Python製PDF解析ライブラリslateをPython3対応する

まずはcloseします。

git clone [email protected]:TakesxiSximada/slate3k.git

Cloning into 'slate3k'...
remote: Counting objects: 195, done.
remote: Total 195 (delta 0), reused 0 (delta 0), pack-reused 195
Receiving objects: 100% (195/195), 66.84 KiB | 0 bytes/s, done.
Resolving deltas: 100% (93/93), done.
Checking connectivity... done.

py3kブランチを作成します。

git checkout -b py3k

Switched to a new branch 'py3k'

作成したブランチをリモートにpushします。

git push --set-upstream origin py3k

Total 0 (delta 0), reused 0 (delta 0)
To [email protected]:TakesxiSximada/slate3k.git
 * [new branch]      py3k -> py3k
Branch py3k set up to track remote branch py3k from origin.

https://github.com/timClicks/slate/pull/32 のパッチは欲しいのでこれを取り込みます。 Pull Requestは`hub am`コマンドを使うことでパッチを当てられます。

hub am -3 https://github.com/timClicks/slate/pull/32

Applying: Fix tests on python 3.5
Applying: add a .gitignore file
Applying: Convert README to reStructuredText so it looks better on github
Applying: Fix unit tests
Applying: Fix demo code
Applying: Fix setup.py so it uses README.rst as readme file
(py3.6.0) $ git push
Counting objects: 24, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (22/22), done.
Writing objects: 100% (24/24), 3.36 KiB | 0 bytes/s, done.
Total 24 (delta 13), reused 0 (delta 0)
remote: Resolving deltas: 100% (13/13), completed with 6 local objects.
To [email protected]:TakesxiSximada/slate3k.git
   3198bfd..c31dbb7  py3k -> py3k

これでPull Requstを取り込めました。slateと名前が衝突するのは面倒なのでslate3kに変更します。まずはソースディレクトリをsrc/slateからsrc/slate3kに変更しました。 その他、修正箇所は以下になります。

git show d5f982e753f6fc22e3c5b08a052d226a53821529

commit d5f982e753f6fc22e3c5b08a052d226a53821529
Author: TakesxiSximada <[email protected]>
Date:   Mon Jan 23 21:02:06 2017 +0900

    rename packaeg name to slate3k

diff --git a/setup.py b/setup.py
index 2a89881..0ec5617 100644
--- a/setup.py
+++ b/setup.py
@@ -12,7 +12,7 @@ else:
 with open('README.rst') as f:
     long_description = f.read()

-setup(name='slate',
+setup(name='slate3k',
       version='0.5.2',
       description='Extract text from PDF documents easily.',
       author='Tim McNamara',

git show a9e79a2d222c9de705ebc244bec1edf1b28e9ea5...

commit c1259d4cbccf370492f19727d95d27022a8a46b5
Author: TakesxiSximada <[email protected]>
Date:   Mon Jan 23 21:06:37 2017 +0900

    rename slate to slate3k

diff --git a/src/slate3k/classes.py b/src/slate3k/classes.py
index 27be417..af32cd0 100644
--- a/src/slate3k/classes.py
+++ b/src/slate3k/classes.py
@@ -23,7 +23,7 @@ try:
 except ImportError:
     from pdfminer.pdfpage import PDFPage
 if PYTHON_3:
-    import slate.utils as utils
+    import .utils as utils
 else:
     import utils

diff --git a/src/slate3k/test_slate.py b/src/slate3k/test_slate.py
index 8416da2..6622e31 100644
--- a/src/slate3k/test_slate.py
+++ b/src/slate3k/test_slate.py
@@ -12,7 +12,7 @@ PYTHON_3 = sys.version_info[0] == 3
 import os

 if PYTHON_3:
-    from slate.classes import PDF
+    from .classes import PDF
 else:
     from classes import PDF

@@ -44,5 +44,3 @@ def get_pdf_path(pdf_file):
     return os.path.join(
         os.path.dirname(__file__),
         pdf_file)
-
-
diff --git a/src/slate3k/unittests.py b/src/slate3k/unittests.py
index 29c2b6b..78dcd68 100644
--- a/src/slate3k/unittests.py
+++ b/src/slate3k/unittests.py
@@ -1,7 +1,7 @@
 import unittest

 import os
-from slate import PDF
+from . import PDF

 class TestSlate(unittest.TestCase):
     def setUp(self):

git show

commit 3960b3eb3f5182278995030ba4de545f41f9b4b7
Author: TakesxiSximada <[email protected]>
Date:   Mon Jan 23 21:15:46 2017 +0900

    Can not use distribute

diff --git a/setup.py b/setup.py
index 0ec5617..2baa9ff 100644
--- a/setup.py
+++ b/setup.py
@@ -23,7 +23,7 @@ setup(name='slate3k',
       packages=find_packages('src'),
       package_dir={'': 'src'},
       requires=[pdfminer],
-      install_requires=['distribute', pdfminer],
+      install_requires=[pdfminer],
       classifiers= [
         'Development Status :: 4 - Beta',
         'Intended Audience :: Developers',

ではediable installします。editable installは開発用のインストール機能です。インストールしたライブラリのコードはsite-packagesディレクトリにコピーされるのですが、editable installの場合はコードはコピーされず、site-packages配下にはegg-linkが作成されます。ソースコード自体は元のコードが使われるのでライブラリとして使いながら開発できるようになります。

pip install -e .

Obtaining file:///Users/sximada/ng2/src/github.com/TakesxiSximada/slate3k
Collecting pdfminer3k (from slate3k==0.5.2)
  Using cached pdfminer3k-1.3.1.tar.gz
Requirement already satisfied: pytest>=2.0 in /Users/sximada/ng2/var/lib/miniconda3/envs/py3.6.0/lib/python3.6/site-packages (from pdfminer3k->slate3k==0.5.2)
Requirement already satisfied: ply>=3.4 in /Users/sximada/ng2/var/lib/miniconda3/envs/py3.6.0/lib/python3.6/site-packages (from pdfminer3k->slate3k==0.5.2)
Requirement already satisfied: py>=1.4.29 in /Users/sximada/ng2/var/lib/miniconda3/envs/py3.6.0/lib/python3.6/site-packages (from pytest>=2.0->pdfminer3k->slate3k==0.5.2)
Building wheels for collected packages: pdfminer3k
  Running setup.py bdist_wheel for pdfminer3k ... done
  Stored in directory: /Users/sximada/Library/Caches/pip/wheels/cd/84/67/3eb20c984d51d38db1ca65ecba0d866407f46d8b3a9e72f7b2
Successfully built pdfminer3k
Installing collected packages: pdfminer3k, slate3k
  Running setup.py develop for slate3k
Successfully installed pdfminer3k-1.3.1 slate3k

インストールできました。