python-calamine 0.3.2__tar.gz → 0.5.0__tar.gz

{python_calamine-0.3.2 → python_calamine-0.5.0}/.github/workflows/CI.yml

@@ -13,33 +13,29 @@ on:
 
 jobs:
   test:
-    name: test ${{ matrix.python-version }} pandas ${{ matrix.pandas-version }} ${{ matrix.disable-gil && 'free-threaded' || '' }}
+    name: test ${{ matrix.python-version }} pandas ${{ matrix.pandas-version }}
     strategy:
       fail-fast: false
       matrix:
         python-version:
-          - "3.9"
           - "3.10"
           - "3.11"
           - "3.12"
+          - "3.13"
+          - "3.14"
         pandas-version:
-          - "2.1.*"
           - "2.2.*"
+          - "2.3.*"
         include:
-          - python-version: "3.8"
-            pandas-version: "none"
           # https://github.com/pandas-dev/pandas/issues/42509
-          - python-version: "pypy3.9"
+          - python-version: "pypy3.11"
             pandas-version: "none"
-          # https://github.com/pandas-dev/pandas/issues/42509
-          - python-version: "pypy3.10"
+          # https://github.com/pandas-dev/pandas/issues/60016
+          - python-version: "3.13t"
             pandas-version: "none"
-          - python-version: "3.13"
-            pandas-version: "2.2.*"
           # https://github.com/pandas-dev/pandas/issues/60016
-          - python-version: "3.13"
+          - python-version: "3.14t"
             pandas-version: "none"
-            disable-gil: true
 
     runs-on: ubuntu-latest
 
@@ -47,7 +43,7 @@ jobs:
       PYTHON: ${{ matrix.python-version }}
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v5
 
       - name: install rust stable
         uses: dtolnay/rust-toolchain@master
@@ -61,29 +57,22 @@ jobs:
           key: v3
 
       - name: set up python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         if: "${{ !matrix.disable-gil }}"
         with:
           python-version: ${{ matrix.python-version }}
           allow-prereleases: true
 
-      - name: set up python (free-threaded)
-        uses: deadsnakes/action@v3.2.0
-        if: "${{ matrix.disable-gil }}"
-        with:
-          python-version: ${{ matrix.python-version }}
-          nogil: ${{ matrix.disable-gil }}
-
       - id: cache-py
         name: cache python
         uses: actions/cache@v4
-        if: "${{ !matrix.disable-gil }}"
         with:
           path: ${{ env.pythonLocation }}
           key: >
             py
             ${{ runner.os }}
-            ${{ env.pythonLocation }}
+            ${{ matrix.python-version }}
+            ${{ matrix.pandas-version }}
 
       - run: |
           if [ ${{ matrix.pandas-version }} == 'none' ]; then
@@ -105,7 +94,7 @@ jobs:
     runs-on: ubuntu-latest
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v5
 
       - name: install rust stable
         uses: dtolnay/rust-toolchain@stable
@@ -115,7 +104,7 @@ jobs:
       - name: cache rust
         uses: Swatinem/rust-cache@v2
 
-      - uses: actions/setup-python@v5
+      - uses: actions/setup-python@v6
         with:
           python-version: "3.13"
 
@@ -155,7 +144,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        os: [ubuntu, macos, windows]
+        os: [ubuntu, macos]
         target: [x86_64, aarch64]
         manylinux: [auto]
         include:
@@ -163,15 +152,17 @@ jobs:
             platform: linux
           - os: windows
             ls: dir
+            target: x86_64
+            interpreter: 3.10 3.11 3.12 3.13 3.13t
           - os: windows
             ls: dir
             target: i686
             python-architecture: x86
-            interpreter: 3.8 3.9 3.10 3.11 3.12 3.13
+            interpreter: 3.10 3.11 3.12 3.13 3.13t
           - os: windows
             ls: dir
             target: aarch64
-            interpreter: 3.11 3.12 3.13
+            interpreter: 3.11 3.12 3.13 3.13t
           - os: ubuntu
             platform: linux
             target: i686
@@ -181,7 +172,7 @@ jobs:
           - os: ubuntu
             platform: linux
             target: armv7
-            interpreter: 3.8 3.9 3.10 3.11 3.12 3.13
+            interpreter: 3.10 3.11 3.12 3.13 3.13t 3.14 3.14t pypy3.11
             # musllinux
           - os: ubuntu
             platform: linux
@@ -194,22 +185,18 @@ jobs:
           - os: ubuntu
             platform: linux
             target: ppc64le
-            interpreter: 3.8 3.9 3.10 3.11 3.12 3.13
+            interpreter: 3.10 3.11 3.12 3.13 3.13t 3.14 3.14t pypy3.11
           - os: ubuntu
             platform: linux
             target: s390x
-            interpreter: 3.8 3.9 3.10 3.11 3.12 3.13
-        exclude:
-          # Windows on arm64 only supports Python 3.11+
-          - os: windows
-            target: aarch64
+            interpreter: 3.10 3.11 3.12 3.13 3.13t 3.14 3.14t pypy3.11
 
     runs-on: ${{ matrix.os }}-latest
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v5
 
       - name: set up python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
           python-version: "3.13"
           architecture: ${{ matrix.python-architecture || 'x64' }}
@@ -229,7 +216,7 @@ jobs:
         with:
           target: ${{ matrix.target }}
           manylinux: ${{ matrix.manylinux || 'auto' }}
-          args: --release --out dist --interpreter ${{ matrix.interpreter || '3.8 3.9 3.10 3.11 3.12 3.13 pypy3.9 pypy3.10' }}
+          args: --release --out dist --interpreter ${{ matrix.interpreter || '3.10 3.11 3.12 3.13 3.13t 3.14 3.14t pypy3.11' }}
           rust-toolchain: stable
 
       - run: ${{ matrix.ls || 'ls -lh' }} dist/
@@ -256,7 +243,7 @@ jobs:
 
     steps:
       - name: Download All Artifacts
-        uses: actions/download-artifact@v4
+        uses: actions/download-artifact@v5
         with:
           path: dist
           pattern: wheels-*

{python_calamine-0.3.2 → python_calamine-0.5.0}/.pre-commit-config.yaml

@@ -38,8 +38,6 @@ repos:
       rev: v1.8.0
       hooks:
         - id: mypy
-          additional_dependencies:
-            - pandas-stubs
           exclude: ^tests/.*$
     -   repo: local
         hooks:

{python_calamine-0.3.2 → python_calamine-0.5.0}/Cargo.lock

@@ -70,8 +70,9 @@ checksum = "1fd0f2584146f6f2ef48085050886acf353beff7305ebd1ae69500e27c67f64b"
 
 [[package]]
 name = "calamine"
-version = "0.26.1"
-source = "git+https://github.com/tafia/calamine.git?rev=5cda5c734661f656bdfb7ac7a1b95e77bf89aa09#5cda5c734661f656bdfb7ac7a1b95e77bf89aa09"
+version = "0.30.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a9aeb09f84576a634da713630e11e431a744b91f1f8114c2ff0760189783a8a1"
 dependencies = [
  "atoi_simd",
  "byteorder",
@@ -133,9 +134,9 @@ checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
 
 [[package]]
 name = "chrono"
-version = "0.4.40"
+version = "0.4.41"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1a7964611d71df112cb1730f2ee67324fcf4d0fc6606acbbe9bfe06df124637c"
+checksum = "c469d952047f47f91b68d1cba3f10d63c11d73e4636f24f08daf0278abf01c4d"
 dependencies = [
  "android-tzdata",
  "iana-time-zone",
@@ -170,12 +171,6 @@ dependencies = [
  "cfg-if",
 ]
 
-[[package]]
-name = "crossbeam-utils"
-version = "0.8.21"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d0a5c400df2834b80a4c3327b3aad3a4c4cd4de0629063962b03235697506a28"
-
 [[package]]
 name = "derive_arbitrary"
 version = "1.4.1"
@@ -187,17 +182,6 @@ dependencies = [
  "syn",
 ]
 
-[[package]]
-name = "displaydoc"
-version = "0.2.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "97369cbbc041bc366949bc74d34658d6cda5621039731c6310521892a3a20ae0"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
 [[package]]
 name = "encoding_rs"
 version = "0.8.35"
@@ -246,11 +230,12 @@ checksum = "37909eebbb50d72f9059c3b6d82c0463f2ff062c9e95845c43a6c9c0355411be"
 
 [[package]]
 name = "flate2"
-version = "1.0.35"
+version = "1.1.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c936bfdafb507ebbf50b8074c54fa31c5be9a1e7e5f467dd659697041407d07c"
+checksum = "4a3d7db9596fecd151c5f638c0ee5d5bd487b6e0ea232e5dc96d5250f6f94b1d"
 dependencies = [
  "crc32fast",
+ "libz-rs-sys",
  "miniz_oxide",
 ]
 
@@ -344,6 +329,15 @@ version = "0.2.169"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b5aba8db14291edd000dfcc4d620c7ebfb122c613afb886ca8803fa4e128a20a"
 
+[[package]]
+name = "libz-rs-sys"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "172a788537a2221661b480fee8dc5f96c580eb34fa88764d3205dc356c7e4221"
+dependencies = [
+ "zlib-rs",
+]
+
 [[package]]
 name = "linux-raw-sys"
 version = "0.4.15"
@@ -379,9 +373,9 @@ dependencies = [
 
 [[package]]
 name = "miniz_oxide"
-version = "0.8.2"
+version = "0.8.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4ffbe83022cedc1d264172192511ae958937694cd57ce297164951b8b3568394"
+checksum = "1fa76a2c86f704bdb222d66965fb3d63269ce38518b83cb0575fca855ebb6316"
 dependencies = [
  "adler2",
 ]
@@ -397,9 +391,9 @@ dependencies = [
 
 [[package]]
 name = "once_cell"
-version = "1.20.2"
+version = "1.21.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1261fe7e33c73b354eab43b1273a57c8f967d0391e80353e51f764ac02cf6775"
+checksum = "42f5e15c9953c5e4ccceeb2e7382a716482c34515315f7b03532b8b4e8393d2d"
 
 [[package]]
 name = "portable-atomic"
@@ -429,11 +423,10 @@ dependencies = [
 
 [[package]]
 name = "pyo3"
-version = "0.24.1"
+version = "0.26.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "17da310086b068fbdcefbba30aeb3721d5bb9af8db4987d6735b2183ca567229"
+checksum = "7ba0117f4212101ee6544044dae45abe1083d30ce7b29c4b5cbdfa2354e07383"
 dependencies = [
- "cfg-if",
  "chrono",
  "indoc",
  "libc",
@@ -448,20 +441,19 @@ dependencies = [
 
 [[package]]
 name = "pyo3-build-config"
-version = "0.24.1"
+version = "0.26.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e27165889bd793000a098bb966adc4300c312497ea25cf7a690a9f0ac5aa5fc1"
+checksum = "4fc6ddaf24947d12a9aa31ac65431fb1b851b8f4365426e182901eabfb87df5f"
 dependencies = [
- "once_cell",
  "python3-dll-a",
  "target-lexicon",
 ]
 
 [[package]]
 name = "pyo3-ffi"
-version = "0.24.1"
+version = "0.26.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "05280526e1dbf6b420062f3ef228b78c0c54ba94e157f5cb724a609d0f2faabc"
+checksum = "025474d3928738efb38ac36d4744a74a400c901c7596199e20e45d98eb194105"
 dependencies = [
  "libc",
  "pyo3-build-config",
@@ -469,9 +461,8 @@ dependencies = [
 
 [[package]]
 name = "pyo3-file"
-version = "0.12.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8d97fbb05588c29774af01833e598fae7e714a3f87d36b286e0466f6b678e4a7"
+version = "0.13.0"
+source = "git+https://github.com/dimastbk/pyo3-file.git?rev=da8ddb5d774d042555b867dd6571f3ce128bc349#da8ddb5d774d042555b867dd6571f3ce128bc349"
 dependencies = [
  "pyo3",
  "skeptic",
@@ -479,9 +470,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros"
-version = "0.24.1"
+version = "0.26.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5c3ce5686aa4d3f63359a5100c62a127c9f15e8398e5fdeb5deef1fed5cd5f44"
+checksum = "2e64eb489f22fe1c95911b77c44cc41e7c19f3082fc81cce90f657cdc42ffded"
 dependencies = [
  "proc-macro2",
  "pyo3-macros-backend",
@@ -491,9 +482,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros-backend"
-version = "0.24.1"
+version = "0.26.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f4cf6faa0cbfb0ed08e89beb8103ae9724eb4750e3a78084ba4017cbe94f3855"
+checksum = "100246c0ecf400b475341b8455a9213344569af29a3c841d29270e53102e0fcf"
 dependencies = [
  "heck",
  "proc-macro2",
@@ -504,7 +495,7 @@ dependencies = [
 
 [[package]]
 name = "python-calamine"
-version = "0.3.2"
+version = "0.5.0"
 dependencies = [
  "calamine",
  "chrono",
@@ -668,26 +659,6 @@ dependencies = [
  "windows-sys",
 ]
 
-[[package]]
-name = "thiserror"
-version = "2.0.10"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a3ac7f54ca534db81081ef1c1e7f6ea8a3ef428d2fc069097c079443d24124d3"
-dependencies = [
- "thiserror-impl",
-]
-
-[[package]]
-name = "thiserror-impl"
-version = "2.0.10"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9e9465d30713b56a37ede7185763c3492a91be2f5fa68d958c44e41ab9248beb"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
 [[package]]
 name = "unicase"
 version = "2.8.1"
@@ -881,21 +852,24 @@ checksum = "589f6da84c646204747d1270a2a5661ea66ed1cced2631d546fdfb155959f9ec"
 
 [[package]]
 name = "zip"
-version = "2.4.1"
+version = "4.5.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "938cc23ac49778ac8340e366ddc422b2227ea176edb447e23fc0627608dddadd"
+checksum = "8835eb39822904d39cb19465de1159e05d371973f0c6df3a365ad50565ddc8b9"
 dependencies = [
  "arbitrary",
  "crc32fast",
- "crossbeam-utils",
- "displaydoc",
  "flate2",
  "indexmap",
  "memchr",
- "thiserror",
  "zopfli",
 ]
 
+[[package]]
+name = "zlib-rs"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "626bd9fa9734751fc50d6060752170984d7053f5a39061f524cda68023d4db8a"
+
 [[package]]
 name = "zopfli"
 version = "0.8.1"

python_calamine-0.5.0/Cargo.toml

@@ -0,0 +1,22 @@
+[package]
+name = "python-calamine"
+version = "0.5.0"
+edition = "2021"
+readme = "README.md"
+
+[lib]
+name = "python_calamine"
+crate-type = ["cdylib"]
+
+[dependencies]
+calamine = { version = "0.30.0", features = ["dates"] }
+pyo3 = { version = "0.26.0", features = [
+    "extension-module",
+    "chrono",
+    "generate-import-lib",
+] }
+chrono = { version = "0.4.38", features = ["serde"] }
+pyo3-file = { git = "https://github.com/dimastbk/pyo3-file.git", rev = "da8ddb5d774d042555b867dd6571f3ce128bc349" }
+
+[build-dependencies]
+pyo3-build-config = "0.26.0"

{python_calamine-0.3.2 → python_calamine-0.5.0}/PKG-INFO

@@ -1,31 +1,24 @@
 Metadata-Version: 2.4
 Name: python-calamine
-Version: 0.3.2
+Version: 0.5.0
 Classifier: Development Status :: 4 - Beta
-Classifier: License :: OSI Approved :: MIT License
 Classifier: Programming Language :: Rust
-Classifier: Programming Language :: Python :: 3.8
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
-Requires-Dist: packaging>=23.1
+Classifier: Programming Language :: Python :: 3.14
 Requires-Dist: maturin~=1.0 ; extra == 'dev'
-Requires-Dist: pre-commit~=3.0 ; extra == 'dev'
+Requires-Dist: pre-commit~=4.3 ; extra == 'dev'
 Requires-Dist: pytest~=8.0 ; extra == 'dev'
-Requires-Dist: pandas[excel]~=2.0 ; extra == 'dev'
-Requires-Dist: numpy~=1.0 ; extra == 'dev'
+Requires-Dist: pandas[excel]~=2.2 ; extra == 'dev'
 Provides-Extra: dev
 License-File: LICENSE
 Summary: Python binding for Rust's library for reading excel and odf file - calamine
-Home-Page: https://github.com/dimastbk/python-calamine
-Author: Dmitriy <dimastbk@proton.me>
 Author-email: Dmitriy <dimastbk@proton.me>
-License: MIT
-Requires-Python: >=3.8
+License-Expression: MIT
+Requires-Python: >=3.10
 Description-Content-Type: text/markdown; charset=UTF-8; variant=GFM
-Project-URL: Source Code, https://github.com/dimastbk/python-calamine
 
 # python-calamine
 [![PyPI - Version](https://img.shields.io/pypi/v/python-calamine)](https://pypi.org/project/python-calamine/)
@@ -79,20 +72,7 @@ workbook = CalamineWorkbook.from_path("file.xlsx").get_sheet_by_name("Sheet1").t
 # ]
 ```
 
-Also, you can use monkeypatch for pandas for use this library as engine in `read_excel()` (only pandas 2.0 and 2.1 are supported).
 Pandas 2.2 and above have built-in support of python-calamine.
-```python
-from pandas import read_excel
-from python_calamine.pandas import pandas_monkeypatch
-
-
-pandas_monkeypatch()
-
-read_excel("file.xlsx", engine="calamine")
-#            1   2   3   4   5   6   7
-# 0          1   2   3   4   5   6   7
-# 1          1   2   3   4   5   6   7
-```
 
 Also, you can find additional examples in [tests](https://github.com/dimastbk/python-calamine/blob/master/tests/test_base.py).
 

{python_calamine-0.3.2 → python_calamine-0.5.0}/README.md

@@ -50,20 +50,7 @@ workbook = CalamineWorkbook.from_path("file.xlsx").get_sheet_by_name("Sheet1").t
 # ]
 ```
 
-Also, you can use monkeypatch for pandas for use this library as engine in `read_excel()` (only pandas 2.0 and 2.1 are supported).
 Pandas 2.2 and above have built-in support of python-calamine.
-```python
-from pandas import read_excel
-from python_calamine.pandas import pandas_monkeypatch
-
-
-pandas_monkeypatch()
-
-read_excel("file.xlsx", engine="calamine")
-#            1   2   3   4   5   6   7
-# 0          1   2   3   4   5   6   7
-# 1          1   2   3   4   5   6   7
-```
 
 Also, you can find additional examples in [tests](https://github.com/dimastbk/python-calamine/blob/master/tests/test_base.py).
 

{python_calamine-0.3.2 → python_calamine-0.5.0}/pyproject.toml

@@ -1,27 +1,27 @@
 [project]
 name = "python-calamine"
-requires-python = ">=3.8"
+description = "Python binding for Rust's library for reading excel and odf file - calamine"
+readme = "README.md"
+requires-python = ">=3.10"
+license = "MIT"
+authors = [{ name = "Dmitriy", email = "dimastbk@proton.me" }]
 classifiers = [
   "Development Status :: 4 - Beta",
-  "License :: OSI Approved :: MIT License",
   "Programming Language :: Rust",
-  "Programming Language :: Python :: 3.8",
-  "Programming Language :: Python :: 3.9",
   "Programming Language :: Python :: 3.10",
   "Programming Language :: Python :: 3.11",
   "Programming Language :: Python :: 3.12",
   "Programming Language :: Python :: 3.13",
+  "Programming Language :: Python :: 3.14",
 ]
-dependencies = ["packaging>=23.1"]
 dynamic = ["version"]
 
 [project.optional-dependencies]
 dev = [
   "maturin~=1.0",
-  "pre-commit~=3.0",
+  "pre-commit~=4.3",
   "pytest~=8.0",
-  "pandas[excel]~=2.0",
-  "numpy~=1.0",
+  "pandas[excel]~=2.2",
 ]
 
 [tool.isort]
@@ -31,7 +31,7 @@ multi_line_output = 3
 profile = "black"
 
 [tool.mypy]
-python_version = "3.8"
+python_version = "3.10"
 ignore_missing_imports = false
 disallow_untyped_defs = true
 check_untyped_defs = true

{python_calamine-0.3.2 → python_calamine-0.5.0}/python/python_calamine/_python_calamine.pyi

@@ -89,6 +89,18 @@ class CalamineSheet:
     ]:
         """Retunrning data from sheet as iterator of lists."""
 
+    @property
+    def merged_cell_ranges(
+        self,
+    ) -> list[tuple[tuple[int, int], tuple[int, int]]] | None:
+        """Return a copy of merged cell ranges.
+
+        Support only for xlsx/xls.
+
+        Returns:
+            list of merged cell ranges (tuple[start coordinate, end coordinate]) or None for unsuported format
+        """
+
 @typing.final
 class CalamineWorkbook(contextlib.AbstractContextManager):
     path: str | None

{python_calamine-0.3.2 → python_calamine-0.5.0}/src/lib.rs

@@ -8,7 +8,7 @@ use crate::types::{
 };
 
 #[pyfunction]
-fn load_workbook(py: Python, path_or_filelike: PyObject) -> PyResult<CalamineWorkbook> {
+fn load_workbook(py: Python, path_or_filelike: Py<PyAny>) -> PyResult<CalamineWorkbook> {
     CalamineWorkbook::from_object(py, path_or_filelike)
 }
 

{python_calamine-0.3.2 → python_calamine-0.5.0}/src/types/cell.rs

@@ -1,8 +1,16 @@
 use std::convert::From;
 
 use calamine::DataType;
+use chrono::Datelike;
 use pyo3::prelude::*;
 
+/// https://docs.python.org/3/library/datetime.html#constants
+/// The smallest year number allowed in a date or datetime object. MINYEAR is 1.
+const MINYEAR: i32 = 1;
+
+/// The largest year number allowed in a date or datetime object. MAXYEAR is 9999.
+const MAXYEAR: i32 = 9999;
+
 #[derive(Debug, Clone)]
 pub enum CellValue {
     Int(i64),
@@ -16,6 +24,14 @@ pub enum CellValue {
     Empty,
 }
 
+fn check_year_range<DT: Datelike>(value: DT) -> Option<DT> {
+    if value.year() < MINYEAR || value.year() > MAXYEAR {
+        None
+    } else {
+        Some(value)
+    }
+}
+
 impl<'py> IntoPyObject<'py> for CellValue {
     type Target = PyAny;
     type Output = Bound<'py, Self::Target>;
@@ -64,19 +80,31 @@ where
             } else if v < 1.0 {
                 value.as_time().map(CellValue::Time)
             } else if v == (v as u64) as f64 {
-                value.as_date().map(CellValue::Date)
+                value
+                    .as_date()
+                    .and_then(check_year_range)
+                    .map(CellValue::Date)
             } else {
-                value.as_datetime().map(CellValue::DateTime)
+                value
+                    .as_datetime()
+                    .and_then(check_year_range)
+                    .map(CellValue::DateTime)
             }
             .unwrap_or(CellValue::Float(v))
         } else if value.is_datetime_iso() {
             let v = value.get_datetime_iso().unwrap();
             if v.contains('T') {
-                value.as_datetime().map(CellValue::DateTime)
+                value
+                    .as_datetime()
+                    .and_then(check_year_range)
+                    .map(CellValue::DateTime)
             } else if v.contains(':') {
                 value.as_time().map(CellValue::Time)
             } else {
-                value.as_date().map(CellValue::Date)
+                value
+                    .as_date()
+                    .and_then(check_year_range)
+                    .map(CellValue::Date)
             }
             .unwrap_or(CellValue::String(v.to_owned()))
         } else if value.is_duration_iso() {

{python_calamine-0.3.2 → python_calamine-0.5.0}/src/types/sheet.rs

@@ -1,7 +1,7 @@
 use std::fmt::Display;
 use std::sync::Arc;
 
-use calamine::{Data, Range, Rows, SheetType, SheetVisible};
+use calamine::{Data, Dimensions, Range, Rows, SheetType, SheetVisible};
 use pyo3::class::basic::CompareOp;
 use pyo3::prelude::*;
 use pyo3::types::PyList;
@@ -25,7 +25,7 @@ pub enum SheetTypeEnum {
 
 impl Display for SheetTypeEnum {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(f, "SheetTypeEnum.{:?}", self)
+        write!(f, "SheetTypeEnum.{self:?}")
     }
 }
 
@@ -54,7 +54,7 @@ pub enum SheetVisibleEnum {
 
 impl Display for SheetVisibleEnum {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(f, "SheetVisibleEnum.{:?}", self)
+        write!(f, "SheetVisibleEnum.{self:?}")
     }
 }
 
@@ -68,6 +68,8 @@ impl From<SheetVisible> for SheetVisibleEnum {
     }
 }
 
+type MergedCellRange = ((u32, u32), (u32, u32));
+
 #[pyclass]
 #[derive(Clone, PartialEq)]
 pub struct SheetMetadata {
@@ -132,13 +134,19 @@ pub struct CalamineSheet {
     #[pyo3(get)]
     name: String,
     range: Arc<Range<Data>>,
+    merged_cell_ranges: Option<Vec<Dimensions>>,
 }
 
 impl CalamineSheet {
-    pub fn new(name: String, range: Range<Data>) -> Self {
+    pub fn new(
+        name: String,
+        range: Range<Data>,
+        merged_cell_ranges: Option<Vec<Dimensions>>,
+    ) -> Self {
         CalamineSheet {
             name,
             range: Arc::new(range),
+            merged_cell_ranges,
         }
     }
 }
@@ -212,6 +220,13 @@ impl CalamineSheet {
     fn iter_rows(&self) -> CalamineCellIterator {
         CalamineCellIterator::from_range(Arc::clone(&self.range))
     }
+
+    #[getter]
+    fn merged_cell_ranges(slf: PyRef<'_, Self>) -> Option<Vec<MergedCellRange>> {
+        slf.merged_cell_ranges
+            .as_ref()
+            .map(|r| r.iter().map(|d| (d.start, d.end)).collect())
+    }
 }
 
 #[pyclass]

{python_calamine-0.3.2 → python_calamine-0.5.0}/src/types/workbook.rs

@@ -2,7 +2,9 @@ use std::fs::File;
 use std::io::{BufReader, Cursor, Read};
 use std::path::PathBuf;
 
-use calamine::{open_workbook_auto, open_workbook_auto_from_rs, Reader, Sheets};
+use calamine::{
+    open_workbook_auto, open_workbook_auto_from_rs, Error as CalamineCrateError, Reader, Sheets,
+};
 use pyo3::exceptions::PyTypeError;
 use pyo3::prelude::*;
 use pyo3::types::{PyString, PyType};
@@ -44,6 +46,35 @@ impl SheetsEnum {
             SheetsEnum::None => Err(Error::WorkbookClosed),
         }
     }
+
+    fn worksheet_merge_cells(
+        &mut self,
+        name: &str,
+    ) -> Result<Option<Vec<calamine::Dimensions>>, Error> {
+        match self {
+            SheetsEnum::File(f) => match f {
+                Sheets::Xls(xls_f) => Ok(xls_f.worksheet_merge_cells(name)),
+                Sheets::Xlsx(xlsx_f) => xlsx_f
+                    .worksheet_merge_cells(name)
+                    .transpose()
+                    .map_err(CalamineCrateError::Xlsx)
+                    .map_err(Error::Calamine)
+                    .map(|inner| inner.or(Some(Vec::new()))),
+                _ => Ok(None),
+            },
+            SheetsEnum::FileLike(f) => match f {
+                Sheets::Xls(xls_f) => Ok(xls_f.worksheet_merge_cells(name)),
+                Sheets::Xlsx(xlsx_f) => xlsx_f
+                    .worksheet_merge_cells(name)
+                    .transpose()
+                    .map_err(CalamineCrateError::Xlsx)
+                    .map_err(Error::Calamine)
+                    .map(|inner| inner.or(Some(Vec::new()))),
+                _ => Ok(None),
+            },
+            SheetsEnum::None => Err(Error::WorkbookClosed),
+        }
+    }
 }
 
 #[pyclass]
@@ -61,7 +92,7 @@ pub struct CalamineWorkbook {
 impl CalamineWorkbook {
     fn __repr__(&self) -> PyResult<String> {
         match &self.path {
-            Some(path) => Ok(format!("CalamineWorkbook(path='{}')", path)),
+            Some(path) => Ok(format!("CalamineWorkbook(path='{path}')")),
             None => Ok("CalamineWorkbook(path='bytes')".to_string()),
         }
     }
@@ -71,7 +102,7 @@ impl CalamineWorkbook {
     fn py_from_object(
         _cls: &Bound<'_, PyType>,
         py: Python<'_>,
-        path_or_filelike: PyObject,
+        path_or_filelike: Py<PyAny>,
     ) -> PyResult<Self> {
         Self::from_object(py, path_or_filelike)
     }
@@ -81,22 +112,22 @@ impl CalamineWorkbook {
     fn py_from_filelike(
         _cls: &Bound<'_, PyType>,
         py: Python<'_>,
-        filelike: PyObject,
+        filelike: Py<PyAny>,
     ) -> PyResult<Self> {
-        py.allow_threads(|| Self::from_filelike(filelike))
+        py.detach(|| Self::from_filelike(filelike))
     }
 
     #[classmethod]
     #[pyo3(name = "from_path")]
-    fn py_from_path(_cls: &Bound<'_, PyType>, py: Python<'_>, path: PyObject) -> PyResult<Self> {
+    fn py_from_path(_cls: &Bound<'_, PyType>, py: Python<'_>, path: Py<PyAny>) -> PyResult<Self> {
         if let Ok(string_ref) = path.downcast_bound::<PyString>(py) {
             let path = string_ref.to_string_lossy().to_string();
-            return py.allow_threads(|| Self::from_path(&path));
+            return py.detach(|| Self::from_path(&path));
         }
 
         if let Ok(string_ref) = path.extract::<PathBuf>(py) {
             let path = string_ref.to_string_lossy().to_string();
-            return py.allow_threads(|| Self::from_path(&path));
+            return py.detach(|| Self::from_path(&path));
         }
 
         Err(PyTypeError::new_err(""))
@@ -104,12 +135,12 @@ impl CalamineWorkbook {
 
     #[pyo3(name = "get_sheet_by_name")]
     fn py_get_sheet_by_name(&mut self, py: Python<'_>, name: &str) -> PyResult<CalamineSheet> {
-        py.allow_threads(|| self.get_sheet_by_name(name))
+        py.detach(|| self.get_sheet_by_name(name))
     }
 
     #[pyo3(name = "get_sheet_by_index")]
     fn py_get_sheet_by_index(&mut self, py: Python<'_>, index: usize) -> PyResult<CalamineSheet> {
-        py.allow_threads(|| self.get_sheet_by_index(index))
+        py.detach(|| self.get_sheet_by_index(index))
     }
 
     fn close(&mut self) -> PyResult<()> {
@@ -129,30 +160,30 @@ impl CalamineWorkbook {
 
     fn __exit__(
         &mut self,
-        _exc_type: PyObject,
-        _exc_value: PyObject,
-        _traceback: PyObject,
+        _exc_type: Py<PyAny>,
+        _exc_value: Py<PyAny>,
+        _traceback: Py<PyAny>,
     ) -> PyResult<()> {
         self.close()
     }
 }
 
 impl CalamineWorkbook {
-    pub fn from_object(py: Python<'_>, path_or_filelike: PyObject) -> PyResult<Self> {
+    pub fn from_object(py: Python<'_>, path_or_filelike: Py<PyAny>) -> PyResult<Self> {
         if let Ok(string_ref) = path_or_filelike.downcast_bound::<PyString>(py) {
             let path = string_ref.to_string_lossy().to_string();
-            return py.allow_threads(|| Self::from_path(&path));
+            return py.detach(|| Self::from_path(&path));
         }
 
         if let Ok(string_ref) = path_or_filelike.extract::<PathBuf>(py) {
             let path = string_ref.to_string_lossy().to_string();
-            return py.allow_threads(|| Self::from_path(&path));
+            return py.detach(|| Self::from_path(&path));
         }
 
-        py.allow_threads(|| Self::from_filelike(path_or_filelike))
+        py.detach(|| Self::from_filelike(path_or_filelike))
     }
 
-    pub fn from_filelike(filelike: PyObject) -> PyResult<Self> {
+    pub fn from_filelike(filelike: Py<PyAny>) -> PyResult<Self> {
         let mut buf = vec![];
         PyFileLikeObject::with_requirements(filelike, true, false, true, false)?
             .read_to_end(&mut buf)?;
@@ -192,14 +223,19 @@ impl CalamineWorkbook {
 
     fn get_sheet_by_name(&mut self, name: &str) -> PyResult<CalamineSheet> {
         let range = self.sheets.worksheet_range(name).map_err(err_to_py)?;
-        Ok(CalamineSheet::new(name.to_owned(), range))
+        let merge_cells_range = self.sheets.worksheet_merge_cells(name).map_err(err_to_py)?;
+        Ok(CalamineSheet::new(
+            name.to_owned(),
+            range,
+            merge_cells_range,
+        ))
     }
 
     fn get_sheet_by_index(&mut self, index: usize) -> PyResult<CalamineSheet> {
         let name = self
             .sheet_names
             .get(index)
-            .ok_or_else(|| WorksheetNotFound::new_err(format!("Worksheet '{}' not found", index)))?
+            .ok_or_else(|| WorksheetNotFound::new_err(format!("Worksheet '{index}' not found")))?
             .to_string();
         self.get_sheet_by_name(&name)
     }

{python_calamine-0.3.2 → python_calamine-0.5.0}/tests/test_base.py

@@ -15,7 +15,7 @@ PATH = Path(__file__).parent / "data"
 
 
 def test_ods_read():
-    names = ["Sheet1", "Sheet2"]
+    names = ["Sheet1", "Sheet2", "Merged Cells"]
     data = [
         ["", "", "", "", "", "", "", "", "", ""],
         [
@@ -59,9 +59,11 @@ def test_ods_read():
     assert [] == reader.get_sheet_by_index(1).to_python()
     assert [] == reader.get_sheet_by_index(1).to_python(skip_empty_area=False)
 
+    assert None is reader.get_sheet_by_name("Merged Cells").merged_cell_ranges
+
 
 def test_xls_read():
-    names = ["Sheet1", "Sheet2"]
+    names = ["Sheet1", "Sheet2", "Merged Cells"]
     data = [
         ["", "", "", "", "", "", "", "", "", ""],
         [
@@ -101,9 +103,13 @@ def test_xls_read():
     assert [] == reader.get_sheet_by_index(1).to_python()
     assert [] == reader.get_sheet_by_index(1).to_python(skip_empty_area=False)
 
+    assert [((0, 0), (1, 0)), ((0, 1), (1, 1))] == reader.get_sheet_by_name(
+        "Merged Cells"
+    ).merged_cell_ranges
+
 
 def test_xlsb_read():
-    names = ["Sheet1", "Sheet2", "Sheet3"]
+    names = ["Sheet1", "Sheet2", "Sheet3", "Merged Cells"]
     data = [
         ["", "", "", "", "", "", "", "", "", ""],
         [
@@ -143,9 +149,11 @@ def test_xlsb_read():
     assert [] == reader.get_sheet_by_index(1).to_python()
     assert [] == reader.get_sheet_by_index(1).to_python(skip_empty_area=False)
 
+    assert None is reader.get_sheet_by_name("Merged Cells").merged_cell_ranges
+
 
 def test_xlsx_read():
-    names = ["Sheet1", "Sheet2", "Sheet3"]
+    names = ["Sheet1", "Sheet2", "Sheet3", "Merged Cells"]
     data = [
         ["", "", "", "", "", "", "", "", "", ""],
         [
@@ -185,9 +193,13 @@ def test_xlsx_read():
     assert [] == reader.get_sheet_by_index(1).to_python()
     assert [] == reader.get_sheet_by_index(1).to_python(skip_empty_area=False)
 
+    assert [((0, 0), (1, 0)), ((0, 1), (1, 1))] == reader.get_sheet_by_name(
+        "Merged Cells"
+    ).merged_cell_ranges
+
 
 def test_xlsx_iter_rows():
-    names = ["Sheet1", "Sheet2", "Sheet3"]
+    names = ["Sheet1", "Sheet2", "Sheet3", "Merged Cells"]
     data = [
         ["", "", "", "", "", "", "", "", "", ""],
         [
@@ -363,3 +375,10 @@ def test_close_workbook_double():
 
     with pytest.raises(WorkbookClosed):
         reader.close()
+
+
+def test_datetime_overflow():
+    reader = CalamineWorkbook.from_path(PATH / "issue139.xlsx")
+
+    # 2994626.0 = 10099-01-01
+    assert reader.get_sheet_by_index(0).to_python() == [["date"], [2994626.0]]

{python_calamine-0.3.2 → python_calamine-0.5.0}/tests/test_pandas.py

@@ -1,8 +1,6 @@
-from importlib.metadata import version
 from pathlib import Path
 
 import pytest
-from packaging.version import Version, parse
 
 try:
     import pandas as pd
@@ -14,17 +12,6 @@ except ImportError:
 PATH = Path(__file__).parent / "data"
 
 
-@pytest.fixture(scope="module", autouse=True)
-def pandas_monkeypatch():
-    if parse(version("pandas")) >= Version("2.2.0"):
-        yield
-    else:
-        from python_calamine.pandas import pandas_monkeypatch
-
-        pandas_monkeypatch()
-        yield
-
-
 @pytest.mark.skipif(not pd, reason="pandas is required")
 def test_ods_pandas(expected_df_ods):
     result = pd.read_excel(PATH / "base.ods", sheet_name="Sheet1", engine="calamine")

python_calamine-0.3.2/Cargo.toml

@@ -1,27 +0,0 @@
-[package]
-name = "python-calamine"
-version = "0.3.2"
-authors = ["Dmitriy <dimastbk@proton.me>"]
-description = "Python binding for Rust's library for reading excel and odf file - calamine"
-edition = "2021"
-homepage = "https://github.com/dimastbk/python-calamine"
-license = "MIT"
-readme = "README.md"
-repository = "https://github.com/dimastbk/python-calamine"
-
-[lib]
-name = "python_calamine"
-crate-type = ["cdylib"]
-
-[dependencies]
-calamine = { git = "https://github.com/tafia/calamine.git", rev = "5cda5c734661f656bdfb7ac7a1b95e77bf89aa09", features = ["dates"] }
-pyo3 = { version = "0.24.0", features = [
-    "extension-module",
-    "chrono",
-    "generate-import-lib",
-] }
-chrono = { version = "0.4.38", features = ["serde"] }
-pyo3-file = "0.12.0"
-
-[build-dependencies]
-pyo3-build-config = "0.24.0"

python_calamine-0.3.2/python/python_calamine/pandas.py

@@ -1,136 +0,0 @@
-from __future__ import annotations
-
-from datetime import date, datetime, time, timedelta
-from importlib.metadata import version
-from typing import TYPE_CHECKING, Union, cast
-
-import pandas as pd
-from packaging.version import Version, parse
-from pandas._typing import Scalar
-from pandas.compat._optional import import_optional_dependency
-from pandas.core.shared_docs import _shared_docs
-from pandas.io.excel import ExcelFile
-from pandas.io.excel._base import (  # type:ignore[attr-defined] # missing in pandas-stubs
-    BaseExcelReader,
-)
-from pandas.util._decorators import (  # type:ignore[attr-defined] # missing in pandas-stubs
-    doc,
-)
-
-if TYPE_CHECKING:
-    from pandas._typing import FilePath, ReadBuffer, StorageOptions
-    from python_calamine import CalamineSheet, CalamineWorkbook
-
-_CellValueT = Union[int, float, str, bool, time, date, datetime, timedelta]
-
-
-PANDAS_VERSION = parse(version("pandas"))
-
-
-class CalamineExcelReader(BaseExcelReader):
-    book: CalamineWorkbook
-
-    @doc(storage_options=_shared_docs["storage_options"])
-    def __init__(
-        self,
-        filepath_or_buffer: FilePath | ReadBuffer[bytes],
-        storage_options: StorageOptions | None = None,
-        engine_kwargs: dict | None = None,
-    ) -> None:
-        """
-        Reader using calamine engine (xlsx/xls/xlsb/ods).
-
-        Parameters
-        ----------
-        filepath_or_buffer : str, path to be parsed or
-            an open readable stream.
-        {storage_options}
-        engine_kwargs : dict, optional
-            Arbitrary keyword arguments passed to excel engine.
-        """
-        import_optional_dependency("python_calamine")
-        if PANDAS_VERSION >= Version("2.2.0"):
-            raise ValueError("Pandas >= 2.2.0 has builtin support of calamine")
-        elif PANDAS_VERSION >= Version("2.1.0"):
-            super().__init__(
-                filepath_or_buffer,
-                storage_options=storage_options,
-                engine_kwargs=engine_kwargs,
-            )
-        elif PANDAS_VERSION >= Version("2.0.0"):
-            super().__init__(
-                filepath_or_buffer,
-                storage_options=storage_options,
-            )
-        else:
-            raise ValueError("Pandas >= 2 is only supported")
-
-    @property
-    def _workbook_class(self) -> type[CalamineWorkbook]:
-        from python_calamine import CalamineWorkbook
-
-        return CalamineWorkbook
-
-    def load_workbook(
-        self,
-        filepath_or_buffer: FilePath | ReadBuffer[bytes],
-        engine_kwargs: dict | None = None,
-    ) -> CalamineWorkbook:
-        from python_calamine import load_workbook
-
-        return load_workbook(filepath_or_buffer, **(engine_kwargs or {}))
-
-    @property
-    def sheet_names(self) -> list[str]:
-        from python_calamine import SheetTypeEnum
-
-        return [
-            sheet.name
-            for sheet in self.book.sheets_metadata
-            if sheet.typ == SheetTypeEnum.WorkSheet
-        ]
-
-    def get_sheet_by_name(self, name: str) -> CalamineSheet:
-        self.raise_if_bad_sheet_by_name(name)
-        return self.book.get_sheet_by_name(name)
-
-    def get_sheet_by_index(self, index: int) -> CalamineSheet:
-        self.raise_if_bad_sheet_by_index(index)
-        return self.book.get_sheet_by_index(index)
-
-    def get_sheet_data(
-        self, sheet: CalamineSheet, file_rows_needed: int | None = None
-    ) -> list[list[Scalar]]:
-        def _convert_cell(value: _CellValueT) -> Scalar:
-            if isinstance(value, float):
-                val = int(value)
-                if val == value:
-                    return val
-                else:
-                    return value
-            elif isinstance(value, date):
-                return pd.Timestamp(value)
-            elif isinstance(value, timedelta):
-                return pd.Timedelta(value)
-            elif isinstance(value, time):
-                # cast needed here because Scalar doesn't include datetime.time
-                return cast(Scalar, value)
-
-            return value
-
-        rows: list[list[_CellValueT]] = sheet.to_python(skip_empty_area=False)
-        data: list[list[Scalar]] = []
-
-        for row in rows:
-            data.append([_convert_cell(cell) for cell in row])
-            if file_rows_needed is not None and len(data) >= file_rows_needed:
-                break
-
-        return data
-
-
-def pandas_monkeypatch() -> None:
-    ExcelFile._engines = {  # type:ignore[attr-defined] # missing in pandas-stubs
-        "calamine": CalamineExcelReader,
-        **ExcelFile._engines,  # type:ignore[attr-defined] # missing in pandas-stubs
-    }