hidden.ts"))
        );

        std::fs::write(
            root.join(".fallowrc.json"),
            r#"{"ignoreFindings":["src/hidden.ts"]}"#,
        )
        .expect("write fallow config");
        let session = AnalysisSession::load(root, None).expect("filtered session loads");
        let hidden_path = root.join("src/hidden.ts");
        assert!(session.files().iter().any(|file| file.path == hidden_path));

        let filtered = session
            .analyze_dead_code_with_artifacts(false, true)
            .expect("filtered analysis succeeds");
        assert!(
            filtered
                .results
                .unused_files
                .iter()
                .all(|finding| finding.file.path != hidden_path)
        );
        assert!(
            filtered
                .graph
                .as_ref()
                .is_some_and(|graph| graph.module_count() == session.files().len())
        );
    }

    #[test]
    fn finding_ignore_normalizes_separators_and_rejects_outside_paths() {
        use fallow_types::output_dead_code::UnusedFileFinding;
        use fallow_types::results::UnusedFile;

        let project = tempfile::tempdir().expect("project");
        let config = serde_json::from_str::<fallow_config::FallowConfig>(
            r#"{"ignoreFindings":["**/*.ts"]}"#,
        )
        .expect("config parses")
        .resolve(
            project.path().to_path_buf(),
            fallow_config::OutputFormat::Human,
            1,
            true,
            true,
            None,
        );
        let outside = project
            .path()
            .parent()
            .expect("project has parent")
            .join("outside.ts");
        let mut results = AnalysisResults {
            unused_files: vec![
                UnusedFileFinding::with_actions(UnusedFile {
                    path: PathBuf::from(r"src\hidden.ts"),
                }),
                UnusedFileFinding::with_actions(UnusedFile {
                    path: outside.clone(),
                }),
            ],
            ..AnalysisResults::default()
        };

        crate::dead_code::filter_configured_ignored_findings(&mut results, &config);

        assert_eq!(results.unused_files.len(), 1);
        assert_eq!(results.unused_files[0].file.path, outside);
    }

    /// Two files under `src/`, with the default config, so the persisted
    /// parse cache is on.
    #[cfg(unix)]
    fn warm_store_project() -> tempfile::TempDir {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(root.join("src/index.ts"), "export const entry = 1;\n").expect("entry");
        std::fs::write(root.join("src/other.ts"), "export const other = 2;\n").expect("other");
        project
    }

    #[cfg(unix)]
    fn warm_session(root: &Path, store: &Arc<WarmParseStore>) -> AnalysisSession {
        AnalysisSession::load_default(root).with_warm_parse(Some(Arc::clone(store)))
    }

    #[cfg(unix)]
    fn exported_names(modules: &[ModuleInfo]) -> Vec<String> {
        let mut names: Vec<String> = modules
            .iter()
            .flat_map(|module| module.exports.iter().map(|export| export.name.to_string()))
            .collect();
        names.sort();
        names
    }

    #[cfg(unix)]
    #[test]
    fn a_warm_store_serves_a_new_session_without_parse_work() {
        let project = warm_store_project();
        let store = Arc::new(WarmParseStore::new(
            crate::warm_parse::WarmParseLimits::default(),
        ));

        let first = warm_session(project.path(), &store).parse_modules(false, None);
        let second = warm_session(project.path(), &store).parse_modules(false, None);

        assert!(Arc::ptr_eq(&first.modules, &second.modules));
        assert_eq!(second.metrics.cache_hits + second.metrics.cache_misses, 0);
        let counts = store.counts();
        assert_eq!(counts.parse_runs, 1);
        assert_eq!(counts.modules_parsed, 2);
        assert_eq!(counts.modules_reused, 2);
    }

    #[cfg(unix)]
    #[test]
    fn a_warm_store_serves_the_health_parse_of_a_new_session() {
        let project = warm_store_project();
        let store = Arc::new(WarmParseStore::new(
            crate::warm_parse::WarmParseLimits::default(),
        ));

        drop(warm_session(project.path(), &store).parse_modules(false, None));
        let health = warm_session(project.path(), &store).parsed_parts_uncached(true);

        assert_eq!(health.modules.len(), 2);
        assert_eq!(health.cache_hits + health.cache_misses, 0);
        assert_eq!(
            store.counts().parse_runs,
            1,
            "the first parse computed complexity"
        );
    }

    #[cfg(unix)]
    #[test]
    fn an_edit_or_a_new_file_makes_the_next_session_parse_again() {
        let project = warm_store_project();
        let root = project.path();
        let store = Arc::new(WarmParseStore::new(
            crate::warm_parse::WarmParseLimits::default(),
        ));
        drop(warm_session(root, &store).parse_modules(false, None));

        std::fs::write(root.join("src/other.ts"), "export const renamed = 22;\n").expect("edit");
        let edited = warm_session(root, &store).parse_modules(false, None);
        assert_eq!(exported_names(&edited.modules), ["entry", "renamed"]);
        assert_eq!(
            edited.metrics.cache_misses, 1,
            "only the edited file is parsed"
        );

        std::fs::write(root.join("src/added.ts"), "export const added = 3;\n").expect("add");
        let added = warm_session(root, &store).parse_modules(false, None);
        assert_eq!(
            exported_names(&added.modules),
            ["added", "entry", "renamed"]
        );

        let counts = store.counts();
        assert_eq!(counts.parse_runs, 3);
        assert_eq!(counts.modules_reused, 0);
        assert_eq!(store.len(), 2, "each file list keeps its latest parse only");
    }

    #[cfg(unix)]
    #[test]
    fn a_session_without_the_persisted_cache_does_not_use_the_store() {
        let project = warm_store_project();
        let store = Arc::new(WarmParseStore::new(
            crate::warm_parse::WarmParseLimits::default(),
        ));

        drop(
            uncached_session(project.path())
                .with_warm_parse(Some(Arc::clone(&store)))
                .parse_modules(false, None),
        );

        assert_eq!(
            store.counts(),
            crate::warm_parse::WarmParseCounts::default()
        );
        assert!(store.is_empty());
    }

    /// A parse records its read failures for the project. A session that
    /// takes kept modules must record them again, or a run between the two
    /// that cleared them would hide the failure.
    #[cfg(unix)]
    #[test]
    fn a_kept_parse_records_its_read_failures_again() {
        let project = warm_store_project();
        let root = project.path();
        std::fs::write(root.join("src/broken.ts"), [0xff, 0xfe, 0x00]).expect("invalid UTF-8");
        let store = Arc::new(WarmParseStore::new(
            crate::warm_parse::WarmParseLimits::default(),
        ));
        let read_failures = |session: &AnalysisSession| {
            session
                .current_workspace_diagnostics()
                .into_iter()
                .filter(|diagnostic| {
                    matches!(
                        diagnostic.kind,
                        fallow_types::workspace::WorkspaceDiagnosticKind::SourceReadFailure { .. }
                    )
                })
                .count()
        };

        let first = warm_session(root, &store);
        drop(first.parse_modules(false, None));
        assert_eq!(read_failures(&first), 1);

        drop(fallow_config::record_source_read_failures(root, &[]));
        let second = warm_session(root, &store);
        drop(second.parse_modules(false, None));
        assert_eq!(store.counts().modules_reused, 2);
        assert_eq!(read_failures(&second), 1);
    }

    #[test]
    fn warm_parse_cache_reuses_module_storage() {
        let (_project, session) = session_with_source("export function value() { return 1; }\n");
        let first = session.parse_modules(true, None);
        let second = session.parse_modules(false, None);

        assert!(
            Arc::ptr_eq(&first.modules, &second.modules),
            "warm session queries must share parsed module storage"
        );
    }

    #[test]
    fn warm_styling_cache_reuses_artifact_allocation() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::write(root.join("styles.css"), ".button { color: red; }\n")
            .expect("write stylesheet");
        let session = AnalysisSession::load_default(root);

        let first = session.styling_analysis_artifacts();
        let second = session.styling_analysis_artifacts();

        assert!(
            Arc::ptr_eq(&first, &second),
            "warm styling queries must share the cached artifact allocation"
        );
    }

    #[test]
    fn shared_parsed_modules_reuse_public_session_storage() {
        let (_project, session) = session_with_source("export const value = 1;\n");
        let first = session.shared_parsed_modules(true);
        let second = session.shared_parsed_modules(false);

        assert!(Arc::ptr_eq(&first, &second));
    }

    #[test]
    fn shared_parsed_parts_reuse_public_session_storage() {
        let (_project, session) = session_with_source("export const value = 1;\n");
        let cached = session.shared_parsed_modules(true);
        let parts = session.shared_parsed_parts(false);

        assert!(Arc::ptr_eq(&cached, &parts.modules));
    }

    #[test]
    fn warm_complexity_artifacts_reuse_cached_module_storage() {
        let (_project, session) = session_with_source("export function value() { return 1; }\n");
        let cached = session.parse_modules(true, None);
        let artifacts = session
            .analyze_dead_code_with_reuse_artifacts(true, true, false)
            .expect("analysis succeeds");
        let retained = artifacts.modules.expect("complexity modules retained");

        assert!(
            Arc::ptr_eq(&cached.modules, &retained),
            "warm complexity artifacts must share parsed module storage"
        );
    }

    #[test]
    fn shared_and_owned_artifacts_preserve_output_bytes() {
        let (_project, session) = session_with_source(
            "export const used = 1;\nexport const unused = 2;\nconsole.log(used);\n",
        );
        let owned = session
            .analyze_dead_code_with_artifacts(true, true)
            .expect("owned analysis succeeds");
        let shared = session
            .analyze_dead_code_with_shared_artifacts(true, true)
            .expect("shared analysis succeeds");

        assert_eq!(
            serde_json::to_vec(&owned.results).expect("serialize owned results"),
            serde_json::to_vec(&shared.results).expect("serialize shared results")
        );
        assert_eq!(owned.file_hashes, shared.file_hashes);
        assert_eq!(
            owned
                .modules
                .as_deref()
                .unwrap_or_default()
                .iter()
                .map(|module| module.content_hash)
                .collect::<Vec<_>>(),
            shared
                .modules
                .as_deref()
                .unwrap_or_default()
                .iter()
                .map(|module| module.content_hash)
                .collect::<Vec<_>>()
        );
    }

    #[test]
    fn route_loader_whole_use_matches_across_cold_and_warm_sessions() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir_all(root.join("app/routes")).expect("create route directory");
        std::fs::write(
            root.join("package.json"),
            r#"{"name":"route-cache-parity","dependencies":{"react-router":"latest"}}"#,
        )
        .expect("write package manifest");
        std::fs::write(
            root.join("app/routes/home.tsx"),
            r#"
import { useLoaderData } from "react-router";
export function loader() { return { opaque: "value" }; }
export default function Home() {
  const data = useLoaderData<typeof loader>();
  const copy = { ...data };
  return JSON.stringify(copy);
}
"#,
        )
        .expect("write route module");

        let cold_session = AnalysisSession::load(root, None).expect("cold session loads");
        let cold_parse = cold_session.parsed_parts(false);
        assert_eq!(cold_parse.cache_hits, 0, "first parse must be cold");
        let cold = cold_session
            .analyze_dead_code()
            .expect("cold analysis succeeds");

        let warm_session = AnalysisSession::load(root, None).expect("warm session loads");
        let warm_parse = warm_session.parsed_parts(false);
        assert!(
            warm_parse.cache_hits > 0,
            "second session must use disk cache"
        );
        let warm = warm_session
            .analyze_dead_code()
            .expect("warm analysis succeeds");

        assert!(
            cold.results.unused_load_data_keys.is_empty(),
            "cold analysis must abstain for an opaque route-loader use"
        );
        assert_eq!(
            serde_json::to_vec(&cold.results).expect("serialize cold results"),
            serde_json::to_vec(&warm.results).expect("serialize warm results"),
            "warm route-loader analysis must match cold analysis"
        );
    }

    #[test]
    fn replaced_module_coverage_matches_across_cold_and_warm_graph_cache() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(
            root.join("package.json"),
            r#"{"name":"mock-cache-parity","main":"src/index.ts","devDependencies":{"vitest":"latest"}}"#,
        )
        .expect("write package manifest");
        std::fs::write(
            root.join("src/dependency.ts"),
            "export function dependency() { return 'real'; }\n",
        )
        .expect("write dependency");
        std::fs::write(
            root.join("src/wrapper.ts"),
            "import { dependency } from './dependency';\nexport function wrapper() { return dependency(); }\n",
        )
        .expect("write wrapper");
        std::fs::write(
            root.join("src/index.ts"),
            "export { wrapper } from './wrapper';\n",
        )
        .expect("write entry point");
        std::fs::write(
            root.join("src/wrapper.test.ts"),
            r#"
import { vi } from "vitest";
vi.mock("./dependency", () => ({ dependency: () => "mock" }));
import { wrapper } from "./wrapper";
wrapper();
"#,
        )
        .expect("write test");

        let cold_session = AnalysisSession::load(root, None).expect("cold session loads");
        let dependency_id = cold_session
            .files()
            .iter()
            .find(|file| file.path == root.join("src/dependency.ts"))
            .expect("dependency discovered")
            .id;
        let cold = cold_session
            .analyze_dead_code_with_artifacts(false, true)
            .expect("cold analysis succeeds");
        let cold_exports = crate::module_graph::module_value_exports(
            cold.graph.as_ref().expect("cold graph retained"),
        );
        assert!(
            fallow_graph::cache::GraphCacheStore::load(&cold_session.config().cache_dir).is_ok(),
            "cold analysis must persist the graph cache"
        );

        let warm_session = AnalysisSession::load(root, None).expect("warm session loads");
        let warm = warm_session
            .analyze_dead_code_with_artifacts(false, true)
            .expect("warm analysis succeeds");
        let warm_exports = crate::module_graph::module_value_exports(
            warm.graph.as_ref().expect("warm graph retained"),
        );

        let dependency = cold_exports
            .iter()
            .find(|export| export.file_id == dependency_id && export.name == "dependency")
            .expect("dependency export retained");
        assert!(!dependency.test_referenced);
        assert_eq!(warm_exports, cold_exports);
    }

    #[test]
    fn session_parse_surfaces_removed_source_with_sparse_file_ids() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(root.join("package.json"), r#"{"name":"read-failure"}"#)
            .expect("write package manifest");
        for name in ["a.ts", "b.ts", "c.ts"] {
            std::fs::write(
                root.join("src").join(name),
                format!("export const {} = 1;\n", name.replace('.', "_")),
            )
            .expect("write source");
        }
        let session = AnalysisSession::load(root, None).expect("session loads");
        let removed_path = root.join("src/b.ts");
        let removed_id = session
            .files()
            .iter()
            .find(|file| file.path == removed_path)
            .expect("removed source discovered")
            .id;
        std::fs::remove_file(&removed_path).expect("remove source after discovery");

        let parts = session.parsed_parts(false);

        assert!(
            parts
                .modules
                .iter()
                .all(|module| module.file_id != removed_id),
            "unreadable file must not receive a placeholder module"
        );
        let diagnostic = parts
            .workspace_diagnostics
            .iter()
            .find(|diagnostic| diagnostic.kind.id() == "source-read-failure")
            .expect("parsed session parts carry source read failure");
        assert_eq!(diagnostic.path, removed_path);
        assert!(
            session
                .current_workspace_diagnostics()
                .iter()
                .any(|diagnostic| {
                    diagnostic.kind.id() == "source-read-failure" && diagnostic.path == removed_path
                }),
            "session output carries parse-time source diagnostics"
        );
    }

    const MALFORMED_PNPM_WORKSPACE_YAML: &str =
        "catalog:\n  react: ^18.2.0\n{this is\nnot: valid: yaml: at: all\n";
    const VALID_PNPM_WORKSPACE_YAML: &str = "catalog:\n  react: ^18.2.0\n";

    fn has_diagnostic_kind(diagnostics: &[WorkspaceDiagnostic], id: &str) -> bool {
        diagnostics
            .iter()
            .any(|diagnostic| diagnostic.kind.id() == id)
    }

    fn write_single_source_project(root: &Path, manifest: &str) {
        std::fs::create_dir(root.join("src")).expect("create source directory");
        std::fs::write(root.join("package.json"), manifest).expect("write package manifest");
        std::fs::write(root.join("src/index.ts"), "export const value = 1;\n")
            .expect("write source");
    }

    /// Issue #2366: engine sessions (the MCP and LSP path) never re-stash the
    /// registry, so a session created after an earlier analysis in the same
    /// process must not keep that analysis's analysis-stage diagnostic once
    /// the cause is fixed: the analyze pass refreshes the entry and the
    /// session snapshot must not pin it.
    #[test]
    fn later_session_drops_stale_analysis_stage_diagnostic_after_cause_is_fixed() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        write_single_source_project(
            root,
            r#"{"name":"issue-2366-engine-session","private":true}"#,
        );
        std::fs::write(
            root.join("pnpm-workspace.yaml"),
            MALFORMED_PNPM_WORKSPACE_YAML,
        )
        .expect("write malformed workspace yaml");

        let broken = AnalysisSession::load(root, None).expect("session loads");
        broken
            .analyze_dead_code()
            .expect("analysis on the malformed yaml succeeds");
        assert!(
            has_diagnostic_kind(
                &broken.current_workspace_diagnostics(),
                "malformed-pnpm-workspace-yaml"
            ),
            "the first session surfaces the malformed yaml: {:?}",
            broken.current_workspace_diagnostics()
        );

        std::fs::write(root.join("pnpm-workspace.yaml"), VALID_PNPM_WORKSPACE_YAML)
            .expect("fix workspace yaml");

        let fixed = AnalysisSession::load(root, None).expect("session loads");
        fixed
            .analyze_dead_code()
            .expect("analysis on the fixed yaml succeeds");
        let current = fixed.current_workspace_diagnostics();
        assert!(
            !has_diagnostic_kind(&current, "malformed-pnpm-workspace-yaml"),
            "a later session must not keep the stale analysis-stage entry (#2366): {current:?}"
        );
    }

    /// Watch-mode rerun shape (issue #2366): the CLI reloads config, which
    /// re-stashes the workspace-discovery set, and builds a fresh session from
    /// the resolved config before re-analyzing. Once a text `bun.lock` exists
    /// the rerun must drop the bun.lockb skip diagnostic. Regression pin: the
    /// old stash wiped analysis-stage entries instead of preserving them, so
    /// this passes before and after the fix.
    #[test]
    fn watch_style_rerun_drops_bun_lockb_skip_once_text_lockfile_exists() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        write_single_source_project(
            root,
            r#"{"name":"issue-2366-watch-rerun","private":true,"overrides":{"ws":"^8.21.0"}}"#,
        );
        std::fs::write(root.join("bun.lockb"), b"placeholder binary lockfile")
            .expect("write bun.lockb placeholder");
        let config = fallow_config::FallowConfig::default().resolve(
            root.to_path_buf(),
            fallow_config::OutputFormat::Json,
            1,
            true,
            true,
            None,
        );
        let reload_config = || {
            let (_, diagnostics) =
                fallow_config::discover_workspaces_with_diagnostics(root, &config.ignore_patterns)
                    .expect("workspace discovery succeeds");
            fallow_config::stash_workspace_diagnostics(root, diagnostics);
        };

        reload_config();
        let first =
            AnalysisSession::from_resolved_config(config.clone()).expect("first session loads");
        first
            .analyze_dead_code()
            .expect("analysis with bun.lockb only succeeds");
        assert!(
            has_diagnostic_kind(
                &first.current_workspace_diagnostics(),
                "bun-lockb-override-resolution-skipped"
            ),
            "the first run surfaces the bun.lockb skip: {:?}",
            first.current_workspace_diagnostics()
        );

        std::fs::write(
            root.join("bun.lock"),
            r#"{"lockfileVersion":1,"workspaces":{"":{"name":"issue-2366-watch-rerun"}},"packages":{"ws":["ws@8.21.3","",{},"sha512-20"]}}"#,
        )
        .expect("write text bun.lock");

        reload_config();
        let rerun =
            AnalysisSession::from_resolved_config(config.clone()).expect("rerun session loads");
        rerun
            .analyze_dead_code()
            .expect("analysis with the text bun.lock succeeds");
        let current = rerun.current_workspace_diagnostics();
        assert!(
            !has_diagnostic_kind(&current, "bun-lockb-override-resolution-skipped"),
            "the rerun drops the skip once a text bun.lock exists (#2366): {current:?}"
        );
    }

    /// Issue #2366: `current_workspace_diagnostics` reads the registry live so
    /// the parse-stage and analyze-stage entries that land after the session
    /// was created still reach the envelope, but it must not import another
    /// walk's skips along with them.
    ///
    /// Combined mode runs the dead-code and duplication walks on the same root
    /// under `rayon::join` whenever a per-analysis `production` split stops
    /// them from sharing a file list, and each walk replaces the registry's
    /// source-discovery set. A session that read that set back would answer
    /// "whichever walk wrote last", which decides where the other walk's skip
    /// lands in the combined root's union and made the array come out in a
    /// different ORDER between runs of the same command.
    #[test]
    fn session_keeps_its_own_walk_skips_and_ignores_another_walks_registry_write() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        write_single_source_project(
            root,
            r#"{"name":"issue-2366-parallel-walks","private":true}"#,
        );
        std::fs::write(root.join("src/huge.ts"), "// filler\n".repeat(400))
            .expect("write oversized source");
        let mut config = fallow_config::FallowConfig::default().resolve(
            root.to_path_buf(),
            fallow_config::OutputFormat::Json,
            1,
            true,
            true,
            None,
        );
        config.max_file_size_bytes = Some(1024);

        let session = AnalysisSession::from_resolved_config(config).expect("session loads");

        // The state a concurrent walk leaves behind: its own skip in this
        // root's registry entry. It writes that through the registry's
        // replace-in-one-operation call, which an architecture guard reserves
        // for the walk itself, so the append is the stand-in here.
        fallow_config::append_workspace_diagnostics(
            root,
            vec![WorkspaceDiagnostic::new(
                root,
                root.join("src/other-walk-only.ts"),
                fallow_types::workspace::WorkspaceDiagnosticKind::SkippedLargeFile {
                    size_bytes: 4096,
                },
            )],
        );

        let current = session.current_workspace_diagnostics();
        let skipped: Vec<&Path> = current
            .iter()
            .filter(|diagnostic| diagnostic.kind.id() == "skipped-large-file")
            .map(|diagnostic| diagnostic.path.as_path())
            .collect();
        assert_eq!(
            skipped.len(),
            1,
            "the session reports its own walk's skips only: {skipped:?}"
        );
        assert!(
            skipped[0].ends_with("src/huge.ts"),
            "the surviving skip is this walk's own: {skipped:?}"
        );
    }

    /// Issue #2366: a config reload that happens AFTER the analyze pass, with
    /// no further pass to re-record, must not wipe the analysis-stage entry
    /// from the process registry. This is the long-lived-server shape: an MCP
    /// or LSP process analyzes once, a later request reloads config for a
    /// different analysis family, and a session built after that reload still
    /// reads the registry live. Pins the analysis-stage preserve in
    /// `stash_workspace_diagnostics`; without it this session reports nothing.
    #[test]
    fn config_reload_after_the_analyze_pass_keeps_the_bun_lockb_skip_readable() {
        let project = tempfile::tempdir().expect("project");
        let root = project.path();
        write_single_source_project(
            root,
            r#"{"name":"issue-2366-reload-preserve","private":true,"overrides":{"ws":"^8.21.0"}}"#,
        );
        std::fs::write(root.join("bun.lockb"), b"placeholder binary lockfile")
            .expect("write bun.lockb placeholder");
        let config = fallow_config::FallowConfig::default().resolve(
            root.to_path_buf(),
            fallow_config::OutputFormat::Json,
            1,
            true,
            true,
            None,
        );
        let reload_config = || {
            let (_, diagnostics) =
                fallow_config::discover_workspaces_with_diagnostics(root, &config.ignore_patterns)
                    .expect("workspace discovery succeeds");
            fallow_config::stash_workspace_diagnostics(root, diagnostics);
        };

        reload_config();
        let analyzing =
            AnalysisSession::from_resolved_config(config.clone()).expect("session loads");
        analyzing
            .analyze_dead_code()
            .expect("analysis with bun.lockb only succeeds");

        // A later request reloads config for another analysis family and never
        // runs a second dead-code pass.
        reload_config();

        let later = AnalysisSession::from_resolved_config(config.clone()).expect("session loads");
        let current = later.current_workspace_diagnostics();
        assert!(
            has_diagnostic_kind(&current, "bun-lockb-override-resolution-skipped"),
            "the reload must preserve the analysis-stage entry the pass recorded (#2366): \
             {current:?}"
        );
    }
}