docpler 0.1.5__tar.gz

docpler-0.1.5/Cargo.lock

@@ -0,0 +1,352 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 4
+
+[[package]]
+name = "adler2"
+version = "2.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "320119579fcad9c21884f5c4861d16174d0e06250625266f50fe6898340abefa"
+
+[[package]]
+name = "autocfg"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c08606f8c3cbf4ce6ec8e28fb0014a2c086708fe954eaa885384a6165172e7e8"
+
+[[package]]
+name = "bumpalo"
+version = "3.20.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5d20789868f4b01b2f2caec9f5c4e0213b41e3e5702a50157d699ae31ced2fcb"
+
+[[package]]
+name = "byteorder"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1fd0f2584146f6f2ef48085050886acf353beff7305ebd1ae69500e27c67f64b"
+
+[[package]]
+name = "cfb"
+version = "0.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d8a4f8e55be323b378facfcf1f06aa97f6ec17cf4ac84fb17325093aaf62da41"
+dependencies = [
+ "byteorder",
+ "fnv",
+ "uuid",
+]
+
+[[package]]
+name = "cfg-if"
+version = "1.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9330f8b2ff13f34540b44e946ef35111825727b38d33286ef986142615121801"
+
+[[package]]
+name = "crc32fast"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9481c1c90cbf2ac953f07c8d4a58aa3945c425b7185c9154d67a65e4230da511"
+dependencies = [
+ "cfg-if",
+]
+
+[[package]]
+name = "docpler-bindings"
+version = "0.1.0"
+dependencies = [
+ "docpler-core",
+ "pyo3",
+]
+
+[[package]]
+name = "docpler-core"
+version = "0.1.0"
+dependencies = [
+ "cfb",
+ "flate2",
+ "thiserror",
+]
+
+[[package]]
+name = "docpler-premium"
+version = "0.1.0"
+dependencies = [
+ "docpler-core",
+]
+
+[[package]]
+name = "flate2"
+version = "1.1.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "843fba2746e448b37e26a819579957415c8cef339bf08564fe8b7ddbd959573c"
+dependencies = [
+ "crc32fast",
+ "miniz_oxide",
+]
+
+[[package]]
+name = "fnv"
+version = "1.0.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3f9eec918d3f24069decb9af1554cad7c880e2da24a9afd88aca000531ab82c1"
+
+[[package]]
+name = "heck"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2304e00983f87ffb38b55b444b5e3b60a884b5d30c0fca7d82fe33449bbe55ea"
+
+[[package]]
+name = "indoc"
+version = "2.0.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "79cf5c93f93228cf8efb3ba362535fb11199ac548a09ce117c9b1adc3030d706"
+dependencies = [
+ "rustversion",
+]
+
+[[package]]
+name = "js-sys"
+version = "0.3.92"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "cc4c90f45aa2e6eacbe8645f77fdea542ac97a494bcd117a67df9ff4d611f995"
+dependencies = [
+ "once_cell",
+ "wasm-bindgen",
+]
+
+[[package]]
+name = "libc"
+version = "0.2.183"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b5b646652bf6661599e1da8901b3b9522896f01e736bad5f723fe7a3a27f899d"
+
+[[package]]
+name = "memoffset"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "488016bfae457b036d996092f6cb448677611ce4449e970ceaf42695203f218a"
+dependencies = [
+ "autocfg",
+]
+
+[[package]]
+name = "miniz_oxide"
+version = "0.8.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1fa76a2c86f704bdb222d66965fb3d63269ce38518b83cb0575fca855ebb6316"
+dependencies = [
+ "adler2",
+ "simd-adler32",
+]
+
+[[package]]
+name = "once_cell"
+version = "1.21.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9f7c3e4beb33f85d45ae3e3a1792185706c8e16d043238c593331cc7cd313b50"
+
+[[package]]
+name = "portable-atomic"
+version = "1.13.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c33a9471896f1c69cecef8d20cbe2f7accd12527ce60845ff44c153bb2a21b49"
+
+[[package]]
+name = "proc-macro2"
+version = "1.0.106"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8fd00f0bb2e90d81d1044c2b32617f68fcb9fa3bb7640c23e9c748e53fb30934"
+dependencies = [
+ "unicode-ident",
+]
+
+[[package]]
+name = "pyo3"
+version = "0.22.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f402062616ab18202ae8319da13fa4279883a2b8a9d9f83f20dbade813ce1884"
+dependencies = [
+ "cfg-if",
+ "indoc",
+ "libc",
+ "memoffset",
+ "once_cell",
+ "portable-atomic",
+ "pyo3-build-config",
+ "pyo3-ffi",
+ "pyo3-macros",
+ "unindent",
+]
+
+[[package]]
+name = "pyo3-build-config"
+version = "0.22.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b14b5775b5ff446dd1056212d778012cbe8a0fbffd368029fd9e25b514479c38"
+dependencies = [
+ "once_cell",
+ "target-lexicon",
+]
+
+[[package]]
+name = "pyo3-ffi"
+version = "0.22.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9ab5bcf04a2cdcbb50c7d6105de943f543f9ed92af55818fd17b660390fc8636"
+dependencies = [
+ "libc",
+ "pyo3-build-config",
+]
+
+[[package]]
+name = "pyo3-macros"
+version = "0.22.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0fd24d897903a9e6d80b968368a34e1525aeb719d568dba8b3d4bfa5dc67d453"
+dependencies = [
+ "proc-macro2",
+ "pyo3-macros-backend",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "pyo3-macros-backend"
+version = "0.22.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "36c011a03ba1e50152b4b394b479826cad97e7a21eb52df179cd91ac411cbfbe"
+dependencies = [
+ "heck",
+ "proc-macro2",
+ "pyo3-build-config",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "quote"
+version = "1.0.45"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "41f2619966050689382d2b44f664f4bc593e129785a36d6ee376ddf37259b924"
+dependencies = [
+ "proc-macro2",
+]
+
+[[package]]
+name = "rustversion"
+version = "1.0.22"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b39cdef0fa800fc44525c84ccb54a029961a8215f9619753635a9c0d2538d46d"
+
+[[package]]
+name = "simd-adler32"
+version = "0.3.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "703d5c7ef118737c72f1af64ad2f6f8c5e1921f818cdcb97b8fe6fc69bf66214"
+
+[[package]]
+name = "syn"
+version = "2.0.117"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e665b8803e7b1d2a727f4023456bbbbe74da67099c585258af0ad9c5013b9b99"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "unicode-ident",
+]
+
+[[package]]
+name = "target-lexicon"
+version = "0.12.16"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "61c41af27dd6d1e27b1b16b489db798443478cef1f06a660c96db617ba5de3b1"
+
+[[package]]
+name = "thiserror"
+version = "1.0.69"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b6aaf5339b578ea85b50e080feb250a3e8ae8cfcdff9a461c9ec2904bc923f52"
+dependencies = [
+ "thiserror-impl",
+]
+
+[[package]]
+name = "thiserror-impl"
+version = "1.0.69"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4fee6c4efc90059e10f81e6d42c60a18f76588c3d74cb83a0b242a2b6c7504c1"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "unicode-ident"
+version = "1.0.24"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
+
+[[package]]
+name = "unindent"
+version = "0.2.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7264e107f553ccae879d21fbea1d6724ac785e8c3bfc762137959b5802826ef3"
+
+[[package]]
+name = "uuid"
+version = "1.23.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5ac8b6f42ead25368cf5b098aeb3dc8a1a2c05a3eee8a9a1a68c640edbfc79d9"
+dependencies = [
+ "js-sys",
+ "wasm-bindgen",
+]
+
+[[package]]
+name = "wasm-bindgen"
+version = "0.2.115"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6523d69017b7633e396a89c5efab138161ed5aafcbc8d3e5c5a42ae38f50495a"
+dependencies = [
+ "cfg-if",
+ "once_cell",
+ "rustversion",
+ "wasm-bindgen-macro",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-macro"
+version = "0.2.115"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4e3a6c758eb2f701ed3d052ff5737f5bfe6614326ea7f3bbac7156192dc32e67"
+dependencies = [
+ "quote",
+ "wasm-bindgen-macro-support",
+]
+
+[[package]]
+name = "wasm-bindgen-macro-support"
+version = "0.2.115"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "921de2737904886b52bcbb237301552d05969a6f9c40d261eb0533c8b055fedf"
+dependencies = [
+ "bumpalo",
+ "proc-macro2",
+ "quote",
+ "syn",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-shared"
+version = "0.2.115"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a93e946af942b58934c604527337bad9ae33ba1d5c6900bbb41c2c07c2364a93"
+dependencies = [
+ "unicode-ident",
+]

docpler-0.1.5/Cargo.toml

@@ -0,0 +1,9 @@
+[workspace]
+members = ["crates/docpler-core", "crates/docpler-bindings"]
+resolver = "2"
+
+[profile.release]
+opt-level = 3
+strip = true
+lto = true
+codegen-units = 1

docpler-0.1.5/PKG-INFO

@@ -0,0 +1,62 @@
+Metadata-Version: 2.4
+Name: docpler
+Version: 0.1.5
+Requires-Dist: markitdown>=0.0.1 ; extra == 'markitdown'
+Provides-Extra: markitdown
+Summary: Document processing tool - converts HWP (and more) to Markdown
+License: MIT
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown; charset=UTF-8; variant=GFM
+
+# docpler
+
+HWP(한글 워드프로세서) 등 문서 파일을 Markdown으로 변환하는 Python 패키지.
+Rust 코어 기반으로 빠르고 정확한 파싱을 제공합니다.
+
+## 지원 포맷
+
+| 포맷 | 읽기 | 출력 |
+|------|------|------|
+| HWP 5.0 | ✅ | Markdown |
+
+## 설치
+
+```bash
+pip install docpler
+```
+
+## 사용법
+
+### 기본 사용
+
+```python
+from docpler.hwp import convert
+
+markdown = convert("document.hwp")
+print(markdown)
+```
+
+### markitdown 플러그인
+
+```python
+from markitdown import MarkItDown
+
+md = MarkItDown(enable_plugins=True)
+result = md.convert("document.hwp")
+print(result.text_content)
+```
+
+## 라이선스
+
+이 프로젝트는 MIT 라이선스로 제공되며, 현재 버전의 사용에 별도의 제약은 없습니다.
+
+- **Python 래퍼 코드**: 오픈소스 (MIT)
+- **Rust 코어 엔진**: 컴파일된 바이너리로 배포되며, 소스 코드는 비공개입니다.
+
+## HWP 포맷 관련 고지
+
+본 제품은 한글과컴퓨터의 한글 문서 파일(.hwp) 공개 문서를 참고하여 개발하였습니다.
+
+HWP 파일 포맷 공개 문서의 저작권은 (주)한글과컴퓨터에 있으며,
+공개 문서의 전문은 [한글과컴퓨터 공식 페이지](https://www.hancom.com/etc/hwpDownload.do)에서 확인할 수 있습니다.
+

docpler-0.1.5/PYPI_README.md

@@ -0,0 +1,51 @@
+# docpler
+
+HWP(한글 워드프로세서) 등 문서 파일을 Markdown으로 변환하는 Python 패키지.
+Rust 코어 기반으로 빠르고 정확한 파싱을 제공합니다.
+
+## 지원 포맷
+
+| 포맷 | 읽기 | 출력 |
+|------|------|------|
+| HWP 5.0 | ✅ | Markdown |
+
+## 설치
+
+```bash
+pip install docpler
+```
+
+## 사용법
+
+### 기본 사용
+
+```python
+from docpler.hwp import convert
+
+markdown = convert("document.hwp")
+print(markdown)
+```
+
+### markitdown 플러그인
+
+```python
+from markitdown import MarkItDown
+
+md = MarkItDown(enable_plugins=True)
+result = md.convert("document.hwp")
+print(result.text_content)
+```
+
+## 라이선스
+
+이 프로젝트는 MIT 라이선스로 제공되며, 현재 버전의 사용에 별도의 제약은 없습니다.
+
+- **Python 래퍼 코드**: 오픈소스 (MIT)
+- **Rust 코어 엔진**: 컴파일된 바이너리로 배포되며, 소스 코드는 비공개입니다.
+
+## HWP 포맷 관련 고지
+
+본 제품은 한글과컴퓨터의 한글 문서 파일(.hwp) 공개 문서를 참고하여 개발하였습니다.
+
+HWP 파일 포맷 공개 문서의 저작권은 (주)한글과컴퓨터에 있으며,
+공개 문서의 전문은 [한글과컴퓨터 공식 페이지](https://www.hancom.com/etc/hwpDownload.do)에서 확인할 수 있습니다.

docpler-0.1.5/crates/docpler-bindings/Cargo.toml

@@ -0,0 +1,13 @@
+[package]
+name = "docpler-bindings"
+version = "0.1.0"
+edition = "2021"
+description = "Python (PyO3) bindings for docpler"
+
+[lib]
+name = "_docpler"
+crate-type = ["cdylib", "rlib"]
+
+[dependencies]
+docpler-core = { path = "../docpler-core" }
+pyo3 = { version = "0.22", features = ["extension-module"] }

docpler-0.1.5/crates/docpler-bindings/src/lib.rs

@@ -0,0 +1,23 @@
+use pyo3::prelude::*;
+
+/// HWP 파일을 Markdown 문자열로 변환한다.
+#[pyfunction]
+fn hwp_to_markdown(path: &str) -> PyResult<String> {
+    let doc = docpler_core::hwp::parse(path)
+        .map_err(|e| pyo3::exceptions::PyRuntimeError::new_err(e.to_string()))?;
+    Ok(docpler_core::markdown::render(&doc))
+}
+
+/// 디버그용: 섹션 레코드의 (tag_id, level, size) 목록을 반환한다.
+#[pyfunction]
+fn hwp_debug_records(path: &str) -> PyResult<Vec<(u16, u16, usize)>> {
+    docpler_core::hwp::debug_records(path)
+        .map_err(|e| pyo3::exceptions::PyRuntimeError::new_err(e.to_string()))
+}
+
+#[pymodule]
+fn _docpler(m: &Bound<'_, PyModule>) -> PyResult<()> {
+    m.add_function(wrap_pyfunction!(hwp_to_markdown, m)?)?;
+    m.add_function(wrap_pyfunction!(hwp_debug_records, m)?)?;
+    Ok(())
+}

docpler-0.1.5/crates/docpler-core/Cargo.toml

@@ -0,0 +1,10 @@
+[package]
+name = "docpler-core"
+version = "0.1.0"
+edition = "2021"
+description = "Core document parsing library (HWP and more)"
+
+[dependencies]
+cfb = "0.10"
+flate2 = "1.0"
+thiserror = "1.0"

docpler-0.1.5/crates/docpler-core/src/error.rs

@@ -0,0 +1,21 @@
+use thiserror::Error;
+
+#[derive(Error, Debug)]
+pub enum DocplerError {
+    #[error("IO error: {0}")]
+    Io(#[from] std::io::Error),
+
+    #[error("Invalid HWP signature")]
+    InvalidSignature,
+
+    #[error("Password-protected HWP files are not supported")]
+    EncryptedFile,
+
+    #[error("Compound file error: {0}")]
+    Cfb(String),
+
+    #[error("Parse error: {0}")]
+    Parse(String),
+}
+
+pub type Result<T> = std::result::Result<T, DocplerError>;

docpler-0.1.5/crates/docpler-core/src/hwp/mod.rs

@@ -0,0 +1,591 @@
+use std::io::Read;
+
+use cfb::CompoundFile;
+use flate2::read::DeflateDecoder;
+
+use crate::error::{DocplerError, Result};
+use crate::reader::ByteReader;
+
+// ─── HWP 파일 시그니처 ───────────────────────────────────────────────────────
+const HWP_SIGNATURE: &[u8] = b"HWP Document File";
+
+// ─── 레코드 태그 ID (HWP 5.0 규격) ──────────────────────────────────────────
+const TAG_PARA_TEXT: u16 = 0x43;   // 문단 텍스트
+const TAG_CTRL_HEADER: u16 = 0x47; // 컨트롤 헤더
+const TAG_LIST_HEADER: u16 = 0x48; // 목록 헤더 (표 셀)
+const TAG_TABLE: u16 = 0x4D;       // 표 정보
+const TAG_EQEDIT: u16 = 0x58;      // 수식 정보
+
+// ─── 컨트롤 타입 ID ──────────────────────────────────────────────────────────
+// Java: CtrlID.make(a,b,c,d) = (a<<24)|(b<<16)|(c<<8)|d → LE로 파일에 저장
+const CTRL_TABLE: u32 = 0x74626C20;    // make('t','b','l',' ')
+const CTRL_EQUATION: u32 = 0x65716564; // make('e','q','e','d')
+const CTRL_GSO: u32 = 0x67736F20;      // make('g','s','o',' ') - 그리기 개체 (글상자 포함)
+
+// ─── 그리기 개체 태그 ────────────────────────────────────────────────────────
+const TAG_SHAPE_COMPONENT: u16 = 0x4C; // 개체 공통 속성
+
+// ─── 문서 모델 ───────────────────────────────────────────────────────────────
+
+pub struct Document {
+    pub blocks: Vec<Block>,
+}
+
+pub enum Block {
+    Paragraph(String),
+    Table(Vec<Vec<String>>),
+    /// HWP EQN 스크립트 (한글 수식 편집기 포맷)
+    Equation(String),
+}
+
+// ─── 내부 레코드 ─────────────────────────────────────────────────────────────
+
+struct Record {
+    tag_id: u16,
+    /// HWP 레코드 헤더 bits[10-19]: 트리 깊이
+    level: u16,
+    data: Vec<u8>,
+}
+
+// ─── 공개 API ────────────────────────────────────────────────────────────────
+
+/// 디버그용: Section0의 레코드 목록을 (tag_id, level, size)로 반환한다.
+pub fn debug_records(path: &str) -> Result<Vec<(u16, u16, usize)>> {
+    let file = std::fs::File::open(path)?;
+    let mut cfb = CompoundFile::open(file)
+        .map_err(|e| DocplerError::Cfb(e.to_string()))?;
+    let compressed = read_file_header(&mut cfb)?;
+    let data = read_stream(&mut cfb, "/BodyText/Section0", compressed)?;
+    let records = parse_records(&data);
+    Ok(records
+        .iter()
+        .map(|r| (r.tag_id, r.level, r.data.len()))
+        .collect())
+}
+
+pub fn parse(path: &str) -> Result<Document> {
+    let file = std::fs::File::open(path)?;
+    let mut cfb = CompoundFile::open(file)
+        .map_err(|e| DocplerError::Cfb(e.to_string()))?;
+
+    let compressed = read_file_header(&mut cfb)?;
+
+    let mut blocks = Vec::new();
+    for section_idx in 0.. {
+        let stream_path = format!("/BodyText/Section{}", section_idx);
+        let data = match read_stream(&mut cfb, &stream_path, compressed) {
+            Ok(d) => d,
+            Err(_) => break,
+        };
+        let records = parse_records(&data);
+        blocks.extend(process_records(&records));
+    }
+
+    Ok(Document { blocks })
+}
+
+// ─── FileHeader 파싱 ─────────────────────────────────────────────────────────
+
+fn read_file_header(cfb: &mut CompoundFile<std::fs::File>) -> Result<bool> {
+    let mut data = Vec::new();
+    cfb.open_stream("/FileHeader")
+        .map_err(|e| DocplerError::Cfb(e.to_string()))?
+        .read_to_end(&mut data)?;
+
+    if !data.starts_with(HWP_SIGNATURE) {
+        return Err(DocplerError::InvalidSignature);
+    }
+    if data.len() < 40 {
+        return Err(DocplerError::Parse("FileHeader too short".into()));
+    }
+
+    // offset 36: 속성 플래그 DWORD  bit0=압축, bit1=암호화
+    let props = u32::from_le_bytes(data[36..40].try_into().unwrap());
+    let compressed = (props & 0x01) != 0;
+    let encrypted = (props & 0x02) != 0;
+
+    if encrypted {
+        return Err(DocplerError::EncryptedFile);
+    }
+
+    Ok(compressed)
+}
+
+// ─── 스트림 읽기 ─────────────────────────────────────────────────────────────
+
+fn read_stream(
+    cfb: &mut CompoundFile<std::fs::File>,
+    path: &str,
+    compressed: bool,
+) -> Result<Vec<u8>> {
+    let mut raw = Vec::new();
+    cfb.open_stream(path)
+        .map_err(|e| DocplerError::Cfb(e.to_string()))?
+        .read_to_end(&mut raw)?;
+
+    if compressed {
+        // HWP는 raw deflate 압축 (zlib 헤더 없음, Java Inflater nowrap=true 동일)
+        let mut decoder = DeflateDecoder::new(raw.as_slice());
+        let mut out = Vec::new();
+        decoder
+            .read_to_end(&mut out)
+            .map_err(|e| DocplerError::Parse(format!("압축 해제 실패: {}", e)))?;
+        Ok(out)
+    } else {
+        Ok(raw)
+    }
+}
+
+// ─── 레코드 파싱 ─────────────────────────────────────────────────────────────
+
+/// 섹션 스트림을 Record 목록으로 변환.
+///
+/// 레코드 헤더 형식 (4바이트 DWORD):
+///   bits  0- 9 : TagID  (10비트)
+///   bits 10-19 : Level  (10비트)
+///   bits 20-31 : Size   (12비트, 0xFFF이면 다음 DWORD가 실제 크기)
+fn parse_records(data: &[u8]) -> Vec<Record> {
+    let mut reader = ByteReader::new(data);
+    let mut records = Vec::new();
+
+    while reader.remaining() >= 4 {
+        let header = reader.read_u32();
+        let tag_id = (header & 0x3FF) as u16;
+        let level = ((header >> 10) & 0x3FF) as u16;
+        let mut size = (header >> 20) as usize;
+
+        if size == 0xFFF {
+            if reader.remaining() < 4 {
+                break;
+            }
+            size = reader.read_u32() as usize;
+        }
+
+        if reader.remaining() < size {
+            break;
+        }
+
+        let record_data = reader.read_bytes(size).to_vec();
+        records.push(Record { tag_id, level, data: record_data });
+    }
+
+    records
+}
+
+// ─── 레코드 처리 ─────────────────────────────────────────────────────────────
+
+/// 레코드 목록을 Block 목록으로 변환.
+///
+/// HWP 5.0 레코드 레벨 구조 (섹션 기준):
+///   level 0 : PARA_HEADER, PARA_TEXT, PARA_CHAR_SHAPE 등 (최상위 문단)
+///   level 1 : CTRL_HEADER, CTRL_DATA, TABLE, LIST_HEADER(셀) 등
+///   level 2 : 셀 내부 문단 텍스트
+fn process_records(records: &[Record]) -> Vec<Block> {
+    let mut blocks = Vec::new();
+    let mut i = 0;
+
+    while i < records.len() {
+        let rec = &records[i];
+
+        match rec.tag_id {
+            TAG_PARA_TEXT if rec.level == 1 => {
+                let text = extract_para_text(&rec.data);
+                if !text.trim().is_empty() {
+                    blocks.push(Block::Paragraph(text));
+                }
+                i += 1;
+            }
+            TAG_CTRL_HEADER => {
+                let ctrl_level = rec.level;
+                let ctrl_id = first_u32(&rec.data);
+                i += 1;
+
+                if ctrl_id == CTRL_TABLE {
+                    let (table_opt, consumed) = parse_table_block(records, i, ctrl_level);
+                    if let Some(table) = table_opt {
+                        blocks.push(Block::Table(table));
+                    }
+                    i += consumed;
+                } else if ctrl_id == CTRL_EQUATION {
+                    let (eq_opt, consumed) = parse_equation_block(records, i, ctrl_level);
+                    if let Some(script) = eq_opt {
+                        blocks.push(Block::Equation(script));
+                    }
+                    i += consumed;
+                } else if ctrl_id == CTRL_GSO {
+                    let (text_opt, consumed) = parse_gso_block(records, i, ctrl_level);
+                    if let Some(text) = text_opt {
+                        blocks.push(Block::Paragraph(text));
+                    }
+                    i += consumed;
+                } else {
+                    // 이 컨트롤에 속한 레코드(level > ctrl_level) 모두 건너뜀
+                    while i < records.len() && records[i].level > ctrl_level {
+                        i += 1;
+                    }
+                }
+            }
+            _ => {
+                i += 1;
+            }
+        }
+    }
+
+    blocks
+}
+
+// ─── 표 파싱 ─────────────────────────────────────────────────────────────────
+
+/// CTRL_HEADER(표) 이후 레코드들에서 표 데이터를 추출.
+///
+/// 실제 레벨 구조 (ctrl_level=1 기준):
+///   level 2 (=ctrl_level+1): TABLE, LIST_HEADER(셀), PARA_HEADER(셀 문단)
+///   level 3 (=ctrl_level+2): 셀 내부 PARA_TEXT
+///
+/// TABLE 레코드 데이터 레이아웃:
+///   props(4) + rowCount(2) + colCount(2) + cellSpacing(2) + margins(4×2=8)
+///   + cellCountOfRow[rowCount](2 each)   → cellCountOfRow 시작 offset = 18
+fn parse_table_block(
+    records: &[Record],
+    start: usize,
+    ctrl_level: u16,
+) -> (Option<Vec<Vec<String>>>, usize) {
+    // TABLE과 셀 LIST_HEADER는 ctrl_level+1에 위치
+    let table_level = ctrl_level + 1;
+    let mut i = start;
+
+    // TABLE 레코드를 찾을 때까지 CTRL_DATA·캡션 LIST_HEADER를 건너뜀
+    while i < records.len() {
+        let rec = &records[i];
+        if rec.level < table_level {
+            return (None, i - start);
+        }
+        if rec.tag_id == TAG_TABLE && rec.level == table_level {
+            break;
+        }
+        if rec.tag_id == TAG_LIST_HEADER && rec.level == table_level {
+            // 캡션 LIST_HEADER → 자식(level > table_level)까지 건너뜀
+            i += 1;
+            while i < records.len() && records[i].level > table_level {
+                i += 1;
+            }
+        } else {
+            i += 1;
+        }
+    }
+
+    if i >= records.len() || records[i].tag_id != TAG_TABLE {
+        return (None, i - start);
+    }
+
+    // TABLE 레코드 데이터 파싱
+    let tdata = &records[i].data;
+    if tdata.len() < 18 {
+        return (None, i - start + 1);
+    }
+
+    let row_count = u16::from_le_bytes([tdata[4], tdata[5]]) as usize;
+    let mut cell_counts: Vec<usize> = Vec::with_capacity(row_count);
+    for j in 0..row_count {
+        let off = 18 + j * 2;
+        if off + 2 <= tdata.len() {
+            cell_counts.push(u16::from_le_bytes([tdata[off], tdata[off + 1]]) as usize);
+        } else {
+            cell_counts.push(0);
+        }
+    }
+
+    i += 1; // TABLE 소비
+
+    // 셀 파싱:
+    //   각 셀 = LIST_HEADER(table_level) 다음에 오는 레코드들
+    //   다음 LIST_HEADER(table_level) 또는 level < table_level이 나올 때까지가 이 셀의 내용
+    //   (PARA_HEADER도 table_level이므로 TAG로만 구분)
+    let mut rows: Vec<Vec<String>> = Vec::new();
+
+    for row_idx in 0..row_count {
+        let col_count = cell_counts.get(row_idx).copied().unwrap_or(0);
+        let mut row: Vec<String> = Vec::with_capacity(col_count);
+
+        for _ in 0..col_count {
+            if i < records.len()
+                && records[i].tag_id == TAG_LIST_HEADER
+                && records[i].level == table_level
+            {
+                i += 1; // LIST_HEADER 소비
+
+                let mut cell_text = String::new();
+                // 다음 LIST_HEADER(table_level) 또는 table_level 미만까지가 이 셀
+                while i < records.len() {
+                    let rec = &records[i];
+                    if rec.level < table_level {
+                        break;
+                    }
+                    if rec.tag_id == TAG_LIST_HEADER && rec.level == table_level {
+                        break; // 다음 셀 시작
+                    }
+                    if rec.tag_id == TAG_PARA_TEXT {
+                        let t = extract_para_text(&rec.data);
+                        let t = t.trim().to_string();
+                        if !t.is_empty() {
+                            if !cell_text.is_empty() {
+                                cell_text.push(' ');
+                            }
+                            cell_text.push_str(&t);
+                        }
+                    }
+                    i += 1;
+                }
+                row.push(cell_text);
+            } else {
+                row.push(String::new());
+            }
+        }
+
+        rows.push(row);
+    }
+
+    // 이 표 컨트롤의 나머지 레코드 건너뜀
+    while i < records.len() && records[i].level >= table_level {
+        i += 1;
+    }
+
+    if rows.is_empty() {
+        (None, i - start)
+    } else {
+        (Some(rows), i - start)
+    }
+}
+
+// ─── 수식 파싱 ───────────────────────────────────────────────────────────────
+
+/// CTRL_HEADER(수식) 이후 레코드들에서 HWP EQN 스크립트를 추출.
+///
+/// 구조 (ctrl_level=1 기준):
+///   [LIST_HEADER: level 2, 캡션 (선택)]
+///   EQEDIT: level 2  → 수식 스크립트 포함
+///
+/// EQEDIT 데이터 레이아웃:
+///   offset 0 (4바이트): property
+///   offset 4 (2바이트): script wchar 길이
+///   offset 6 (length×2 바이트): HWP EQN 스크립트 (UTF-16LE)
+fn parse_equation_block(
+    records: &[Record],
+    start: usize,
+    ctrl_level: u16,
+) -> (Option<String>, usize) {
+    let child_level = ctrl_level + 1;
+    let mut i = start;
+
+    // EQEDIT 레코드 탐색 (캡션 LIST_HEADER 건너뜀)
+    while i < records.len() {
+        let rec = &records[i];
+        if rec.level < child_level {
+            return (None, i - start);
+        }
+        if rec.tag_id == TAG_EQEDIT && rec.level == child_level {
+            break;
+        }
+        i += 1;
+    }
+
+    if i >= records.len() || records[i].tag_id != TAG_EQEDIT {
+        return (None, i - start);
+    }
+
+    let script = extract_eqedit_script(&records[i].data);
+    i += 1;
+
+    // 나머지 컨트롤 레코드 건너뜀
+    while i < records.len() && records[i].level > ctrl_level {
+        i += 1;
+    }
+
+    (script, i - start)
+}
+
+/// EQEDIT 레코드 데이터에서 HWP EQN 스크립트를 추출한다.
+fn extract_eqedit_script(data: &[u8]) -> Option<String> {
+    if data.len() < 6 {
+        return None;
+    }
+    // property(4) 건너뜀
+    let len_wchars = u16::from_le_bytes([data[4], data[5]]) as usize;
+    if len_wchars == 0 {
+        return None;
+    }
+    let script_start = 6;
+    let script_end = script_start + len_wchars * 2;
+    if script_end > data.len() {
+        return None;
+    }
+    // UTF-16LE → String
+    let utf16: Vec<u16> = data[script_start..script_end]
+        .chunks_exact(2)
+        .map(|c| u16::from_le_bytes([c[0], c[1]]))
+        .collect();
+    String::from_utf16(&utf16).ok()
+}
+
+// ─── 글상자(GSO) 파싱 ───────────────────────────────────────────────────────
+
+/// CTRL_HEADER(gso) 이후 레코드들에서 글상자 텍스트를 추출.
+///
+/// 구조 (ctrl_level=1 기준):
+///   [LIST_HEADER: level 2, 캡션 (선택)]
+///   [CTRL_DATA:   level 2 (선택)]
+///   SHAPE_COMPONENT: level 2   → 첫 4바이트 = gsoId
+///   [CTRL_DATA:   level 2 (선택)]
+///   LIST_HEADER:  level 2      → 글상자 본문 (텍스트 박스)
+///     PARA_HEADER: level 2
+///     PARA_TEXT:   level 3
+///   SHAPE_COMPONENT_RECTANGLE: level 2
+///
+/// 전략: SHAPE_COMPONENT를 찾은 뒤, 이후 첫 LIST_HEADER의 하위 PARA_TEXT를 수집.
+fn parse_gso_block(
+    records: &[Record],
+    start: usize,
+    ctrl_level: u16,
+) -> (Option<String>, usize) {
+    let child_level = ctrl_level + 1;
+    let mut i = start;
+
+    // SHAPE_COMPONENT 찾기 (캡션 LIST_HEADER, CTRL_DATA 건너뜀)
+    while i < records.len() {
+        let rec = &records[i];
+        if rec.level <= ctrl_level {
+            return (None, i - start);
+        }
+        if rec.tag_id == TAG_SHAPE_COMPONENT && rec.level == child_level {
+            break;
+        }
+        i += 1;
+    }
+
+    if i >= records.len() {
+        return (None, i - start);
+    }
+    i += 1; // SHAPE_COMPONENT 소비
+
+    // SHAPE_COMPONENT 이후 첫 LIST_HEADER(글상자 본문) 찾기
+    let mut found_list = false;
+    while i < records.len() {
+        let rec = &records[i];
+        if rec.level <= ctrl_level {
+            break;
+        }
+        if rec.tag_id == TAG_LIST_HEADER && rec.level == child_level {
+            found_list = true;
+            i += 1; // LIST_HEADER 소비
+            break;
+        }
+        i += 1;
+    }
+
+    if !found_list {
+        // GSO 블록 나머지 건너뜀
+        while i < records.len() && records[i].level > ctrl_level {
+            i += 1;
+        }
+        return (None, i - start);
+    }
+
+    // LIST_HEADER 이후 PARA_TEXT 수집.
+    // PARA_HEADER는 child_level(=2)에, PARA_TEXT는 child_level+1(=3)에 위치.
+    // SHAPE_COMPONENT_RECTANGLE(child_level)이 나오면 본문 종료.
+    let mut texts: Vec<String> = Vec::new();
+    while i < records.len() && records[i].level > ctrl_level {
+        let rec = &records[i];
+        if rec.tag_id == TAG_PARA_TEXT {
+            let t = extract_para_text(&rec.data);
+            let t = t.trim().to_string();
+            if !t.is_empty() {
+                texts.push(t);
+            }
+        }
+        i += 1;
+    }
+
+    // GSO 블록 나머지 건너뜀 (혹시 남은 경우)
+    while i < records.len() && records[i].level > ctrl_level {
+        i += 1;
+    }
+
+    if texts.is_empty() {
+        (None, i - start)
+    } else {
+        (Some(texts.join("\n")), i - start)
+    }
+}
+
+// ─── 텍스트 추출 ─────────────────────────────────────────────────────────────
+
+/// 레코드 데이터 첫 4바이트를 LE u32로 읽는다.
+fn first_u32(data: &[u8]) -> u32 {
+    if data.len() >= 4 {
+        u32::from_le_bytes(data[..4].try_into().unwrap())
+    } else {
+        0
+    }
+}
+
+/// PARA_TEXT 레코드 데이터에서 가시 텍스트를 추출.
+///
+/// 각 문자는 2바이트(UTF-16LE). 제어 코드 처리:
+///   code > 31              → 일반 문자 (Normal)
+///   1,2,3,11~18,21~23     → ControlExtend: 총 16바이트 (2+12+2)
+///   4~9,19,20             → ControlInline: 총 16바이트 (2+12+2)
+///   0,10,13,24~31         → ControlChar:   2바이트만
+pub fn extract_para_text(data: &[u8]) -> String {
+    let mut text = String::new();
+    let mut i = 0;
+
+    while i + 2 <= data.len() {
+        let code = u16::from_le_bytes([data[i], data[i + 1]]) as u32;
+
+        if code > 31 {
+            // 일반 문자 (UTF-16LE)
+            if (0xD800..0xDC00).contains(&code) {
+                // 상위 서로게이트 → 하위 서로게이트와 쌍
+                if i + 4 <= data.len() {
+                    let low = u16::from_le_bytes([data[i + 2], data[i + 3]]) as u32;
+                    if (0xDC00..0xE000).contains(&low) {
+                        let cp = 0x10000 + ((code - 0xD800) << 10) + (low - 0xDC00);
+                        if let Some(ch) = char::from_u32(cp) {
+                            text.push(ch);
+                        }
+                        i += 4;
+                        continue;
+                    }
+                }
+                i += 2;
+            } else if (0xDC00..0xE000).contains(&code) {
+                // 고립된 하위 서로게이트 → 건너뜀
+                i += 2;
+            } else if let Some(ch) = char::from_u32(code) {
+                text.push(ch);
+                i += 2;
+            } else {
+                i += 2;
+            }
+        } else {
+            // 제어 코드
+            match code {
+                1 | 2 | 3 | 11 | 12 | 14 | 15 | 16 | 17 | 18 | 21 | 22 | 23 => {
+                    // ControlExtend: 16바이트
+                    i += if i + 16 <= data.len() { 16 } else { break };
+                }
+                4 | 5 | 6 | 7 | 8 | 9 | 19 | 20 => {
+                    // ControlInline: 16바이트
+                    i += if i + 16 <= data.len() { 16 } else { break };
+                }
+                _ => {
+                    // ControlChar: 2바이트 (줄바꿈 등 무시)
+                    i += 2;
+                }
+            }
+        }
+    }
+
+    text
+}

docpler-0.1.5/crates/docpler-core/src/lib.rs

@@ -0,0 +1,4 @@
+pub mod error;
+pub mod hwp;
+pub mod markdown;
+pub mod reader;

docpler-0.1.5/crates/docpler-core/src/markdown.rs

@@ -0,0 +1,70 @@
+use crate::hwp::{Block, Document};
+
+pub fn render(doc: &Document) -> String {
+    let mut out = String::new();
+
+    for block in &doc.blocks {
+        match block {
+            Block::Paragraph(text) => {
+                let text = text.trim();
+                if !text.is_empty() {
+                    out.push_str(text);
+                    out.push_str("\n\n");
+                }
+            }
+            Block::Table(rows) => {
+                if !rows.is_empty() {
+                    render_table(&mut out, rows);
+                    out.push('\n');
+                }
+            }
+            Block::Equation(script) => {
+                // HWP EQN 포맷 스크립트를 코드 블록으로 출력.
+                // TODO: HWP EQN → LaTeX 변환
+                out.push_str("$$\n");
+                out.push_str(script.trim());
+                out.push_str("\n$$\n\n");
+            }
+        }
+    }
+
+    out.trim_end().to_string()
+}
+
+fn render_table(out: &mut String, rows: &[Vec<String>]) {
+    let max_cols = rows.iter().map(|r| r.len()).max().unwrap_or(0);
+    if max_cols == 0 {
+        return;
+    }
+
+    let cell = |row: &Vec<String>, col: usize| -> String {
+        row.get(col)
+            .map(|s| s.replace('|', "\\|").replace('\n', " "))
+            .unwrap_or_default()
+    };
+
+    // 헤더 행
+    out.push('|');
+    for col in 0..max_cols {
+        out.push_str(&cell(&rows[0], col));
+        out.push('|');
+    }
+    out.push('\n');
+
+    // 구분선
+    out.push('|');
+    for _ in 0..max_cols {
+        out.push_str("---|");
+    }
+    out.push('\n');
+
+    // 데이터 행
+    for row in &rows[1..] {
+        out.push('|');
+        for col in 0..max_cols {
+            out.push_str(&cell(row, col));
+            out.push('|');
+        }
+        out.push('\n');
+    }
+}

docpler-0.1.5/crates/docpler-core/src/reader.rs

@@ -0,0 +1,37 @@
+/// 바이트 슬라이스에서 리틀엔디안 정수를 읽는 헬퍼.
+/// HWP 레코드 스트림 파싱에 사용된다.
+pub struct ByteReader<'a> {
+    data: &'a [u8],
+    pos: usize,
+}
+
+impl<'a> ByteReader<'a> {
+    pub fn new(data: &'a [u8]) -> Self {
+        Self { data, pos: 0 }
+    }
+
+    pub fn remaining(&self) -> usize {
+        self.data.len().saturating_sub(self.pos)
+    }
+
+    pub fn read_u32(&mut self) -> u32 {
+        if self.pos + 4 <= self.data.len() {
+            let val = u32::from_le_bytes(
+                self.data[self.pos..self.pos + 4].try_into().unwrap(),
+            );
+            self.pos += 4;
+            val
+        } else {
+            self.pos = self.data.len();
+            0
+        }
+    }
+
+    /// n 바이트를 읽어 슬라이스로 반환. 데이터 부족 시 가능한 만큼만 반환.
+    pub fn read_bytes(&mut self, n: usize) -> &'a [u8] {
+        let start = self.pos;
+        let end = (self.pos + n).min(self.data.len());
+        self.pos = end;
+        &self.data[start..end]
+    }
+}

docpler-0.1.5/pyproject.toml

@@ -0,0 +1,25 @@
+[build-system]
+requires = ["maturin>=1.7,<2.0"]
+build-backend = "maturin"
+
+[project]
+name = "docpler"
+version = "0.1.5"
+description = "Document processing tool - converts HWP (and more) to Markdown"
+requires-python = ">=3.9"
+license = { text = "MIT" }
+readme = "PYPI_README.md"
+dependencies = []
+
+[project.optional-dependencies]
+markitdown = ["markitdown>=0.0.1"]
+
+# markitdown 플러그인으로 자동 등록
+[project.entry-points."markitdown.plugin"]
+docpler = "docpler.hwp:HwpConverterPlugin"
+
+[tool.maturin]
+manifest-path = "crates/docpler-bindings/Cargo.toml"
+python-source = "python"
+module-name = "docpler._docpler"
+features = ["pyo3/extension-module"]

docpler-0.1.5/python/docpler/__init__.py

@@ -0,0 +1,4 @@
+"""
+docpler - Document processing tool.
+HWP, and more formats to Markdown.
+"""

docpler-0.1.5/python/docpler/hwp.py

@@ -0,0 +1,57 @@
+"""
+HWP (한글 워드프로세서) → Markdown 변환기.
+markitdown 플러그인 및 독립 사용 모두 지원.
+"""
+from pathlib import Path
+from typing import Any, BinaryIO, Optional
+
+
+ACCEPTED_EXTENSIONS = frozenset({".hwp"})
+ACCEPTED_MIME_TYPES = frozenset({
+    "application/x-hwp",
+    "application/haansofthwp",
+    "application/vnd.hancom.hwp",
+})
+
+
+def convert(path: str | Path) -> str:
+    """HWP 파일을 Markdown 문자열로 변환한다."""
+    from ._docpler import hwp_to_markdown
+    return hwp_to_markdown(str(path))
+
+
+class HwpConverter:
+    """markitdown 컨버터: HWP 파일을 Markdown으로 변환."""
+
+    def accepts(
+        self,
+        file_stream: BinaryIO,
+        stream_info: Any,
+        **kwargs,
+    ) -> bool:
+        ext = getattr(stream_info, "extension", "") or ""
+        mime = getattr(stream_info, "mimetype", "") or ""
+        return ext.lower() in ACCEPTED_EXTENSIONS or mime in ACCEPTED_MIME_TYPES
+
+    def convert(
+        self,
+        file_stream: BinaryIO,
+        stream_info: Any,
+        **kwargs,
+    ):
+        from markitdown import DocumentConverterResult
+
+        local_path = getattr(stream_info, "local_path", None)
+        if not local_path:
+            return None
+
+        markdown_text = convert(local_path)
+        return DocumentConverterResult(markdown=markdown_text)
+
+
+class HwpConverterPlugin:
+    """markitdown 플러그인 엔트리포인트."""
+
+    @staticmethod
+    def register_converters(markitdown_instance, **kwargs):
+        markitdown_instance.register_converter(HwpConverter())