wip: pause checkpoint — P6 feature complete, gate interrupted (not for merge)

fuzz/Cargo.toml

@@ -78,6 +78,20 @@ test = false
 doc = false
 bench = false
 
+[[bin]]
+name = "fuzz_extract_pdf"
+path = "fuzz_targets/fuzz_extract_pdf.rs"
+test = false
+doc = false
+bench = false
+
+[[bin]]
+name = "fuzz_extract_xlsx"
+path = "fuzz_targets/fuzz_extract_xlsx.rs"
+test = false
+doc = false
+bench = false
+
 [[bin]]
 name = "fuzz_config_parse"
 path = "fuzz_targets/fuzz_config_parse.rs"

fuzz/fuzz_targets/fuzz_extract_pdf.rs

@@ -0,0 +1,17 @@
+//! The PDF extractor on arbitrary bytes (plan P6): reading text, a page's
+//! pixels and the page count is a refusal or an answer, never a panic or a
+//! hang (D-32: it runs in the server's process).
+#![no_main]
+
+use libfuzzer_sys::fuzz_target;
+
+fuzz_target!(|data: &[u8]| {
+    let _ = ledger_import::extract::pdf_text(data);
+    let _ = ledger_import::extract::pdf_lines(data);
+    if let Ok(pages) = ledger_import::extract::page_count(data) {
+        for page in [0, pages.saturating_sub(1), pages] {
+            let _ = ledger_import::extract::page_png(data, page, 64);
+        }
+    }
+    let _ = ledger_import::extract::text_of(data, "application/pdf", "receipt.pdf");
+});

fuzz/fuzz_targets/fuzz_extract_xlsx.rs

@@ -0,0 +1,15 @@
+//! The XLSX extractor on arbitrary bytes (plan P6): a zip or not, a sheet
+//! or not, it is a value or a refusal, never a panic; a cell's number is
+//! read as the decimal string the XML holds.
+#![no_main]
+
+use libfuzzer_sys::fuzz_target;
+
+fuzz_target!(|data: &[u8]| {
+    let _ = ledger_import::extract::xlsx_text(data);
+    let _ = ledger_import::extract::text_of(
+        data,
+        "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+        "sheet.xlsx",
+    );
+});

fuzz/fuzz_targets/fuzz_store_replay.rs

@@ -24,7 +24,8 @@ use serde_json::{Map, Value, json};
 const TYPES: &[&str] = &[
     "account.added", "account.renamed", "account.tax_set", "asset.depreciation_posted",
     "asset.disposed",
-    "asset.registered", "attachment.withdrawn", "bill.created", "bill.paid", "class.added",
+    "asset.registered", "attachment.annotated", "attachment.text_extracted", "attachment.withdrawn",
+    "bill.created", "bill.paid", "class.added",
     "customer.added", "document.filed", "entry.created", "entry.line_added",
     "entry.line_class_set", "entry.posted", "entry.reversed", "entry.vendor_set",
     "fx.rate_set", "invoice.created", "invoice.payment_received", "journal.tail_repaired",
@@ -40,7 +41,7 @@ const TYPES: &[&str] = &[
     "entry.party_set", "party.added", "party.role_set", "party.tin_set", "party.updated",
     "party.w4_set", "party.w9_set", "payroll.taxes_remitted", "reconciliation.completed",
     "reconciliation.voided", "role.assigned", "statement.recorded", "statement.voided",
-    "project.set",
+    "project.set", "entry.settles",
 ];
 
 const FIELDS: &[&str] = &[