feat: replace carquet C library with pure Zig zig-parquet

.gitignore

@@ -2,13 +2,10 @@
 zig-out/
 # SQLite amalgamation — generated during build setup, not committed
 lib/
-# Vendored C source (linenoise, yaml, carquet, compression libs)
+# Vendored C source (linenoise, yaml)
 !lib/linenoise/
 !lib/yaml/
-!lib/carquet/
-!lib/zstd/
-!lib/lz4/
-!lib/zlib/
+!lib/parquet/
 sqlite.zip
 # Nix build output symlink
 result

build.zig

@@ -54,6 +54,16 @@ pub fn build(b: *std.Build) void {
     });
     exe.root_module.addImport("yaml", translate_yaml.createModule());
 
+    // Parquet reader: pure Zig library with pure-Zig compression codecs (zstd, gzip, snappy, lz4, brotli).
+    // Replaces bundled carquet C library + zstd/lz4/zlib C sources.
+    // codecs=zig-only: no C compression sources are compiled (all codecs are pure Zig).
+    const zig_parquet = b.dependency("zig_parquet", .{
+        .target = target,
+        .optimize = optimize,
+        .codecs = "zig-only",
+    });
+    exe.root_module.addImport("zig_parquet", zig_parquet.module("parquet"));
+
     if (bundle_sqlite) {
         exe.root_module.addIncludePath(b.path("lib"));
         exe.root_module.addCSourceFile(.{
@@ -81,63 +91,6 @@ pub fn build(b: *std.Build) void {
     exe.root_module.addCSourceFile(.{ .file = b.path("lib/yaml/writer.c"), .flags = &.{} });
     exe.root_module.addCSourceFile(.{ .file = b.path("lib/yaml/loader.c"), .flags = &.{} });
 
-    // Parquet support via carquet C library (MIT, by Johan Natter).
-    // Compression libs (zstd, lz4, zlib) bundled as C source alongside carquet.
-    // All statically linked — zero runtime deps.
-    // NetBSD: Zig's bundled libc stdio.h uses GCC extensions @cImport can't
-    // parse. Shadow it with a minimal shim that only declares what carquet needs.
-    if (target.result.os.tag == .netbsd) {
-        exe.root_module.addIncludePath(b.path("lib/carquet/netbsd-shim"));
-    }
-
-    exe.root_module.addIncludePath(b.path("lib/carquet/include"));
-    exe.root_module.addIncludePath(b.path("lib/carquet/src"));
-    exe.root_module.addIncludePath(b.path("lib/zstd"));
-    exe.root_module.addIncludePath(b.path("lib/lz4"));
-    exe.root_module.addIncludePath(b.path("lib/zlib"));
-
-    const carquet_src_root = "lib/carquet/src";
-    const carquet_flags = &.{ "-std=gnu11", "-D_GNU_SOURCE" };
-    inline for (.{
-        "core/arena.c", "core/allocator.c", "core/buffer.c",
-        "core/bitpack.c", "core/endian.c", "core/error.c", "core/geo_wkb.c",
-        "thrift/thrift_decode.c", "thrift/thrift_encode.c", "thrift/parquet_types.c",
-        "encoding/plain.c", "encoding/rle.c", "encoding/delta.c",
-        "encoding/delta_length.c", "encoding/delta_strings.c",
-        "encoding/dictionary.c", "encoding/byte_stream_split.c",
-        "compression/lz4.c", "compression/snappy.c", "compression/zstd.c",
-        "compression/gzip.c", "compression/custom.c",
-        "simd/detect.c", "simd/dispatch.c",
-        "reader/file_reader.c", "reader/batch_reader.c", "reader/column_reader.c",
-        "reader/page_reader.c", "reader/row_group_reader.c",
-        "reader/mmap_reader.c", "reader/statistics.c", "reader/page_filter.c",
-        "reader/worker_pool.c", "reader/arrow_c_export.c",
-        "reader/arrow_c_read.c", "reader/arrow_schema_read.c",
-        "writer/file_writer.c", "writer/row_group_writer.c", "writer/column_writer.c",
-        "writer/page_writer.c", "writer/arrow_schema.c", "writer/arrow_c_import.c",
-        "metadata/schema.c", "metadata/bloom_filter.c", "metadata/page_index.c",
-        "util/crc32.c", "util/xxhash.c",
-    }) |src_file| {
-        exe.root_module.addCSourceFile(.{
-            .file = b.path(carquet_src_root ++ "/" ++ src_file),
-            .flags = carquet_flags,
-        });
-    }
-
-    // Bundled compression libraries (C source, cross-compiles everywhere)
-    exe.root_module.addCSourceFile(.{ .file = b.path("lib/zstd/zstd.c"), .flags = &.{"-std=gnu11"} });
-    exe.root_module.addCSourceFile(.{ .file = b.path("lib/lz4/lz4.c"), .flags = &.{"-std=gnu11"} });
-    inline for (.{
-        "adler32.c", "compress.c", "crc32.c", "deflate.c",
-        "infback.c", "inffast.c",
-        "inflate.c", "inftrees.c", "trees.c", "uncompr.c", "zutil.c",
-    }) |zf| {
-        exe.root_module.addCSourceFile(.{ .file = b.path("lib/zlib/" ++ zf), .flags = &.{"-std=gnu11"} });
-    }
-
-    exe.root_module.linkSystemLibrary("pthread", .{});
-    exe.root_module.linkSystemLibrary("m", .{});
-
     b.installArtifact(exe);
 
     // Generate man page from scdoc source if scdoc (and gzip) are available (optional dependencies)
@@ -3708,6 +3661,18 @@ pub fn build(b: *std.Build) void {
         test_parquet_columns.step.dependOn(b.getInstallStep());
         test_step.dependOn(&test_parquet_columns.step);
 
+        // Integration test 206g: Parquet INT96 legacy timestamps → TEXT column with ISO values
+        const test_parquet_int96 = b.addSystemCommand(&.{
+            "bash", "-c",
+            \\result=$(./zig-out/bin/sql-pipe tests/fixtures/int96.parquet 'SELECT ts FROM int96 ORDER BY ts')
+            \\expected=$(printf '2023-06-01 00:00:00\n2024-01-15 10:30:00')
+            \\[ "$result" = "$expected" ]
+            \\schema=$(./zig-out/bin/sql-pipe tests/fixtures/int96.parquet --schema)
+            \\echo "$schema" | grep -q '"ts" TEXT'
+        });
+        test_parquet_int96.step.dependOn(b.getInstallStep());
+        test_step.dependOn(&test_parquet_int96.step);
+
         // Fuzzing tests: malformed Parquet files must not crash
         const test_parquet_fuzz_truncated = b.addSystemCommand(&.{
             "bash", "-c",

build.zig.zon

@@ -4,5 +4,9 @@
     .fingerprint = 0xf649b9ac95d768ab,
     .minimum_zig_version = "0.16.0",
     .paths = .{"."},
-    .dependencies = .{},
-}
+    .dependencies = .{
+        .zig_parquet = .{
+            .path = "lib/parquet",
+        },
+    },
+}

codemap.md

@@ -2,12 +2,12 @@
 
 ## Project Responsibility
 
-CLI tool that pipes structured data (CSV, TSV, JSON, NDJSON, XML, YAML, Parquet) into an in-memory SQLite engine, runs a user-supplied SQL query, and emits results in eight formats (CSV, TSV, JSON, NDJSON, XML, Markdown, HTML table, SQL INSERT, pretty-printed table). Also provides ancillary modes for column listing, validation, sampling, statistics, schema DDL generation, and fused `--inspect` — plus a native interactive `--repl` — and shell completion for bash/zsh/fish. Single binary, zero external dependencies, bundles SQLite amalgamation, libyaml subset, and carquet (Parquet) C library.
+CLI tool that pipes structured data (CSV, TSV, JSON, NDJSON, XML, YAML, Parquet) into an in-memory SQLite engine, runs a user-supplied SQL query, and emits results in eight formats (CSV, TSV, JSON, NDJSON, XML, Markdown, HTML table, SQL INSERT, pretty-printed table). Also provides ancillary modes for column listing, validation, sampling, statistics, schema DDL generation, and fused `--inspect` — plus a native interactive `--repl` — and shell completion for bash/zsh/fish. Single binary, zero external dependencies, bundles SQLite amalgamation, libyaml subset, and zig-parquet (Parquet) pure Zig library.
 
 ## System Entry Points
 
 - `src/main.zig` — CLI entry point, argument parsing, mode dispatch, pipeline orchestration
-- `build.zig` — Zig build system with 120+ integration tests, bundles C deps (sqlite3, libyaml, carquet)
+- `build.zig` — Zig build system with 120+ integration tests, bundles C deps (sqlite3, libyaml)
 - `build.zig.zon` — Package manifest (name=`sql_pipe`, version=`0.0.0-dev`, min Zig `0.16.0`)
 
 ## Directory Map
@@ -16,7 +16,7 @@ CLI tool that pipes structured data (CSV, TSV, JSON, NDJSON, XML, YAML, Parquet)
 |-----------|---------------|--------------|
 | `src/` | Core pipeline: argument parsing, multi-format I/O loaders (incl. Parquet), SQLite wrappers, output formatters (15 modules) | [View Map](src/codemap.md) |
 | `src/modes/` | CLI sub-command modes: `--inspect` (fused columns/validate/sample/stats/schema), `--repl`, legacy flags (8 modules) | [View Map](src/modes/codemap.md) |
-| `lib/` | Vendored C deps: SQLite amalgamation (`sqlite3.c/h`), libyaml subset, carquet (Parquet) | (vendored) |
+| `lib/` | Vendored C deps: SQLite amalgamation (`sqlite3.c/h`), libyaml subset, zig-parquet (Parquet) | (vendored) |
 | `tests/` | Test fixtures (CSV, JSON, NDJSON, XML sample data) + HTTP test server | (fixtures) |
 | `docs/` | Man page source (`sql-pipe.1.scd`) | — |
 | `packaging/` | nfpm, winget packaging configs | — |
@@ -65,7 +65,7 @@ CLI args → parseArgs() → dispatch
 | NDJSON | `.ndjson` | `json.zig` | Newline-delimited, one object per line |
 | XML | `.xml` | `xml.zig` | Custom streaming parser, configurable container/row elements |
 | YAML | `.yaml` | `yaml.zig` | Sequence of mappings via libyaml FFI |
-| Parquet | `.parquet` | `parquet.zig` | Columnar via carquet FFI, batch inserts, logical-type conversion |
+| Parquet | `.parquet` | `parquet.zig` | Columnar via zig-parquet DynamicReader, batch inserts, logical-type conversion |
 
 ## Output Formats
 
@@ -82,9 +82,9 @@ CSV, TSV, JSON (array), NDJSON, XML, Markdown table, HTML table, SQL INSERT, pre
 
 ## Integration Points
 
-- **FFI**: SQLite3 C API (`sqlite3_open`, `sqlite3_prepare_v2`, `sqlite3_step`, etc.), libyaml C API (`yaml_parser_parse`, etc.), carquet C API (Parquet reading, logical-type conversion)
+- **FFI**: SQLite3 C API (`sqlite3_open`, `sqlite3_prepare_v2`, `sqlite3_step`, etc.), libyaml C API (`yaml_parser_parse`, etc.)
 - **HTTP**: `std.http.Client` for HTTPS URL input sources (`http.zig`)
-- **Build**: `c` module (SQLite + libyaml + carquet C bindings), `yaml` module (libyaml Zig bindings), `build_options.VERSION`
+- **Build**: `c` module (SQLite + libyaml C bindings), `yaml` module (libyaml Zig bindings), `zig_parquet` module (zig-parquet pure Zig), `build_options.VERSION`
 
 ## Build & Test