refactor(reader): extract.rs moves into ledger-reader (U0d)

D-61 (Q-9): the writer sheds the typesetter and the crate map is layered
again. The PDF/XLSX reader (extract.rs, plan P6) is now the plain library
crate ledger-reader, moved whole with git mv and no behavior change.
ledger-store depends on it instead of ledger-import, so the store no
longer builds Typst or the fonts; ledger-import re-exports it as
ledger_import::extract so gui, mcp, host and the fuzz targets compile
unchanged, and still never depends on the store. New crate carries only
the dependencies extract.rs imports (hayro, hayro-interpret, kurbo, zip,
quick-xml, thiserror); fuzz/Cargo.lock gains the new package.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HS8WZziDnyxunC23MUxrLp

crates/ledger-import/Cargo.toml

@@ -18,17 +18,8 @@ typst = "0.15.1"
 serde_json = "1"
 typst-layout = "0.15.1"
 qrcode = { version = "0.14.1", default-features = false, features = ["svg"] }
-# Extraction (extract.rs, plan P6): a PDF's text, a page's pixels, an
-# XLSX's cells. hayro (the renderer) and hayro-interpret (the text
-# interpreter) are pure Rust and already the viewer's PDF reader, so the
-# server and the viewer read a receipt through one implementation rather
-# than two; `zip` and `quick-xml` read a spreadsheet's XML. None of them
-# touches money: a number is kept as the decimal string the XML holds.
-hayro = { version = "0.8", features = ["embed-fonts"] }
-hayro-interpret = "0.8"
-kurbo = "0.13"
-zip = { version = "9.0.1", default-features = false, features = ["deflate"] }
-quick-xml = "0.38"
+# The reader (extract.rs moved out: D-61); re-exported as `extract`.
+ledger-reader = { path = "../ledger-reader" }
 # BSD-3-Clause: the IIF's transliteration to ASCII (iif.rs, rules.md B-34).
 # Pinned exactly: a new table would change what a CPA's file says.
 deunicode = { version = "=1.6.0", default-features = false }
@@ -38,4 +29,5 @@ workspace = true
 
 [dev-dependencies]
 proptest = "1"
+zip = { version = "9.0.1", default-features = false, features = ["deflate"] }
 rqrr = { version = "0.11.0", default-features = false }

crates/ledger-import/src/lib.rs

@@ -16,7 +16,8 @@
 pub mod asset_register;
 pub mod bank_csv;
 pub mod chart_csv;
-pub mod extract;
+/// The one reader, now the `ledger-reader` crate (D-61), kept at its old path.
+pub use ledger_reader as extract;
 pub mod fonts;
 pub mod iif;
 pub mod invoice_pdf;

crates/ledger-import/tests/extract.rs

@@ -1,7 +1,7 @@
 //! The one extractor (plan P6 step 1): a PDF's text and pages, and an
 //! XLSX's cells as exact decimal strings.
 //!
-//! Expected behavior comes from `crates/ledger-import/src/extract.rs`'s
+//! Expected behavior comes from `crates/ledger-reader/src/lib.rs`'s
 //! INVARIANTS, which come from plan/served.md P6 step 1 and D-15, D-32 and
 //! D-53.
 

crates/ledger-reader/Cargo.toml

@@ -0,0 +1,23 @@
+[package]
+name = "ledger-reader"
+version = "0.1.0"
+edition = "2024"
+license = "MIT OR Apache-2.0"
+description = "The one reader: a PDF's text and pages, an XLSX's cells as exact decimal strings (plan P6, D-61)."
+
+[dependencies]
+thiserror = "2"
+# Extraction (plan P6): a PDF's text, a page's pixels, an XLSX's cells.
+# hayro (the renderer) and hayro-interpret (the text interpreter) are pure
+# Rust and already the viewer's PDF reader, so the server and the viewer
+# read a receipt through one implementation rather than two; `zip` and
+# `quick-xml` read a spreadsheet's XML. None of them touches money: a
+# number is kept as the decimal string the XML holds.
+hayro = { version = "0.8", features = ["embed-fonts"] }
+hayro-interpret = "0.8"
+kurbo = "0.13"
+zip = { version = "9.0.1", default-features = false, features = ["deflate"] }
+quick-xml = "0.38"
+
+[lints]
+workspace = true

crates/ledger-store/Cargo.toml

@@ -8,11 +8,11 @@ description = "Durable persistence for wanderbooks: a hash-chained JSONL journal
 
 [dependencies]
 ledger-core = { path = "../ledger-core" }
-# The one extractor (plan P6 step 1, D-53): the writer journals a file's
+# The one reader (plan P6 step 1, D-53, D-61: its own crate): the writer journals a file's
 # text at attach time and on `wanderbooks books extract`, so the extractor
 # must be callable here rather than only from the server. It is pure and
 # deterministic (no I/O, no network, no time source).
-ledger-import = { path = "../ledger-import" }
+ledger-reader = { path = "../ledger-reader" }
 rusqlite = { version = "0.32", features = ["bundled"] }
 serde = { version = "1", features = ["derive"] }
 serde_json = "1"

crates/ledger-store/src/store.rs

@@ -45,7 +45,7 @@ use ledger_core::{
     payroll::{PayrollAccounts, payroll_entry_plan, payroll_remit_plan},
     reports,
 };
-use ledger_import::extract::{EXTRACTOR as E_EXTRACTOR, VERSION as E_VERSION};
+use ledger_reader::{EXTRACTOR as E_EXTRACTOR, VERSION as E_VERSION};
 use rusqlite::{Connection, OptionalExtension, params};
 use rust_decimal::Decimal;
 use serde_json::{Value, json};
@@ -368,7 +368,7 @@ pub struct Extraction {
     pub attachment_id: u64,
     /// The text's blob (its SHA-256), when there is text.
     pub text_sha256: Option<String>,
-    /// What read it (`ledger_import::extract::EXTRACTOR`).
+    /// What read it (`ledger_reader::EXTRACTOR`).
     pub extractor: String,
     /// That extractor's version.
     pub version: String,
@@ -4620,7 +4620,7 @@ impl Store {
         if let Some(already) = self.extraction_for(attachment_id, E_EXTRACTOR, E_VERSION)? {
             return Ok(already);
         }
-        let got = ledger_import::extract::extract(bytes, mime_type, file_name);
+        let got = ledger_reader::extract(bytes, mime_type, file_name);
         let text_sha = match &got.text {
             Some(text) => Some(self.blobs.put(text.as_bytes())?.0),
             None => None,
@@ -4644,9 +4644,9 @@ impl Store {
             version: E_VERSION.into(),
             pages: got.pages,
             status: match got.status {
-                ledger_import::extract::ExtractStatus::Ok => ExtractionStatus::Ok,
-                ledger_import::extract::ExtractStatus::None => ExtractionStatus::None,
-                ledger_import::extract::ExtractStatus::Failed => ExtractionStatus::Failed,
+                ledger_reader::ExtractStatus::Ok => ExtractionStatus::Ok,
+                ledger_reader::ExtractStatus::None => ExtractionStatus::None,
+                ledger_reader::ExtractStatus::Failed => ExtractionStatus::Failed,
             },
             seq: line.seq,
         })

crates/ledger-store/tests/attachments.rs

@@ -1471,8 +1471,8 @@ fn extraction_is_journaled_once_per_version() {
     let x = s.extraction(id).unwrap().expect("an extraction record");
     assert_eq!(x.status, ExtractionStatus::Ok);
     assert_eq!(x.pages, 1);
-    assert_eq!(x.extractor, ledger_import::extract::EXTRACTOR);
-    assert_eq!(x.version, ledger_import::extract::VERSION);
+    assert_eq!(x.extractor, ledger_reader::EXTRACTOR);
+    assert_eq!(x.version, ledger_reader::VERSION);
     assert_eq!(
         s.extraction_text(id).unwrap().as_deref(),
         Some(String::from_utf8_lossy(text).as_ref())