Differences between revisions 3 and 32 (spanning 29 versions)
⇤ ← Revision 3 as of 2005-09-14 06:33:16 →
Size: 3758
Editor: limodou
Comment:
← Revision 32 as of 2007-06-26 09:28:27 → ⇥
Size: 10240
Editor: limodou
Comment:
Deletions are marked like this. Additions are marked like this.
Line 12: Line 12:
 * You must create section first, so you can access its options. So if you didn't create x['name']={} section, so you cann't do x['name']['o'] = 3
 * The option's data can be saved as string format, but as read out again, Config4Obj cann't convert it to their original value type, so you must conver it yourself		  * You must create section first, then you can access its options. So if you didn't create x['name']={} section, so you cann't do x['name']['o'] = 3
 * The option's data can be saved as string format, but as read out again, Config4Obj cann't convert it to their original value type, so you must conver it yourself. I didn't tried validate module ships with Config4Obj.
Line 18: Line 18:
So I decide to ''reinvent'' a new module to solve there lacks, I named it as Dict4Ini, it means you can see the object is just like a dict. So I decide to ''reinvent'' a new module to solve these lacks, I named it as Dict4Ini, it means you can access the config object just like a dict.
Line 27: Line 27:
 * It's a litter module, just for my mind, so if you like, you should try it, but if you don't like, just skip it, that's ok  * can save config items order
 * support multi data types: string, unicode, int, float, list/tuple, dict, etc, you can save them to or regain them from config file
 * can convert to dict '''new'''
 * auto detect BOM of utf-8 file '''new'''
 * It's a little module, just for my mind, so if you like, you could try it, but if you don't like, just skip it, that's ok

== Where can I download it? ==

 * Visit the http://wiki.woodpecker.org.cn/moin/Dict4Ini site
 * Download it from here attachment:dict4ini.py
 * Dict4Ini also ships with UliPad, you can also download UliPad, and get it from modules/dict4ini.py, that's the same.

== About license ==

 * before GPL
 * now BSD you can see it in the source code

== What's new? ==
 * 2007/06/26 Version 0.9.1
  * Fix float convert bug
 * 2007/06/13 Version 0.9
  * Thanks for Victor Stinner giving a output format patch, so you can define your own output format "%s = %s" to anything you want, for example "%s=%s" or "%s:%s". And Dict4Ini will auto remove '%s' from the fromat string, and the strip() the rest of the string, then use it to split the key and value in Ini file. For example, if you using "%s:%s", Dict4Ini will get "%s:%s".replace('%s', '').strip(), then use ':' to split the line.
 * 2007/04/20 Version 0.8
  * Add exception process when parsing file
  * Add BOM detect
 * 2007/04/19 Version 0.7
  * Fix '\' escape
 * 2006/03/21 Version 0.6
  * Fix ordereditems bug.
 * 2006/01/04 Version 0.5
  * 2006/01/04
   * Add ordereditems() method, so you can get the items according the ini file definition
  * 2005/12/30
   * Support onelevel parameter, you can set it True, than only one level can be used, so that the key can include section delimeter char in it.
   * Support sectiondelimeter parmeter, you can set the section delimeter to which you want
 * 2005/12/12 Version 0.4
  * Fixed a bug about "\" in option's value, thanks to Andreas Kaiser
 * 2005/12/09 Version 0.3
  * Adding dict() method, then you can change the DictIni object to dict type, so you can really use it as a dict alternative
 * 2005/10/16 Version 0.2
  * Saving config items order
  * Support float format
Line 31: Line 72:
=== Examples 1 Create a ini file === === Example 1 Create a ini file ===
Line 38: Line 79:
    x.common.list = [3, 'Hello', 'have spaces']     x.common.list = [3, 1.2, 'Hello', 'have spaces']
Line 44: Line 85:
list = 3,Hello,"have spaces", list = 3,1.2,Hello,"have spaces",
Line 50: Line 91:
In this time, the Dict4Ini support number, list, string, unicode data type, for others you should convert yourself. In this time, the Dict4Ini support int, float, list/tuple, string, unicode data type, for others you should convert yourself.
Line 65: Line 106:
[3, 'Hello', 'have spaces'] [3, 1.2, 'Hello', 'have spaces']
Line 68: Line 109:
So you also can see, the order of options don't be holding. In many cases, it's not very seriously.

=== Example 3 Default value ===

Many times, you may want to set default values of options, once it is not set in configuration file, using Dict4Ini, you have many ways to do that:		 The data is auto converted to its original type.

=== Example 3 Dealing default values ===

Many times, you may want to set default values of options, once it is not set in configuration file. Using Dict4Ini, you have many ways to do that:
Line 97: Line 138:
=== Example 4 Saving comments in ini file ===

{{{#!python
    import dict4ini
    x = dict4ini.DictIni('test.ini')

    x.common._comment = 'This is a comment test.\nThis is the second line.'
    x.common.name = 'limodou'
    x.common.comment('name', 'Input your name')
    x.common.bool = 1
    x.common.comment('bool', 'Boot type')
    x.common.list = ['3', 'Hello', 'have spaces']
    x.common.comment('list', 'list testing')
    x.save()}}}

You can save comments in configuration also. Adding comments to section, you should using x.section._comment = 'comments'. Comments could be multi lines. Or you could use more commonly method, x.comment(). Just like x.comment('common', 'comments'). Add comments to options, you can only using comment() method, just like x.common.comment('list', 'comments').

The result of the ini file will be:

{{{# This is a comment test.
# This is the second line.
[common]
# Boot type
bool = 1
# list testing
list = "3",Hello,"have spaces",
# Input your name
name = limodou}}}

=== Example 5 Using unicode in ini file ===
==== Using default encoding ====
{{{#!python
#coding=utf-8
import dict4ini
x = dict4ini.DictIni('test.ini')
x.common.name = u'中文名'
x.save()}}}

Note: You should specify the coding used in the .py file. In this case is utf-8. Then I assign x.common.name with a unicode string. If you don't specify the encoding in create instance of the DictIni, the Dict4Ini will auto find the default encoding in the system in turns of:

 * local.getdefaultlocale()[1]
 * sys.getfilesystemencoding()
 * utf-8

The result of the ini file will be:

{{{[common]
name = u"中文名"
}}}

You should notice the file encoding will be utf-8, and the name's value is like python unicode syntax. For easiness, it doesn't support using unicode in comments.

==== Specifying encoding name ====
You can also specify the encoding of ini file, just like:

{{{#!python
#coding=utf-8
import dict4ini
x = dict4ini.DictIni('test.ini', encoding='gbk')
x.common.name = u'中文名'
x.save()}}}

It's easy to set an encoding of ini file.

=== Example 6 Using multi section ===

{{{#!python
import dict4ini
x = dict4ini.DictIni('test.ini')
x.common.settings.a = 1
x.common.settings.b = ["3", "hello"]
x.special.name = 'limodou'
x.special.homepage = 'http://www.donews.net/limodou'
x.save()}}}

You don't need to care if subsection is created, you need to just use it.

The result of the ini file will be:

{{{ [common/settings]
    a = 1
    b = "3",hello,


    [special]
    homepage = http://www.donews.net/limodou
    name = limodou}}}

=== Example 7 Getting ordered options ===
Sometimes we need to keep the order or the options according to the ini file, so how to get the ordered items?

{{{#!python
ini = dict4ini.DictIni('x.ini')
for key, value in ini.ordereditems(ini):
   print key, value
}}}
This example is dealing with the first level section.

{{{#!python
ini = dict4ini.DictIni('x.ini')
for key, value in ini.ordereditems(ini.autostring):
   print key, value
}}}
This example is dealing with certain section.
Line 98: Line 244:

== FAQ ==

=== 1. Can I delete an option? ===

A: Yes. For example:
{{{#!python
import dict4ini
x = dict4ini.DictIni('test.ini')
del x.a
x.save()}}}

=== 2. How to use 'xxx.xxx' style option key? ===

A: Easy. Just using dict syntax, for example:

{{{#!python
x['common']['xxx.xxx'] = 'a'}}}

or

{{{#!python
x.common['xxx.xxx'] = 'a'}}}

=== 3. How to deal the key including section delimeter char, just like '/' ===

A: As you creating the DictIni instance, you can specify a "onelevel=True" parameter:

{{{#!python
x = dict4ini.DictIni('inifile', onelevel=True)
}}}

But it'll not support multi section again. So you can also defined another sectiondelimeter char different from '/', just like:

{{{#!python
x = dict4ini.DictIni('inifile', sectiondelimeter='@')
}}}

But every time you called dict4ini.DictIni() you may need including sectiondelimeter parameter.
Line 101: Line 286:
[[PageComment2]]
 * This is a neat piece of software, but why is it GPL licenced? -- cmkl.
 * Because I like GPL :) -- limodou
 * 重要需求，可以支持多行内容的读取？比如说:
{{{
[code]
c1 = """
import sys
print dir(sys)
"""
}}} 这样类似的代码内容？ -- ZoomQuiet

::-- limodou [DateTime(2005-09-14T06:31:19Z)] TableOfContents

1. Dict4Ini

This module is used to process ini format configuration file. It acts just like a dict, but you can also access it's sections and options with attribute syntax, just like x.test.

1.1. Why reinvent this module?

I used Config4Obj module for GoogleTalkBot software (confbot) to deal with configuration file. But I found its lacks on:

  • Only can access options as x['name']['o'], but not as x.name.o
  • You must create section first, then you can access its options. So if you didn't create x['name']={} section, so you cann't do x['name']['o'] = 3

  • The option's data can be saved as string format, but as read out again, Config4Obj cann't convert it to their original value type, so you must conver it yourself. I didn't tried validate module ships with Config4Obj.

  • Didn't support unicode

Above is only my opinions, so they may be not right.

So I decide to reinvent a new module to solve these lacks, I named it as Dict4Ini, it means you can access the config object just like a dict.

1.2. What's it features

  • as simple as others
  • you can access options according to dict syntax, just like x['name']['o'] = 1, x['name'].keys(), x['name'].values(), etc.
  • you also can access options according to attr syntax, just like x.name.o = 1, x.name.keys(), x.name.values(), etc. So the name must be Identifier or single word.
  • you can save comments in it(but this feature is not tested so much)
  • support multi level section, subsection name will just like: [firsub/secsub]
  • can save config items order
  • support multi data types: string, unicode, int, float, list/tuple, dict, etc, you can save them to or regain them from config file

  • can convert to dict new

  • auto detect BOM of utf-8 file new

  • It's a little module, just for my mind, so if you like, you could try it, but if you don't like, just skip it, that's ok

1.3. Where can I download it?

1.4. About license

  • before GPL
  • now BSD you can see it in the source code

1.5. What's new?

  • 2007/06/26 Version 0.9.1
    • Fix float convert bug
  • 2007/06/13 Version 0.9

    • Thanks for Victor Stinner giving a output format patch, so you can define your own output format "%s = %s" to anything you want, for example "%s=%s" or "%s:%s". And Dict4Ini will auto remove '%s' from the fromat string, and the strip() the rest of the string, then use it to split the key and value in Ini file. For example, if you using "%s:%s", Dict4Ini will get "%s:%s".replace('%s', ).strip(), then use ':' to split the line.

  • 2007/04/20 Version 0.8
    • Add exception process when parsing file
    • Add BOM detect
  • 2007/04/19 Version 0.7
    • Fix '\' escape
  • 2006/03/21 Version 0.6
    • Fix ordereditems bug.
  • 2006/01/04 Version 0.5
    • 2006/01/04
      • Add ordereditems() method, so you can get the items according the ini file definition
    • 2005/12/30
      • Support onelevel parameter, you can set it True, than only one level can be used, so that the key can include section delimeter char in it.
      • Support sectiondelimeter parmeter, you can set the section delimeter to which you want
  • 2005/12/12 Version 0.4
    • Fixed a bug about "\" in option's value, thanks to Andreas Kaiser
  • 2005/12/09 Version 0.3

    • Adding dict() method, then you can change the DictIni object to dict type, so you can really use it as a dict alternative

  • 2005/10/16 Version 0.2
    • Saving config items order
    • Support float format

1.6. Examples

1.6.1. Example 1 Create a ini file

   1     import dict4ini
   2 
   3     x = dict4ini.DictIni('test.ini')
   4     x.common.name = 'limodou'
   5     x.common.bool = 1
   6     x.common.list = [3, 1.2, 'Hello', 'have spaces']
   7     x.save()

This example will save option to test.ini. As you can see, you needn't create section "common" at first, just think it's there, it's ok. The result of test.ini is:

{{{[common] list = 3,1.2,Hello,"have spaces", bool = 1 name = limodou}}}

And you can see, once the value has special chars, just like ' ', ',', '\"', etc, the string will be quoted by double quoter. But "Hello" is a single word, and it has not the special chars, so it won't be quoted. If the value is number, it'll be just like number literal, but if the value is number string, it'll be quoted by double quoter.

In this time, the Dict4Ini support int, float, list/tuple, string, unicode data type, for others you should convert yourself.

1.6.2. Example 2 Open an existed ini file

   1     import dict4ini
   2 
   3     x = dict4ini.DictIni('test.ini')
   4     print x.common.bool
   5     print x.common.list
   6     print x.common.name

So it's easy. The result will be:

Dict4Ini (last edited 2009-12-25 07:13:47 by localhost)