pdf-rs 0.1.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- pdf_rs-0.1.0/.gitattributes +3 -0
- pdf_rs-0.1.0/.github/workflows/ci.yml +131 -0
- pdf_rs-0.1.0/.github/workflows/release.yml +102 -0
- pdf_rs-0.1.0/.gitignore +9 -0
- pdf_rs-0.1.0/CONTRIBUTING.md +46 -0
- pdf_rs-0.1.0/Cargo.lock +291 -0
- pdf_rs-0.1.0/Cargo.toml +29 -0
- pdf_rs-0.1.0/PKG-INFO +100 -0
- pdf_rs-0.1.0/README.md +88 -0
- pdf_rs-0.1.0/dist/pdf_rs-0.1.0-cp39-abi3-manylinux_2_34_x86_64.whl +0 -0
- pdf_rs-0.1.0/pyproject.toml +21 -0
- pdf_rs-0.1.0/src/annotation.rs +43 -0
- pdf_rs-0.1.0/src/cmap.rs +307 -0
- pdf_rs-0.1.0/src/content.rs +258 -0
- pdf_rs-0.1.0/src/document.rs +1719 -0
- pdf_rs-0.1.0/src/embedded_file.rs +58 -0
- pdf_rs-0.1.0/src/error.rs +94 -0
- pdf_rs-0.1.0/src/form.rs +74 -0
- pdf_rs-0.1.0/src/inspect.rs +111 -0
- pdf_rs-0.1.0/src/lexer.rs +306 -0
- pdf_rs-0.1.0/src/lib.rs +53 -0
- pdf_rs-0.1.0/src/link.rs +22 -0
- pdf_rs-0.1.0/src/llm.rs +398 -0
- pdf_rs-0.1.0/src/main.rs +109 -0
- pdf_rs-0.1.0/src/markdown.rs +167 -0
- pdf_rs-0.1.0/src/metadata.rs +107 -0
- pdf_rs-0.1.0/src/object.rs +473 -0
- pdf_rs-0.1.0/src/outline.rs +10 -0
- pdf_rs-0.1.0/src/page.rs +106 -0
- pdf_rs-0.1.0/src/page_label.rs +136 -0
- pdf_rs-0.1.0/src/parser.rs +240 -0
- pdf_rs-0.1.0/src/python.rs +956 -0
- pdf_rs-0.1.0/src/resource.rs +331 -0
- pdf_rs-0.1.0/src/writer.rs +525 -0
- pdf_rs-0.1.0/tests/basic_parse.rs +2310 -0
- pdf_rs-0.1.0/tests/gen/complex.typ +35 -0
- pdf_rs-0.1.0/tests/gen/references.typ +14 -0
- pdf_rs-0.1.0/tests/gen/test1.typ +2 -0
- pdf_rs-0.1.0/tests/python_api.py +125 -0
- pdf_rs-0.1.0/tests/real_world_smoke.rs +41 -0
- pdf_rs-0.1.0/tools/compare_pdf_perf.py +221 -0
- pdf_rs-0.1.0/tools/profile_pdf_rs.py +103 -0
|
@@ -0,0 +1,131 @@
|
|
|
1
|
+
name: CI
|
|
2
|
+
|
|
3
|
+
on:
|
|
4
|
+
push:
|
|
5
|
+
branches: [main]
|
|
6
|
+
pull_request:
|
|
7
|
+
workflow_dispatch:
|
|
8
|
+
|
|
9
|
+
permissions:
|
|
10
|
+
contents: read
|
|
11
|
+
|
|
12
|
+
concurrency:
|
|
13
|
+
group: ci-${{ github.workflow }}-${{ github.ref }}
|
|
14
|
+
cancel-in-progress: true
|
|
15
|
+
|
|
16
|
+
env:
|
|
17
|
+
CARGO_TERM_COLOR: always
|
|
18
|
+
|
|
19
|
+
jobs:
|
|
20
|
+
rust:
|
|
21
|
+
name: Rust checks (${{ matrix.os }})
|
|
22
|
+
runs-on: ${{ matrix.os }}
|
|
23
|
+
strategy:
|
|
24
|
+
fail-fast: false
|
|
25
|
+
matrix:
|
|
26
|
+
os: [ubuntu-latest, windows-latest, macos-latest]
|
|
27
|
+
|
|
28
|
+
steps:
|
|
29
|
+
- name: Checkout
|
|
30
|
+
uses: actions/checkout@v6
|
|
31
|
+
|
|
32
|
+
- name: Set up Python
|
|
33
|
+
uses: actions/setup-python@v6
|
|
34
|
+
with:
|
|
35
|
+
python-version: "3.12"
|
|
36
|
+
|
|
37
|
+
- name: Install Rust components
|
|
38
|
+
run: rustup component add rustfmt clippy
|
|
39
|
+
|
|
40
|
+
- name: Cache cargo
|
|
41
|
+
uses: Swatinem/rust-cache@v2
|
|
42
|
+
with:
|
|
43
|
+
shared-key: cargo-${{ runner.os }}
|
|
44
|
+
|
|
45
|
+
- name: Format
|
|
46
|
+
run: cargo fmt --check
|
|
47
|
+
|
|
48
|
+
- name: Test
|
|
49
|
+
run: cargo test
|
|
50
|
+
|
|
51
|
+
- name: Clippy
|
|
52
|
+
run: cargo clippy --all-targets -- -D warnings
|
|
53
|
+
|
|
54
|
+
- name: Check mimalloc feature
|
|
55
|
+
run: cargo check --features mimalloc
|
|
56
|
+
|
|
57
|
+
- name: Clippy Python feature
|
|
58
|
+
run: cargo clippy --features python --all-targets -- -D warnings
|
|
59
|
+
|
|
60
|
+
typst-fixtures:
|
|
61
|
+
name: Typst fixture tests
|
|
62
|
+
runs-on: ubuntu-latest
|
|
63
|
+
|
|
64
|
+
steps:
|
|
65
|
+
- name: Checkout
|
|
66
|
+
uses: actions/checkout@v6
|
|
67
|
+
|
|
68
|
+
- name: Set up Python
|
|
69
|
+
uses: actions/setup-python@v6
|
|
70
|
+
with:
|
|
71
|
+
python-version: "3.12"
|
|
72
|
+
|
|
73
|
+
- name: Install Rust components
|
|
74
|
+
run: rustup component add rustfmt clippy
|
|
75
|
+
|
|
76
|
+
- name: Cache cargo
|
|
77
|
+
uses: Swatinem/rust-cache@v2
|
|
78
|
+
with:
|
|
79
|
+
shared-key: typst-fixtures
|
|
80
|
+
|
|
81
|
+
- name: Install Typst
|
|
82
|
+
run: cargo install typst-cli --locked
|
|
83
|
+
|
|
84
|
+
- name: Run Typst-backed tests
|
|
85
|
+
env:
|
|
86
|
+
PDF_RS_REQUIRE_TYPST: "1"
|
|
87
|
+
run: cargo test typst -- --nocapture
|
|
88
|
+
|
|
89
|
+
python-wheel:
|
|
90
|
+
name: Python wheel smoke (${{ matrix.os }})
|
|
91
|
+
runs-on: ${{ matrix.os }}
|
|
92
|
+
strategy:
|
|
93
|
+
fail-fast: false
|
|
94
|
+
matrix:
|
|
95
|
+
os: [ubuntu-latest, windows-latest, macos-latest]
|
|
96
|
+
|
|
97
|
+
steps:
|
|
98
|
+
- name: Checkout
|
|
99
|
+
uses: actions/checkout@v6
|
|
100
|
+
|
|
101
|
+
- name: Set up Python
|
|
102
|
+
uses: actions/setup-python@v6
|
|
103
|
+
with:
|
|
104
|
+
python-version: "3.12"
|
|
105
|
+
|
|
106
|
+
- name: Cache cargo
|
|
107
|
+
uses: Swatinem/rust-cache@v2
|
|
108
|
+
with:
|
|
109
|
+
shared-key: wheel-${{ runner.os }}
|
|
110
|
+
|
|
111
|
+
- name: Install maturin
|
|
112
|
+
run: python -m pip install --upgrade pip maturin
|
|
113
|
+
|
|
114
|
+
- name: Build wheel
|
|
115
|
+
run: maturin build --features python --out dist
|
|
116
|
+
|
|
117
|
+
- name: Install wheel
|
|
118
|
+
shell: bash
|
|
119
|
+
run: |
|
|
120
|
+
python -m pip install --force-reinstall dist/*.whl
|
|
121
|
+
|
|
122
|
+
- name: Python smoke
|
|
123
|
+
run: python tests/python_api.py
|
|
124
|
+
|
|
125
|
+
- name: Upload wheel artifact
|
|
126
|
+
uses: actions/upload-artifact@v6
|
|
127
|
+
with:
|
|
128
|
+
name: wheel-${{ matrix.os }}
|
|
129
|
+
path: dist/*
|
|
130
|
+
if-no-files-found: error
|
|
131
|
+
retention-days: 7
|
|
@@ -0,0 +1,102 @@
|
|
|
1
|
+
name: Release
|
|
2
|
+
|
|
3
|
+
on:
|
|
4
|
+
push:
|
|
5
|
+
tags:
|
|
6
|
+
- "v*.*.*"
|
|
7
|
+
|
|
8
|
+
permissions:
|
|
9
|
+
contents: write
|
|
10
|
+
id-token: write
|
|
11
|
+
|
|
12
|
+
concurrency:
|
|
13
|
+
group: release-${{ github.workflow }}-${{ github.ref }}
|
|
14
|
+
cancel-in-progress: false
|
|
15
|
+
|
|
16
|
+
env:
|
|
17
|
+
CARGO_TERM_COLOR: always
|
|
18
|
+
|
|
19
|
+
jobs:
|
|
20
|
+
build:
|
|
21
|
+
name: Build distributions (${{ matrix.os }})
|
|
22
|
+
runs-on: ${{ matrix.os }}
|
|
23
|
+
strategy:
|
|
24
|
+
fail-fast: false
|
|
25
|
+
matrix:
|
|
26
|
+
include:
|
|
27
|
+
- os: ubuntu-latest
|
|
28
|
+
artifact: linux
|
|
29
|
+
build_sdist: true
|
|
30
|
+
- os: windows-latest
|
|
31
|
+
artifact: windows
|
|
32
|
+
build_sdist: false
|
|
33
|
+
- os: macos-latest
|
|
34
|
+
artifact: macos
|
|
35
|
+
build_sdist: false
|
|
36
|
+
|
|
37
|
+
steps:
|
|
38
|
+
- name: Checkout
|
|
39
|
+
uses: actions/checkout@v6
|
|
40
|
+
|
|
41
|
+
- name: Set up Python
|
|
42
|
+
uses: actions/setup-python@v6
|
|
43
|
+
with:
|
|
44
|
+
python-version: "3.12"
|
|
45
|
+
|
|
46
|
+
- name: Cache cargo
|
|
47
|
+
uses: Swatinem/rust-cache@v2
|
|
48
|
+
with:
|
|
49
|
+
shared-key: release-${{ runner.os }}
|
|
50
|
+
|
|
51
|
+
- name: Install maturin
|
|
52
|
+
run: python -m pip install --upgrade pip maturin
|
|
53
|
+
|
|
54
|
+
- name: Build wheel
|
|
55
|
+
run: maturin build --release --features python --out dist
|
|
56
|
+
|
|
57
|
+
- name: Build sdist
|
|
58
|
+
if: matrix.build_sdist
|
|
59
|
+
run: maturin sdist --out dist
|
|
60
|
+
|
|
61
|
+
- name: Upload distributions
|
|
62
|
+
uses: actions/upload-artifact@v6
|
|
63
|
+
with:
|
|
64
|
+
name: dist-${{ matrix.artifact }}
|
|
65
|
+
path: dist/*
|
|
66
|
+
if-no-files-found: error
|
|
67
|
+
retention-days: 14
|
|
68
|
+
|
|
69
|
+
publish:
|
|
70
|
+
name: Publish PyPI and GitHub release
|
|
71
|
+
needs: build
|
|
72
|
+
runs-on: ubuntu-latest
|
|
73
|
+
environment: pypi
|
|
74
|
+
|
|
75
|
+
steps:
|
|
76
|
+
- name: Checkout
|
|
77
|
+
uses: actions/checkout@v6
|
|
78
|
+
|
|
79
|
+
- name: Download distributions
|
|
80
|
+
uses: actions/download-artifact@v7
|
|
81
|
+
with:
|
|
82
|
+
pattern: dist-*
|
|
83
|
+
path: dist
|
|
84
|
+
merge-multiple: true
|
|
85
|
+
|
|
86
|
+
- name: List distributions
|
|
87
|
+
run: ls -lah dist
|
|
88
|
+
|
|
89
|
+
- name: Publish to PyPI
|
|
90
|
+
uses: pypa/gh-action-pypi-publish@release/v1
|
|
91
|
+
with:
|
|
92
|
+
packages-dir: dist
|
|
93
|
+
password: ${{ secrets.PYPI_API_TOKEN }}
|
|
94
|
+
skip-existing: true
|
|
95
|
+
|
|
96
|
+
- name: Create GitHub release
|
|
97
|
+
env:
|
|
98
|
+
GH_TOKEN: ${{ github.token }}
|
|
99
|
+
run: |
|
|
100
|
+
gh release create "${GITHUB_REF_NAME}" dist/* \
|
|
101
|
+
--title "${GITHUB_REF_NAME}" \
|
|
102
|
+
--generate-notes
|
pdf_rs-0.1.0/.gitignore
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
1
|
+
# Contributing
|
|
2
|
+
|
|
3
|
+
## Development Loop
|
|
4
|
+
|
|
5
|
+
1. Keep changes focused on the parser, Python API, tools, tests, or release
|
|
6
|
+
automation being changed.
|
|
7
|
+
2. Add tests under `tests/` for parser behavior and corner cases.
|
|
8
|
+
3. Run formatting, tests, and clippy before opening a pull request.
|
|
9
|
+
|
|
10
|
+
```powershell
|
|
11
|
+
cargo fmt --check
|
|
12
|
+
cargo test
|
|
13
|
+
cargo clippy --all-targets -- -D warnings
|
|
14
|
+
```
|
|
15
|
+
|
|
16
|
+
When touching Python bindings, also run:
|
|
17
|
+
|
|
18
|
+
```powershell
|
|
19
|
+
cargo clippy --features python --all-targets -- -D warnings
|
|
20
|
+
uvx maturin build --features python --out target\wheels
|
|
21
|
+
```
|
|
22
|
+
|
|
23
|
+
## Commit Messages
|
|
24
|
+
|
|
25
|
+
Use:
|
|
26
|
+
|
|
27
|
+
```text
|
|
28
|
+
action(scope): detailed changes
|
|
29
|
+
```
|
|
30
|
+
|
|
31
|
+
Examples:
|
|
32
|
+
|
|
33
|
+
```text
|
|
34
|
+
feat(parser): add batched object stream expansion
|
|
35
|
+
ci(release): build abi3 wheels for tagged releases
|
|
36
|
+
test(python): cover LLM chunk and OCR request APIs
|
|
37
|
+
```
|
|
38
|
+
|
|
39
|
+
## Release Checklist
|
|
40
|
+
|
|
41
|
+
1. Update versions in `Cargo.toml` and `pyproject.toml`.
|
|
42
|
+
2. Run the full local check set from `README.md`.
|
|
43
|
+
3. Push changes to `main`.
|
|
44
|
+
4. Create and push a semver tag such as `v0.1.0`.
|
|
45
|
+
5. Confirm the GitHub release workflow builds all artifacts, publishes to PyPI,
|
|
46
|
+
and attaches the distributions to the GitHub release.
|
pdf_rs-0.1.0/Cargo.lock
ADDED
|
@@ -0,0 +1,291 @@
|
|
|
1
|
+
# This file is automatically @generated by Cargo.
|
|
2
|
+
# It is not intended for manual editing.
|
|
3
|
+
version = 4
|
|
4
|
+
|
|
5
|
+
[[package]]
|
|
6
|
+
name = "adler2"
|
|
7
|
+
version = "2.0.1"
|
|
8
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
9
|
+
checksum = "320119579fcad9c21884f5c4861d16174d0e06250625266f50fe6898340abefa"
|
|
10
|
+
|
|
11
|
+
[[package]]
|
|
12
|
+
name = "cc"
|
|
13
|
+
version = "1.2.62"
|
|
14
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
15
|
+
checksum = "a1dce859f0832a7d088c4f1119888ab94ef4b5d6795d1ce05afb7fe159d79f98"
|
|
16
|
+
dependencies = [
|
|
17
|
+
"find-msvc-tools",
|
|
18
|
+
"shlex",
|
|
19
|
+
]
|
|
20
|
+
|
|
21
|
+
[[package]]
|
|
22
|
+
name = "cfg-if"
|
|
23
|
+
version = "1.0.4"
|
|
24
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
25
|
+
checksum = "9330f8b2ff13f34540b44e946ef35111825727b38d33286ef986142615121801"
|
|
26
|
+
|
|
27
|
+
[[package]]
|
|
28
|
+
name = "crc32fast"
|
|
29
|
+
version = "1.5.0"
|
|
30
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
31
|
+
checksum = "9481c1c90cbf2ac953f07c8d4a58aa3945c425b7185c9154d67a65e4230da511"
|
|
32
|
+
dependencies = [
|
|
33
|
+
"cfg-if",
|
|
34
|
+
]
|
|
35
|
+
|
|
36
|
+
[[package]]
|
|
37
|
+
name = "find-msvc-tools"
|
|
38
|
+
version = "0.1.9"
|
|
39
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
40
|
+
checksum = "5baebc0774151f905a1a2cc41989300b1e6fbb29aff0ceffa1064fdd3088d582"
|
|
41
|
+
|
|
42
|
+
[[package]]
|
|
43
|
+
name = "flate2"
|
|
44
|
+
version = "1.1.9"
|
|
45
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
46
|
+
checksum = "843fba2746e448b37e26a819579957415c8cef339bf08564fe8b7ddbd959573c"
|
|
47
|
+
dependencies = [
|
|
48
|
+
"crc32fast",
|
|
49
|
+
"miniz_oxide",
|
|
50
|
+
]
|
|
51
|
+
|
|
52
|
+
[[package]]
|
|
53
|
+
name = "heck"
|
|
54
|
+
version = "0.5.0"
|
|
55
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
56
|
+
checksum = "2304e00983f87ffb38b55b444b5e3b60a884b5d30c0fca7d82fe33449bbe55ea"
|
|
57
|
+
|
|
58
|
+
[[package]]
|
|
59
|
+
name = "libc"
|
|
60
|
+
version = "0.2.186"
|
|
61
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
62
|
+
checksum = "68ab91017fe16c622486840e4c83c9a37afeff978bd239b5293d61ece587de66"
|
|
63
|
+
|
|
64
|
+
[[package]]
|
|
65
|
+
name = "libmimalloc-sys"
|
|
66
|
+
version = "0.1.47"
|
|
67
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
68
|
+
checksum = "2d1eacfa31c33ec25e873c136ba5669f00f9866d0688bea7be4d3f7e43067df6"
|
|
69
|
+
dependencies = [
|
|
70
|
+
"cc",
|
|
71
|
+
]
|
|
72
|
+
|
|
73
|
+
[[package]]
|
|
74
|
+
name = "memmap2"
|
|
75
|
+
version = "0.9.10"
|
|
76
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
77
|
+
checksum = "714098028fe011992e1c3962653c96b2d578c4b4bce9036e15ff220319b1e0e3"
|
|
78
|
+
dependencies = [
|
|
79
|
+
"libc",
|
|
80
|
+
]
|
|
81
|
+
|
|
82
|
+
[[package]]
|
|
83
|
+
name = "mimalloc"
|
|
84
|
+
version = "0.1.50"
|
|
85
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
86
|
+
checksum = "b3627c4272df786b9260cabaa46aec1d59c93ede723d4c3ef646c503816b0640"
|
|
87
|
+
dependencies = [
|
|
88
|
+
"libmimalloc-sys",
|
|
89
|
+
]
|
|
90
|
+
|
|
91
|
+
[[package]]
|
|
92
|
+
name = "miniz_oxide"
|
|
93
|
+
version = "0.8.9"
|
|
94
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
95
|
+
checksum = "1fa76a2c86f704bdb222d66965fb3d63269ce38518b83cb0575fca855ebb6316"
|
|
96
|
+
dependencies = [
|
|
97
|
+
"adler2",
|
|
98
|
+
"simd-adler32",
|
|
99
|
+
]
|
|
100
|
+
|
|
101
|
+
[[package]]
|
|
102
|
+
name = "once_cell"
|
|
103
|
+
version = "1.21.4"
|
|
104
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
105
|
+
checksum = "9f7c3e4beb33f85d45ae3e3a1792185706c8e16d043238c593331cc7cd313b50"
|
|
106
|
+
|
|
107
|
+
[[package]]
|
|
108
|
+
name = "pdf-rs"
|
|
109
|
+
version = "0.1.0"
|
|
110
|
+
dependencies = [
|
|
111
|
+
"flate2",
|
|
112
|
+
"memmap2",
|
|
113
|
+
"mimalloc",
|
|
114
|
+
"pyo3",
|
|
115
|
+
"thiserror",
|
|
116
|
+
"tracing",
|
|
117
|
+
]
|
|
118
|
+
|
|
119
|
+
[[package]]
|
|
120
|
+
name = "pin-project-lite"
|
|
121
|
+
version = "0.2.17"
|
|
122
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
123
|
+
checksum = "a89322df9ebe1c1578d689c92318e070967d1042b512afbe49518723f4e6d5cd"
|
|
124
|
+
|
|
125
|
+
[[package]]
|
|
126
|
+
name = "portable-atomic"
|
|
127
|
+
version = "1.13.1"
|
|
128
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
129
|
+
checksum = "c33a9471896f1c69cecef8d20cbe2f7accd12527ce60845ff44c153bb2a21b49"
|
|
130
|
+
|
|
131
|
+
[[package]]
|
|
132
|
+
name = "proc-macro2"
|
|
133
|
+
version = "1.0.106"
|
|
134
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
135
|
+
checksum = "8fd00f0bb2e90d81d1044c2b32617f68fcb9fa3bb7640c23e9c748e53fb30934"
|
|
136
|
+
dependencies = [
|
|
137
|
+
"unicode-ident",
|
|
138
|
+
]
|
|
139
|
+
|
|
140
|
+
[[package]]
|
|
141
|
+
name = "pyo3"
|
|
142
|
+
version = "0.28.3"
|
|
143
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
144
|
+
checksum = "91fd8e38a3b50ed1167fb981cd6fd60147e091784c427b8f7183a7ee32c31c12"
|
|
145
|
+
dependencies = [
|
|
146
|
+
"libc",
|
|
147
|
+
"once_cell",
|
|
148
|
+
"portable-atomic",
|
|
149
|
+
"pyo3-build-config",
|
|
150
|
+
"pyo3-ffi",
|
|
151
|
+
"pyo3-macros",
|
|
152
|
+
]
|
|
153
|
+
|
|
154
|
+
[[package]]
|
|
155
|
+
name = "pyo3-build-config"
|
|
156
|
+
version = "0.28.3"
|
|
157
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
158
|
+
checksum = "e368e7ddfdeb98c9bca7f8383be1648fd84ab466bf2bc015e94008db6d35611e"
|
|
159
|
+
dependencies = [
|
|
160
|
+
"target-lexicon",
|
|
161
|
+
]
|
|
162
|
+
|
|
163
|
+
[[package]]
|
|
164
|
+
name = "pyo3-ffi"
|
|
165
|
+
version = "0.28.3"
|
|
166
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
167
|
+
checksum = "7f29e10af80b1f7ccaf7f69eace800a03ecd13e883acfacc1e5d0988605f651e"
|
|
168
|
+
dependencies = [
|
|
169
|
+
"libc",
|
|
170
|
+
"pyo3-build-config",
|
|
171
|
+
]
|
|
172
|
+
|
|
173
|
+
[[package]]
|
|
174
|
+
name = "pyo3-macros"
|
|
175
|
+
version = "0.28.3"
|
|
176
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
177
|
+
checksum = "df6e520eff47c45997d2fc7dd8214b25dd1310918bbb2642156ef66a67f29813"
|
|
178
|
+
dependencies = [
|
|
179
|
+
"proc-macro2",
|
|
180
|
+
"pyo3-macros-backend",
|
|
181
|
+
"quote",
|
|
182
|
+
"syn",
|
|
183
|
+
]
|
|
184
|
+
|
|
185
|
+
[[package]]
|
|
186
|
+
name = "pyo3-macros-backend"
|
|
187
|
+
version = "0.28.3"
|
|
188
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
189
|
+
checksum = "c4cdc218d835738f81c2338f822078af45b4afdf8b2e33cbb5916f108b813acb"
|
|
190
|
+
dependencies = [
|
|
191
|
+
"heck",
|
|
192
|
+
"proc-macro2",
|
|
193
|
+
"pyo3-build-config",
|
|
194
|
+
"quote",
|
|
195
|
+
"syn",
|
|
196
|
+
]
|
|
197
|
+
|
|
198
|
+
[[package]]
|
|
199
|
+
name = "quote"
|
|
200
|
+
version = "1.0.45"
|
|
201
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
202
|
+
checksum = "41f2619966050689382d2b44f664f4bc593e129785a36d6ee376ddf37259b924"
|
|
203
|
+
dependencies = [
|
|
204
|
+
"proc-macro2",
|
|
205
|
+
]
|
|
206
|
+
|
|
207
|
+
[[package]]
|
|
208
|
+
name = "shlex"
|
|
209
|
+
version = "1.3.0"
|
|
210
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
211
|
+
checksum = "0fda2ff0d084019ba4d7c6f371c95d8fd75ce3524c3cb8fb653a3023f6323e64"
|
|
212
|
+
|
|
213
|
+
[[package]]
|
|
214
|
+
name = "simd-adler32"
|
|
215
|
+
version = "0.3.9"
|
|
216
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
217
|
+
checksum = "703d5c7ef118737c72f1af64ad2f6f8c5e1921f818cdcb97b8fe6fc69bf66214"
|
|
218
|
+
|
|
219
|
+
[[package]]
|
|
220
|
+
name = "syn"
|
|
221
|
+
version = "2.0.117"
|
|
222
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
223
|
+
checksum = "e665b8803e7b1d2a727f4023456bbbbe74da67099c585258af0ad9c5013b9b99"
|
|
224
|
+
dependencies = [
|
|
225
|
+
"proc-macro2",
|
|
226
|
+
"quote",
|
|
227
|
+
"unicode-ident",
|
|
228
|
+
]
|
|
229
|
+
|
|
230
|
+
[[package]]
|
|
231
|
+
name = "target-lexicon"
|
|
232
|
+
version = "0.13.5"
|
|
233
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
234
|
+
checksum = "adb6935a6f5c20170eeceb1a3835a49e12e19d792f6dd344ccc76a985ca5a6ca"
|
|
235
|
+
|
|
236
|
+
[[package]]
|
|
237
|
+
name = "thiserror"
|
|
238
|
+
version = "2.0.18"
|
|
239
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
240
|
+
checksum = "4288b5bcbc7920c07a1149a35cf9590a2aa808e0bc1eafaade0b80947865fbc4"
|
|
241
|
+
dependencies = [
|
|
242
|
+
"thiserror-impl",
|
|
243
|
+
]
|
|
244
|
+
|
|
245
|
+
[[package]]
|
|
246
|
+
name = "thiserror-impl"
|
|
247
|
+
version = "2.0.18"
|
|
248
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
249
|
+
checksum = "ebc4ee7f67670e9b64d05fa4253e753e016c6c95ff35b89b7941d6b856dec1d5"
|
|
250
|
+
dependencies = [
|
|
251
|
+
"proc-macro2",
|
|
252
|
+
"quote",
|
|
253
|
+
"syn",
|
|
254
|
+
]
|
|
255
|
+
|
|
256
|
+
[[package]]
|
|
257
|
+
name = "tracing"
|
|
258
|
+
version = "0.1.44"
|
|
259
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
260
|
+
checksum = "63e71662fa4b2a2c3a26f570f037eb95bb1f85397f3cd8076caed2f026a6d100"
|
|
261
|
+
dependencies = [
|
|
262
|
+
"pin-project-lite",
|
|
263
|
+
"tracing-attributes",
|
|
264
|
+
"tracing-core",
|
|
265
|
+
]
|
|
266
|
+
|
|
267
|
+
[[package]]
|
|
268
|
+
name = "tracing-attributes"
|
|
269
|
+
version = "0.1.31"
|
|
270
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
271
|
+
checksum = "7490cfa5ec963746568740651ac6781f701c9c5ea257c58e057f3ba8cf69e8da"
|
|
272
|
+
dependencies = [
|
|
273
|
+
"proc-macro2",
|
|
274
|
+
"quote",
|
|
275
|
+
"syn",
|
|
276
|
+
]
|
|
277
|
+
|
|
278
|
+
[[package]]
|
|
279
|
+
name = "tracing-core"
|
|
280
|
+
version = "0.1.36"
|
|
281
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
282
|
+
checksum = "db97caf9d906fbde555dd62fa95ddba9eecfd14cb388e4f491a66d74cd5fb79a"
|
|
283
|
+
dependencies = [
|
|
284
|
+
"once_cell",
|
|
285
|
+
]
|
|
286
|
+
|
|
287
|
+
[[package]]
|
|
288
|
+
name = "unicode-ident"
|
|
289
|
+
version = "1.0.24"
|
|
290
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
|
291
|
+
checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
|
pdf_rs-0.1.0/Cargo.toml
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
1
|
+
[package]
|
|
2
|
+
name = "pdf-rs"
|
|
3
|
+
version = "0.1.0"
|
|
4
|
+
edition = "2024"
|
|
5
|
+
description = "Rust PDF parser with Python bindings and LLM-friendly extraction"
|
|
6
|
+
readme = "README.md"
|
|
7
|
+
repository = "https://github.com/MosRat/pdf-rs"
|
|
8
|
+
keywords = ["pdf", "parser", "markdown", "pyo3"]
|
|
9
|
+
categories = ["parser-implementations"]
|
|
10
|
+
|
|
11
|
+
[lib]
|
|
12
|
+
crate-type = ["rlib", "cdylib"]
|
|
13
|
+
|
|
14
|
+
[[bin]]
|
|
15
|
+
name = "pdf-rs-cli"
|
|
16
|
+
path = "src/main.rs"
|
|
17
|
+
|
|
18
|
+
[dependencies]
|
|
19
|
+
flate2 = "1.1.9"
|
|
20
|
+
memmap2 = "0.9.10"
|
|
21
|
+
mimalloc = { version = "0.1.50", optional = true }
|
|
22
|
+
pyo3 = { version = "0.28.3", features = ["extension-module", "abi3-py39"], optional = true }
|
|
23
|
+
thiserror = "2.0.18"
|
|
24
|
+
tracing = "0.1.44"
|
|
25
|
+
|
|
26
|
+
[features]
|
|
27
|
+
mimalloc = ["dep:mimalloc"]
|
|
28
|
+
python = ["pyo3"]
|
|
29
|
+
pyo3 = ["dep:pyo3"]
|