/[gentoo]/xml/htdocs/proj/en/glep/glep-0047.html
Gentoo

Diff of /xml/htdocs/proj/en/glep/glep-0047.html

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

Revision 1.3 Revision 1.9
1<?xml version="1.0" encoding="utf-8" ?> 1<?xml version="1.0" encoding="utf-8" ?>
2<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 2<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
3<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"> 3<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
4<!-- 4
5This HTML is auto-generated. DO NOT EDIT THIS FILE! If you are writing a new
6PEP, see http://www.python.org/peps/pep-0001.html for instructions and links
7to templates. DO NOT USE THIS HTML FILE AS YOUR TEMPLATE!
8-->
9<head> 5<head>
10 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> 6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
11 <meta name="generator" content="Docutils 0.3.5: http://docutils.sourceforge.net/" /> 7 <meta name="generator" content="Docutils 0.4: http://docutils.sourceforge.net/" />
12 <title>GLEP 47 -- Creating 'safe' environment variables</title> 8 <title>GLEP 47 -- Creating 'safe' environment variables</title>
13 <link rel="stylesheet" href="tools/glep.css" type="text/css" /> 9 <link rel="stylesheet" href="tools/glep.css" type="text/css" />
14</head> 10</head>
15<body bgcolor="white"> 11<body bgcolor="white">
16<table class="navigation" cellpadding="0" cellspacing="0" 12<table class="navigation" cellpadding="0" cellspacing="0"
… …
20<img src="http://www.gentoo.org/images/gentoo-new.gif" alt="[Gentoo]" 16<img src="http://www.gentoo.org/images/gentoo-new.gif" alt="[Gentoo]"
21 border="0" width="150" height="35" /></a></td> 17 border="0" width="150" height="35" /></a></td>
22<td class="textlinks" align="left"> 18<td class="textlinks" align="left">
23[<b><a href="http://www.gentoo.org/">Gentoo Linux Home</a></b>] 19[<b><a href="http://www.gentoo.org/">Gentoo Linux Home</a></b>]
24[<b><a href="http://www.gentoo.org/proj/en/glep">GLEP Index</a></b>] 20[<b><a href="http://www.gentoo.org/proj/en/glep">GLEP Index</a></b>]
25[<b><a href="./glep-0047.txt">GLEP Source</a></b>] 21[<b><a href="http://www.gentoo.org/proj/en/glep/glep-0047.txt">GLEP Source</a></b>]
26</td></tr></table> 22</td></tr></table>
27<div class="document">
28<table class="rfc2822 field-list" frame="void" rules="none"> 23<table class="rfc2822 docutils field-list" frame="void" rules="none">
29<col class="field-name" /> 24<col class="field-name" />
30<col class="field-body" /> 25<col class="field-body" />
31<tbody valign="top"> 26<tbody valign="top">
32<tr class="field"><th class="field-name">GLEP:</th><td class="field-body">47</td> 27<tr class="field"><th class="field-name">GLEP:</th><td class="field-body">47</td>
33</tr> 28</tr>
34<tr class="field"><th class="field-name">Title:</th><td class="field-body">Creating 'safe' environment variables</td> 29<tr class="field"><th class="field-name">Title:</th><td class="field-body">Creating 'safe' environment variables</td>
35</tr> 30</tr>
36<tr class="field"><th class="field-name">Version:</th><td class="field-body">1.2</td> 31<tr class="field"><th class="field-name">Version:</th><td class="field-body">1.6</td>
37</tr> 32</tr>
38<tr class="field"><th class="field-name">Last-Modified:</th><td class="field-body"><a class="reference" href="http://www.gentoo.org/cgi-bin/viewcvs/xml/htdocs/proj/en/glep/glep-0047.txt?cvsroot=gentoo">2006/02/11 21:43:14</a></td> 33<tr class="field"><th class="field-name">Last-Modified:</th><td class="field-body"><a class="reference" href="http://www.gentoo.org/cgi-bin/viewcvs.cgi/xml/htdocs/proj/en/glep/glep-0047.txt?cvsroot=gentoo">2007/07/22 10:03:18</a></td>
39</tr> 34</tr>
40<tr class="field"><th class="field-name">Author:</th><td class="field-body">Diego Pettenò, Fabian Groffen</td> 35<tr class="field"><th class="field-name">Author:</th><td class="field-body">Diego Pettenò, Fabian Groffen</td>
41</tr> 36</tr>
42<tr class="field"><th class="field-name">Status:</th><td class="field-body">Draft</td> 37<tr class="field"><th class="field-name">Status:</th><td class="field-body">Rejected</td>
43</tr> 38</tr>
44<tr class="field"><th class="field-name">Type:</th><td class="field-body">Standards Track</td> 39<tr class="field"><th class="field-name">Type:</th><td class="field-body">Standards Track</td>
45</tr> 40</tr>
46<tr class="field"><th class="field-name">Content-Type:</th><td class="field-body"><a class="reference" href="glep-0012.html">text/x-rst</a></td> 41<tr class="field"><th class="field-name">Content-Type:</th><td class="field-body"><a class="reference" href="glep-0002.html">text/x-rst</a></td>
47</tr> 42</tr>
48<tr class="field"><th class="field-name">Created:</th><td class="field-body">14-Oct-2005</td> 43<tr class="field"><th class="field-name">Created:</th><td class="field-body">14-Oct-2005</td>
49</tr> 44</tr>
50<tr class="field"><th class="field-name">Post-History:</th><td class="field-body">09-Feb-2006</td> 45<tr class="field"><th class="field-name">Post-History:</th><td class="field-body">09-Feb-2006</td>
51</tr> 46</tr>
52</tbody> 47</tbody>
53</table> 48</table>
54<hr /> 49<hr />
55<div class="contents topic" id="contents"> 50<div class="contents topic">
56<p class="topic-title first"><a name="contents">Contents</a></p> 51<p class="topic-title first"><a id="contents" name="contents">Contents</a></p>
57<ul class="simple"> 52<ul class="simple">
58<li><a class="reference" href="#credits" id="id5" name="id5">Credits</a></li> 53<li><a class="reference" href="#credits" id="id5" name="id5">Credits</a></li>
59<li><a class="reference" href="#abstract" id="id6" name="id6">Abstract</a></li> 54<li><a class="reference" href="#abstract" id="id6" name="id6">Abstract</a></li>
60<li><a class="reference" href="#motivation" id="id7" name="id7">Motivation</a></li> 55<li><a class="reference" href="#motivation" id="id7" name="id7">Motivation</a></li>
61<li><a class="reference" href="#rationale" id="id8" name="id8">Rationale</a></li> 56<li><a class="reference" href="#rationale" id="id8" name="id8">Rationale</a></li>
… …
66</li> 61</li>
67<li><a class="reference" href="#references" id="id12" name="id12">References</a></li> 62<li><a class="reference" href="#references" id="id12" name="id12">References</a></li>
68<li><a class="reference" href="#copyright" id="id13" name="id13">Copyright</a></li> 63<li><a class="reference" href="#copyright" id="id13" name="id13">Copyright</a></li>
69</ul> 64</ul>
70</div> 65</div>
71<div class="section" id="credits"> 66<div class="section">
72<h1><a class="toc-backref" href="#id5" name="credits">Credits</a></h1> 67<h1><a class="toc-backref" href="#id5" id="credits" name="credits">Credits</a></h1>
73<p>The text of this GLEP is a result of a discussion and input of the 68<p>The text of this GLEP is a result of a discussion and input of the
74following persons, in no particular order: Mike Frysinger, Diego 69following persons, in no particular order: Mike Frysinger, Diego
75Pettenò, Fabian Groffen and Finn Thain.</p> 70Pettenò, Fabian Groffen and Finn Thain.</p>
76</div> 71</div>
77<div class="section" id="abstract"> 72<div class="section">
78<h1><a class="toc-backref" href="#id6" name="abstract">Abstract</a></h1> 73<h1><a class="toc-backref" href="#id6" id="abstract" name="abstract">Abstract</a></h1>
79<p>In order for ebuilds and eclasses to be able to make host specific 74<p>In order for ebuilds and eclasses to be able to make host specific
80decisions, it is necessary to have a number of environmental variables 75decisions, it is necessary to have a number of environmental variables
81which allow for such decisions. This GLEP introduces some measures that 76which allow for such decisions. This GLEP introduces some measures that
82need to be made to make these decisions 'safe', by making sure the 77need to be made to make these decisions 'safe', by making sure the
83variables the decisions are based on are 'safe'. A small overlap with 78variables the decisions are based on are 'safe'. A small overlap with
84GLEP 22 <a class="footnote-reference" href="#id3" id="id1" name="id1">[1]</a> is being handled in this GLEP where the use of 2-tuple 79GLEP 22 <a class="footnote-reference" href="#id3" id="id1" name="id1">[1]</a> is being handled in this GLEP where the use of 2-tuple
85keywords are being kept instead of 4-tuple keywords. Additionally, the 80keywords are being kept instead of 4-tuple keywords. Additionally, the
86<tt class="literal"><span class="pre">ELIBC</span></tt>, <tt class="literal"><span class="pre">KERNEL</span></tt> and <tt class="literal"><span class="pre">ARCH</span></tt> get auto filled starting from 81<tt class="docutils literal"><span class="pre">ELIBC</span></tt>, <tt class="docutils literal"><span class="pre">KERNEL</span></tt> and <tt class="docutils literal"><span class="pre">ARCH</span></tt> get auto filled starting from
87<tt class="literal"><span class="pre">CHOST</span></tt> and the 2-tuple keyword, instead of solely from they 4-tuple 82<tt class="docutils literal"><span class="pre">CHOST</span></tt> and the 2-tuple keyword, instead of solely from they 4-tuple
88keyword as proposed in GLEP 22.</p> 83keyword as proposed in GLEP 22.</p>
89<p>The destiny of the <tt class="literal"><span class="pre">USERLAND</span></tt> variable is out of the scope of this 84<p>The destiny of the <tt class="docutils literal"><span class="pre">USERLAND</span></tt> variable is out of the scope of this
90GLEP. Depending on its presence in the tree, it may be decided to set 85GLEP. Depending on its presence in the tree, it may be decided to set
91this variable the same way we propose to set <tt class="literal"><span class="pre">ELIBC</span></tt>, <tt class="literal"><span class="pre">KERNEL</span></tt> and 86this variable the same way we propose to set <tt class="docutils literal"><span class="pre">ELIBC</span></tt>, <tt class="docutils literal"><span class="pre">KERNEL</span></tt> and
92<tt class="literal"><span class="pre">ARCH</span></tt>, or alternatively, e.g. via the profiles.</p> 87<tt class="docutils literal"><span class="pre">ARCH</span></tt>, or alternatively, e.g. via the profiles.</p>
93</div> 88</div>
94<div class="section" id="motivation"> 89<div class="section">
95<h1><a class="toc-backref" href="#id7" name="motivation">Motivation</a></h1> 90<h1><a class="toc-backref" href="#id7" id="motivation" name="motivation">Motivation</a></h1>
96<p>The Gentoo/Alt project is in an emerging state to get ready to serve a 91<p>The Gentoo/Alt project is in an emerging state to get ready to serve a
97plethora of 'alternative' configurations such as FreeBSD, NetBSD, 92plethora of 'alternative' configurations such as FreeBSD, NetBSD,
98DragonflyBSD, GNU/kFreeBSD, Mac OS X, (Open)Darwin, (Open)Solaris and so 93DragonflyBSD, GNU/kFreeBSD, Mac OS X, (Open)Darwin, (Open)Solaris and so
99on. As such, the project is in need for a better grip on the actual 94on. As such, the project is in need for a better grip on the actual
100host being built on. This information on the host environment is 95host being built on. This information on the host environment is
101necessary to make proper (automated) decisions on settings that are 96necessary to make proper (automated) decisions on settings that are
102highly dependant on the build environment, such as platform or C-library 97highly dependant on the build environment, such as platform or C-library
103implementation.</p> 98implementation.</p>
104</div> 99</div>
105<div class="section" id="rationale"> 100<div class="section">
106<h1><a class="toc-backref" href="#id8" name="rationale">Rationale</a></h1> 101<h1><a class="toc-backref" href="#id8" id="rationale" name="rationale">Rationale</a></h1>
107<p>Gentoo's unique Portage system allows easy installation of applications 102<p>Gentoo's unique Portage system allows easy installation of applications
108from source packages. Compiling sources is prone to many environmental 103from source packages. Compiling sources is prone to many environmental
109settings and availability of certain tools. Only recently the Gentoo 104settings and availability of certain tools. Only recently the Gentoo
110for FreeBSD project has started, as second Gentoo project that operates 105for FreeBSD project has started, as second Gentoo project that operates
111on a foreign host operating system using foreign (non-GNU) C-libraries 106on a foreign host operating system using foreign (non-GNU) C-libraries
… …
116environmental differences, information regarding it should be supplied. 111environmental differences, information regarding it should be supplied.
117Since decisions based on this information can be vital, it is of high 112Since decisions based on this information can be vital, it is of high
118importance that this information can be trusted and the values can be 113importance that this information can be trusted and the values can be
119considered 'safe' and correct.</p> 114considered 'safe' and correct.</p>
120</div> 115</div>
121<div class="section" id="backwards-compatibility"> 116<div class="section">
122<h1><a class="toc-backref" href="#id9" name="backwards-compatibility">Backwards Compatibility</a></h1> 117<h1><a class="toc-backref" href="#id9" id="backwards-compatibility" name="backwards-compatibility">Backwards Compatibility</a></h1>
123<p>The proposed keywording scheme in this GLEP is fully compatible with the 118<p>The proposed keywording scheme in this GLEP is fully compatible with the
124current situation of the portage tree, this in contrast to GLEP 22. The 119current situation of the portage tree, this in contrast to GLEP 22. The
125variables provided by GLEP 22 can't be extracted from the new keyword, 120variables provided by GLEP 22 can't be extracted from the new keyword,
126but since GLEP 22-style keywords aren't in the tree at the moment, that 121but since GLEP 22-style keywords aren't in the tree at the moment, that
127is not a problem. The same information can be extracted from the CHOST 122is not a problem. The same information can be extracted from the CHOST
128variable, if necessary. No modifications to ebuilds will have to be 123variable, if necessary. No modifications to ebuilds will have to be
129made.</p> 124made.</p>
130</div> 125</div>
131<div class="section" id="specification"> 126<div class="section">
132<h1><a class="toc-backref" href="#id10" name="specification">Specification</a></h1> 127<h1><a class="toc-backref" href="#id10" id="specification" name="specification">Specification</a></h1>
133<p>Unlike GLEP 22 the currently used keyword scheme is not changed. 128<p>Unlike GLEP 22 the currently used keyword scheme is not changed.
134Instead of proposing a 4-tuple <a class="footnote-reference" href="#id4" id="id2" name="id2">[2]</a> keyword, a 2-tuple keyword is chosen 129Instead of proposing a 4-tuple <a class="footnote-reference" href="#id4" id="id2" name="id2">[2]</a> keyword, a 2-tuple keyword is chosen
135for archs that require them. Archs for which a 1-tuple keyword 130for archs that require them. Archs for which a 1-tuple keyword
136suffices, can keep that keyword. Since this doesn't change anything to 131suffices, can keep that keyword. Since this doesn't change anything to
137the current situation in the tree, it is considered to be a big 132the current situation in the tree, it is considered to be a big
138advantage over the 4-tuple keyword from GLEP 22. This GLEP is an 133advantage over the 4-tuple keyword from GLEP 22. This GLEP is an
139official specification of the syntax of the keyword.</p> 134official specification of the syntax of the keyword.</p>
140<p>Keywords will consist out of two parts separated by a hyphen ('-'). The 135<p>Keywords will consist out of two parts separated by a hyphen ('-'). The
141part up to the first hyphen from the left of the keyword 2-tuple is the 136part up to the first hyphen from the left of the keyword 2-tuple is the
142architecture, such as ppc64, sparc and x86. Allowed characters for the 137architecture, such as ppc64, sparc and x86. Allowed characters for the
143architecture name are in <tt class="literal"><span class="pre">a-z0-9</span></tt>. The remaining part on the right of 138architecture name are in <tt class="docutils literal"><span class="pre">a-z0-9</span></tt>. The remaining part on the right of
144the first hyphen from the left indicates the operating system or 139the first hyphen from the left indicates the operating system or
145distribution, such as linux, macos, darwin, obsd, et-cetera. If the 140distribution, such as linux, macos, darwin, obsd, et-cetera. If the
146right hand part is omitted, it implies the operating system/distribution 141right hand part is omitted, it implies the operating system/distribution
147type is Gentoo GNU/Linux. In such case the hyphen is also omitted, and 142type is Gentoo GNU/Linux. In such case the hyphen is also omitted, and
148the keyword consists of solely the architecture. The operating system 143the keyword consists of solely the architecture. The operating system
149or distribution name can consist out of characters in <tt class="literal"><span class="pre">a-zA-Z0-9_+:-</span></tt>. 144or distribution name can consist out of characters in <tt class="docutils literal"><span class="pre">a-zA-Z0-9_+:-</span></tt>.
150Please note that the hyphen is an allowed character, and therefore the 145Please note that the hyphen is an allowed character, and therefore the
151separation of the two fields in the keyword is only determinable by 146separation of the two fields in the keyword is only determinable by
152scanning for the first hyphen character from the start of the keyword 147scanning for the first hyphen character from the start of the keyword
153string. Examples of keywords following this specification are 148string. Examples of keywords following this specification are
154ppc-darwin and x86. This is fully compatible with the current use of 149ppc-darwin and x86. This is fully compatible with the current use of
155keywords in the tree.</p> 150keywords in the tree.</p>
156<p>The variables <tt class="literal"><span class="pre">ELIBC</span></tt>, <tt class="literal"><span class="pre">KERNEL</span></tt> and <tt class="literal"><span class="pre">ARCH</span></tt> are currently set in 151<p>The variables <tt class="docutils literal"><span class="pre">ELIBC</span></tt>, <tt class="docutils literal"><span class="pre">KERNEL</span></tt> and <tt class="docutils literal"><span class="pre">ARCH</span></tt> are currently set in
157the profiles when other than their defaults for a GNU/Linux system. 152the profiles when other than their defaults for a GNU/Linux system.
158They can as such easily be overridden and defined by the user. To 153They can as such easily be overridden and defined by the user. To
159prevent this from happening, the variables should be auto filled by 154prevent this from happening, the variables should be auto filled by
160Portage itself, based on the <tt class="literal"><span class="pre">CHOST</span></tt> variable. While the <tt class="literal"><span class="pre">CHOST</span></tt> 155Portage itself, based on the <tt class="docutils literal"><span class="pre">CHOST</span></tt> variable. While the <tt class="docutils literal"><span class="pre">CHOST</span></tt>
161variable can be as easy as the others set by the user, it still is 156variable can be as easy as the others set by the user, it still is
162assumed to be 'safe'. This assumption is grounded in the fact that the 157assumed to be 'safe'. This assumption is grounded in the fact that the
163variable itself is being used in various other places with the same 158variable itself is being used in various other places with the same
164intention, and that an invalid <tt class="literal"><span class="pre">CHOST</span></tt> will cause major malfunctioning 159intention, and that an invalid <tt class="docutils literal"><span class="pre">CHOST</span></tt> will cause major malfunctioning
165of the system. A user that changes the <tt class="literal"><span class="pre">CHOST</span></tt> into something that is 160of the system. A user that changes the <tt class="docutils literal"><span class="pre">CHOST</span></tt> into something that is
166not valid for the system, is already warned that this might render the 161not valid for the system, is already warned that this might render the
167system unusable. Concluding, the 'safeness' of the <tt class="literal"><span class="pre">CHOST</span></tt> variable 162system unusable. Concluding, the 'safeness' of the <tt class="docutils literal"><span class="pre">CHOST</span></tt> variable
168is based on externally assumed 'safeness', which's discussion falls 163is based on externally assumed 'safeness', which's discussion falls
169outside this GLEP.</p> 164outside this GLEP.</p>
170<p>Current USE-expansion of the variables is being maintained, as this 165<p>Current USE-expansion of the variables is being maintained, as this
171results in full backward compatibility. Since the variables themself 166results in full backward compatibility. Since the variables themself
172don't change in what they represent, but only how they are being 167don't change in what they represent, but only how they are being
… …
181 ... 176 ...
182 use elibc_FreeBSD &amp;&amp; myconf=&quot;${myconf} -Dlibc=/usr/lib/libc.a&quot; 177 use elibc_FreeBSD &amp;&amp; myconf=&quot;${myconf} -Dlibc=/usr/lib/libc.a&quot;
183 ... 178 ...
184} 179}
185</pre> 180</pre>
186<p>Alternatively, the variables <tt class="literal"><span class="pre">ELIBC</span></tt>, <tt class="literal"><span class="pre">KERNEL</span></tt> and <tt class="literal"><span class="pre">ARCH</span></tt> 181<p>Alternatively, the variables <tt class="docutils literal"><span class="pre">ELIBC</span></tt>, <tt class="docutils literal"><span class="pre">KERNEL</span></tt> and <tt class="docutils literal"><span class="pre">ARCH</span></tt>
187are available in the ebuild evironment and they can be used instead of 182are available in the ebuild evironment and they can be used instead of
188invoking <tt class="literal"><span class="pre">xxx_Xxxx</span></tt> or in switch statements where they are actually 183invoking <tt class="docutils literal"><span class="pre">xxx_Xxxx</span></tt> or in switch statements where they are actually
189necessary.</p> 184necessary.</p>
190<p>A map file can be used to have the various <tt class="literal"><span class="pre">CHOST</span></tt> values being 185<p>A map file can be used to have the various <tt class="docutils literal"><span class="pre">CHOST</span></tt> values being
191translated to the correct values for the four variables. This change is 186translated to the correct values for the four variables. This change is
192invisible for ebuilds and eclasses, but allows to rely on these 187invisible for ebuilds and eclasses, but allows to rely on these
193variables as they are based on a 'safe' value -- the <tt class="literal"><span class="pre">CHOST</span></tt> variable. 188variables as they are based on a 'safe' value -- the <tt class="docutils literal"><span class="pre">CHOST</span></tt> variable.
194Ebuilds should not be sensitive to the keyword value, but use the 189Ebuilds should not be sensitive to the keyword value, but use the
195aforementioned four variables instead. They allow specific tests for 190aforementioned four variables instead. They allow specific tests for
196properties. If this is undesirable, the full <tt class="literal"><span class="pre">CHOST</span></tt> variable can be 191properties. If this is undesirable, the full <tt class="docutils literal"><span class="pre">CHOST</span></tt> variable can be
197used to match a complete operating system.</p> 192used to match a complete operating system.</p>
198<div class="section" id="variable-assignment"> 193<div class="section">
199<h2><a class="toc-backref" href="#id11" name="variable-assignment">Variable Assignment</a></h2> 194<h2><a class="toc-backref" href="#id11" id="variable-assignment" name="variable-assignment">Variable Assignment</a></h2>
200<p>The <tt class="literal"><span class="pre">ELIBC</span></tt>, <tt class="literal"><span class="pre">KERNEL</span></tt>, <tt class="literal"><span class="pre">ARCH</span></tt> variables are filled from a profile 195<p>The <tt class="docutils literal"><span class="pre">ELIBC</span></tt>, <tt class="docutils literal"><span class="pre">KERNEL</span></tt>, <tt class="docutils literal"><span class="pre">ARCH</span></tt> variables are filled from a profile
201file. The file can be overlaid, such that the following entries in the 196file. The file can be overlaid, such that the following entries in the
202map file (on the left of the arrow) will result in the assigned 197map file (on the left of the arrow) will result in the assigned
203variables on the right hand side of the arrow:</p> 198variables on the right hand side of the arrow:</p>
204<pre class="literal-block"> 199<pre class="literal-block">
205*-*-linux-* -&gt; KERNEL=&quot;linux&quot; 200*-*-linux-* -&gt; KERNEL=&quot;linux&quot;
206*-*-*-gnu -&gt; ELIBC=&quot;glibc&quot; 201*-*-*-gnu -&gt; ELIBC=&quot;glibc&quot;
207*-*-kfreebsd-gnu -&gt; KERNEL=&quot;FreeBSD&quot; ELIBC=&quot;glibc&quot; 202*-*-kfreebsd-gnu -&gt; KERNEL=&quot;FreeBSD&quot; ELIBC=&quot;glibc&quot;
208*-*-freebsd* -&gt; KERNEL=&quot;FreeBSD&quot; ELIBC=&quot;FreeBSD&quot; 203*-*-freebsd* -&gt; KERNEL=&quot;FreeBSD&quot; ELIBC=&quot;FreeBSD&quot;
209*-*-darwin* -&gt; KERNEL=&quot;Darwin&quot; ELIBC=&quot;Darwin&quot; 204*-*-darwin* -&gt; KERNEL=&quot;Darwin&quot; ELIBC=&quot;Darwin&quot;
210*-*-netbsd* -&gt; KERNEL=&quot;NetBSD&quot; ELIBC=&quot;NetBSD&quot; 205*-*-netbsd* -&gt; KERNEL=&quot;NetBSD&quot; ELIBC=&quot;NetBSD&quot;
211*-*-solaris* -&gt; KERNEL=&quot;Solaris&quot; ELIBC=&quot;Solaris&quot; 206*-*-solaris* -&gt; KERNEL=&quot;Solaris&quot; ELIBC=&quot;Solaris&quot;
212</pre> 207</pre>
213<p>A way to achieve this is proposed by Mike Frysinger, which 208<p>A way to achieve this is proposed by Mike Frysinger, which
214suggests to have a env-map file, for instance filled with:</p> 209suggests to have an env-map file, for instance filled with:</p>
215<pre class="literal-block"> 210<pre class="literal-block">
216% cat env-map 211% cat env-map
217*-linux-* KERNEL=linux 212*-linux-* KERNEL=linux
218*-gnu ELIBC=glibc 213*-gnu ELIBC=glibc
219x86_64-* ARCH=amd64 214x86_64-* ARCH=amd64
220</pre> 215</pre>
221<p>then the following bash script can be used to set the four variables to 216<p>then the following bash script can be used to set the four variables to
222their correct values:</p> 217their correct values:</p>
223<pre class="literal-block"> 218<pre class="literal-block">
224% cat readmap 219% cat readmap
225#!/bin/bash 220#!/bin/bash
226 221
227CBUILD=${CBUILD:-${CHOST=${CHOST:-$1}}} 222CBUILD=${CBUILD:-${CHOST=${CHOST:-$1}}}
228[[ -z ${CHOST} ]] &amp;&amp; echo need chost 223[[ -z ${CHOST} ]] &amp;&amp; echo need chost
229 224
230unset KERNEL ELIBC ARCH 225unset KERNEL ELIBC ARCH
231 226
232while read LINE ; do 227while read LINE ; do
233 set -- ${LINE} 228 set -- ${LINE}
234 targ=$1 229 targ=$1
235 shift 230 shift
236 [[ ${CBUILD} == ${targ} ]] &amp;&amp; eval $&#64; 231 [[ ${CBUILD} == ${targ} ]] &amp;&amp; eval $&#64;
237done &lt; env-map 232done &lt; env-map
238 233
239echo ARCH=${ARCH} KERNEL=${KERNEL} ELIBC=${ELIBC} 234echo ARCH=${ARCH} KERNEL=${KERNEL} ELIBC=${ELIBC}
240</pre> 235</pre>
241<p>Given the example env-map file, this script would result in:</p> 236<p>Given the example env-map file, this script would result in:</p>
242<pre class="literal-block"> 237<pre class="literal-block">
243% ./readmap x86_64-pc-linux-gnu 238% ./readmap x86_64-pc-linux-gnu
244ARCH=amd64 KERNEL=linux ELIBC=glibc 239ARCH=amd64 KERNEL=linux ELIBC=glibc
245</pre> 240</pre>
246<p>The entries in the <tt class="literal"><span class="pre">env-map</span></tt> file will be evaluated in a forward 241<p>The entries in the <tt class="docutils literal"><span class="pre">env-map</span></tt> file will be evaluated in a forward
247linear full scan. A side-effect of this exhaustive search is that the 242linear full scan. A side-effect of this exhaustive search is that the
248variables can be re-assigned if multiple entries match the given 243variables can be re-assigned if multiple entries match the given
249<tt class="literal"><span class="pre">CHOST</span></tt>. Because of this, the order of the entries does matter. 244<tt class="docutils literal"><span class="pre">CHOST</span></tt>. Because of this, the order of the entries does matter.
250Because the <tt class="literal"><span class="pre">env-map</span></tt> file size is assumed not to exceed the block 245Because the <tt class="docutils literal"><span class="pre">env-map</span></tt> file size is assumed not to exceed the block
251size of the file system, the performance penalty of a full scan versus 246size of the file system, the performance penalty of a full scan versus
252'first-hit-stop technique' is assumed to be minimal.</p> 247'first-hit-stop technique' is assumed to be minimal.</p>
253<p>It should be noted, however, that the above bash script is a proof of 248<p>It should be noted, however, that the above bash script is a proof of
254concept implementation. Since Portage is largerly written in Python, it 249concept implementation. Since Portage is largerly written in Python, it
255will be more efficient to write an equivalent of this code in Python 250will be more efficient to write an equivalent of this code in Python
256also. Coding wise, this is considered to be a non-issue, but the format 251also. Coding wise, this is considered to be a non-issue, but the format
257of the <tt class="literal"><span class="pre">env-map</span></tt> file, and especially its wildcard characters, might 252of the <tt class="docutils literal"><span class="pre">env-map</span></tt> file, and especially its wildcard characters, might
258not be the best match with Python. For this purpose, the format 253not be the best match with Python. For this purpose, the format
259specification of the <tt class="literal"><span class="pre">env-map</span></tt> file is deferred to the Python 254specification of the <tt class="docutils literal"><span class="pre">env-map</span></tt> file is deferred to the Python
260implementation, and only the requirements are given here.</p> 255implementation, and only the requirements are given here.</p>
261<p>The <tt class="literal"><span class="pre">env-map</span></tt> file should be capable of encoding a <tt class="literal"><span class="pre">key</span></tt>, <tt class="literal"><span class="pre">value</span></tt> 256<p>The <tt class="docutils literal"><span class="pre">env-map</span></tt> file should be capable of encoding a <tt class="docutils literal"><span class="pre">key</span></tt>, <tt class="docutils literal"><span class="pre">value</span></tt>
262pair, where <tt class="literal"><span class="pre">key</span></tt> is a (regular) expression that matches a 257pair, where <tt class="docutils literal"><span class="pre">key</span></tt> is a (regular) expression that matches a
263chost-string, and <tt class="literal"><span class="pre">value</span></tt> contains at least one, distinct variable 258chost-string, and <tt class="docutils literal"><span class="pre">value</span></tt> contains at least one, distinct variable
264assignment for the variables <tt class="literal"><span class="pre">ARCH</span></tt>, <tt class="literal"><span class="pre">KERNEL</span></tt> and <tt class="literal"><span class="pre">ELIBC</span></tt>. The 259assignment for the variables <tt class="docutils literal"><span class="pre">ARCH</span></tt>, <tt class="docutils literal"><span class="pre">KERNEL</span></tt> and <tt class="docutils literal"><span class="pre">ELIBC</span></tt>. The
265interpreter of the <tt class="literal"><span class="pre">env-map</span></tt> file must scan the file linearly and 260interpreter of the <tt class="docutils literal"><span class="pre">env-map</span></tt> file must scan the file linearly and
266continue trying to match the <tt class="literal"><span class="pre">key</span></tt>s and assign variables if 261continue trying to match the <tt class="docutils literal"><span class="pre">key</span></tt>s and assign variables if
267appropriate until the end of file.</p> 262appropriate until the end of file.</p>
268<p>Since Portage will use the <tt class="literal"><span class="pre">env-map</span></tt> file, the location of the file is 263<p>Since Portage will use the <tt class="docutils literal"><span class="pre">env-map</span></tt> file, the location of the file is
269beyond the scope of this GLEP and up to the Portage implementors.</p> 264beyond the scope of this GLEP and up to the Portage implementors.</p>
270</div> 265</div>
271</div> 266</div>
272<div class="section" id="references"> 267<div class="section">
273<h1><a class="toc-backref" href="#id12" name="references">References</a></h1> 268<h1><a class="toc-backref" href="#id12" id="references" name="references">References</a></h1>
274<table class="footnote" frame="void" id="id3" rules="none"> 269<table class="docutils footnote" frame="void" id="id3" rules="none">
275<colgroup><col class="label" /><col /></colgroup> 270<colgroup><col class="label" /><col /></colgroup>
276<tbody valign="top"> 271<tbody valign="top">
277<tr><td class="label"><a class="fn-backref" href="#id1" name="id3">[1]</a></td><td>GLEP 22, New &quot;keyword&quot; system to incorporate various 272<tr><td class="label"><a class="fn-backref" href="#id1" name="id3">[1]</a></td><td>GLEP 22, New &quot;keyword&quot; system to incorporate various
278userlands/kernels/archs, Goodyear, 273userlands/kernels/archs, Goodyear,
279(<a class="reference" href="http://glep.gentoo.org/glep-0022.html">http://glep.gentoo.org/glep-0022.html</a>)</td></tr> 274(<a class="reference" href="http://glep.gentoo.org/glep-0022.html">http://glep.gentoo.org/glep-0022.html</a>)</td></tr>
280</tbody> 275</tbody>
281</table> 276</table>
282<table class="footnote" frame="void" id="id4" rules="none"> 277<table class="docutils footnote" frame="void" id="id4" rules="none">
283<colgroup><col class="label" /><col /></colgroup> 278<colgroup><col class="label" /><col /></colgroup>
284<tbody valign="top"> 279<tbody valign="top">
285<tr><td class="label"><a class="fn-backref" href="#id2" name="id4">[2]</a></td><td>For the purpose of readability, we will refer to 1, 2 and 280<tr><td class="label"><a class="fn-backref" href="#id2" name="id4">[2]</a></td><td>For the purpose of readability, we will refer to 1, 2 and
2864-tuples, even though tuple in itself suggest a field consisting of 2814-tuples, even though tuple in itself suggest a field consisting of
287two values. For clarity: a 1-tuple describes a single value field, 282two values. For clarity: a 1-tuple describes a single value field,
288while a 4-tuple describes a field consisting out of four values.</td></tr> 283while a 4-tuple describes a field consisting out of four values.</td></tr>
289</tbody> 284</tbody>
290</table> 285</table>
291</div> 286</div>
292<div class="section" id="copyright"> 287<div class="section">
293<h1><a class="toc-backref" href="#id13" name="copyright">Copyright</a></h1> 288<h1><a class="toc-backref" href="#id13" id="copyright" name="copyright">Copyright</a></h1>
294<p>This document has been placed in the public domain.</p> 289<p>This document has been placed in the public domain.</p>
295</div> 290</div>
296</div>
297 291
292</div>
293<div class="footer">
298<hr class="footer" /> 294<hr class="footer" />
299<div class="footer">
300<a class="reference" href="glep-0047.txt">View document source</a>. 295<a class="reference" href="glep-0047.txt">View document source</a>.
301Generated on: 2006-02-12 19:57 UTC. 296Generated on: 2007-10-13 13:39 UTC.
302Generated by <a class="reference" href="http://docutils.sourceforge.net/">Docutils</a> from <a class="reference" href="http://docutils.sourceforge.net/rst.html">reStructuredText</a> source. 297Generated by <a class="reference" href="http://docutils.sourceforge.net/">Docutils</a> from <a class="reference" href="http://docutils.sourceforge.net/rst.html">reStructuredText</a> source.
298
303</div> 299</div>
304</body> 300</body>
305</html> 301</html>
306 302
Legend:
Removed from v.1.3  
changed lines
  Added in v.1.9
  ViewVC Help
Powered by ViewVC 1.1.13