mieki256's diary



2014/06/15() [n年前の日記]

#1 [python] Python 2.7.6でsetuptoolsやpipをインストールしようとしてハマった

以下の症状でハマりました。

_2013/12/19 Python-2.7.6がWindows環境でmimetypes.init()に失敗する場合がありsetuptoolsがインストールできない話 - 清水川Web
_[Python-ml-jp 5663] Windows 用の Python 2.7.6 に setuptools 2.0.1 を設置できなかった話
_[Python-ml-jp 5664] Re: Windows 用の Python 2.7.6 に setuptools 2.0.1 を設置できなかった話
_姫踊子草の楽屋裏 Windows 用の Python 2.7.6 に setuptools 2.0.1 を設置できなかった話
_Issue 9291: mimetypes initialization fails on Windows because of non-Latin characters in registry - Python tracker
_cpython: 18cfc2a42772

以下を参考にして作業。
ノート 2014/5/5追記

Python本家のバグチケット( http://bugs.python.org/issue9291 )がクローズされたようです。 http://hg.python.org/cpython/rev/18cfc2a42772 にある差分(か http://hg.python.org/cpython/raw-file/18cfc2a42772/Lib/mimetypes.py のファイル) をPython-2.7.6に適用したところ、問題が再現しなくなりました。

2013/12/19 Python-2.7.6がWindows環境でmimetypes.init()に失敗する場合がありsetuptoolsがインストールできない話 - 清水川Web より


_http://hg.python.org/cpython/raw-file/18cfc2a42772/Lib/mimetypes.py から、mimetypes.py をDLして、Python2.7.6インストールフォルダ\Lib\mimetypes.py を置き換えました。

と、ここまでやってたら、 _Windows 7 / Python 2.7.6 にて ez_setup.py にて UnicodeDecodeError が出て困ったけど解決した件 - secretbase.log という記事に遭遇。
Python 2.7.7 にて下記不具合として修正されていますので、2.7.7を使いましょう。

Windows 7 / Python 2.7.6 にて ez_setup.py にて UnicodeDecodeError が出て困ったけど解決した件 - secretbase.log より

2.7.7が出てたのかー。

Python 2.7.7 をインストール。 :

Python 2.7.6 をアンインストールして、Python 2.7.7 をインストールした。

ふと気づいたら、GIMP 2.6.12 の Python-fu が動かなくなっていた。Python 2.6.6 をアンインストールしてインストールし直したり。

相変わらず Python-fu が動かない。何度かアンインストール、再インストールを繰り返したけど問題解決せず。

ふと、別のショートカットファイルから GIMP を起動したら、Python-fu が動くことに気付いた。ショートカットファイルの設定で、作業フォルダに %USERPROFILE% を指定している場合は、Python-fu が動くらしい。…なんでだろ。

#2 [windows] Windows7 x64 に curl をインストール

Python 関係のパッケージ?インストール解説記事中で curl というコマンドラインのダウンローダ?を使っていたので、試しに自分も導入。

_cURL - Download の下の方にある 「Win32 - Generic」のどれかを選択。

今回は、 _cURL groks URLs - CURL 7.37.0 から、curl_737_0_ssl.zip をDLした。

解凍して、curl.exe を、パスの通ったところにコピー。

このままだと、https でアクセスできないらしいので、以下を参考に作業。

_Why can't cURL properly verify a certificate on Windows? - Super User
_cURL - Extract CA Certs from Mozilla

cacert.pem をDLして、curl.exe と同じ階層に、curl-ca-bundle.crt というファイル名でコピー。

あるいは、--insecure をつけることでも、https でアクセスできる?

_WindowsR環境にcURLコマンドを実行できるようにインストールする手順 | ええかげんブログ(本店)

curl --insecure https://www.google.co.jp/
といった指定でもイケるらしい。

#3 [python] 画像の周辺の黒い部分をトリミングするPythonスクリプト

Python + PIL で、トリミング作業の実験。
from optparse import OptionParser
import os.path
import glob
try:
    import Image
except ImportError:
    from PIL import Image

ver = "0.0.1"

def chk_args():
    """コマンドラインオプションを解析."""

    parser = OptionParser(usage = "%prog [--trim] img\*.png",
        version = "%prog " + ver)
    parser.add_option("--trim", dest="trim",
        action="store_true", default=False, help="Trim exec")

    (opts, args) = parser.parse_args()
    if len(args) == 0:
       parser.print_help()
       exit
    return (opts, args)

def open_image(fn, enable_trim):
    """画像ファイルを開いて情報を調べる."""

    if not os.path.isfile(fn):
        print "Not Found %s" % fn
        return

    im = Image.open(fn).convert('RGB')
    w, h = im.size
    xmin, ymin, xmax, ymax = im.getbbox()
    xw = xmax - xmin
    yh = ymax - ymin
    print "DrawArea (%d, %d) - (%d, %d) : %d x %d : Img Size %d x %d : %s" % \
        (xmin, ymin, xmax, ymax, xw, yh, w, h, fn)

    if enable_trim:
        region = im.crop((xmin, ymin, xmax, ymax))
        dn = os.path.dirname(fn)
        bn = os.path.basename(fn)
        new_fn = os.path.join(dn, "trim_%s" % bn)
        region.save(new_fn)

def main():
    (opts, args) = chk_args()
    for fng in args:
        for fn in glob.glob(fng):
            open_image(fn, opts.trim)

if __name__ == '__main__':
    main()

getbbox() で、周辺の黒い部分を取り除いた領域の値を得られるらしい。

そんな便利なメソッドがあったとは知らなかったものだから、画像のドット数分、getpixel() を使って1ドットずつチマチマと全部チェックして、コレはさすがに処理が遅すぎるわ、高速化できんかなー、とか思ってました。getbbox() を使ったら一瞬で処理が終わってしまった…。

2014/06/17追記。 :


以上、1 日分です。

過去ログ表示

Prev - 2014/06 - Next
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30

カテゴリで表示

検索機能は Namazu for hns で提供されています。(詳細指定/ヘルプ


注意: 現在使用の日記自動生成システムは Version 2.19.6 です。
公開されている日記自動生成システムは Version 2.19.5 です。

Powered by hns-2.19.6, HyperNikkiSystem Project