|
|
LXR 真是难安装，搞了大半天才将其装上，而且还是有点糊里糊涂的，甚至一度想放弃不装了，下面记录一下过程。
安装的 LXR版本： 0.96 (截至发帖时为最新版)
安装当前版本 LXR 需要确保以下软件都已安装：
1) ctags
在 http://sf.net/projects/ctags 可以获得。
2) MySQL (4.x/5.x)
在 http://www.mysql.com 可以获得。
3)Apache ( Apache-1 或 Apache-2)
在 http://httpd.apache.org/ 可以获得。
4) Glimpse 或 swish-e
可以在 http://glimpse.cs.arizona.deu 可以获得 glimpse
可以在 http://swish-e.org/download/index.html 可以获得 swish-e
5) Perl 数据库驱动接口 DBI 以及相关DBD 驱动
6) Perl 文件 : MMagic 模块、mod_perl 模块
我的是 Mandriva 2009.1 ，在“安装/删除“软件里似乎找不到这个模块，只好上网找，在：
http://fr2.rpmfind.net/linux/RPM ... v2009.0.noarch.html
这里可以下载得到。
7) 如果需要 CVS 支持，还需要安装 rcs
至于安装上面这些软件，现在发行版的 Linux 里都有类似的傻瓜安装软件管理器，不需要手动编译，添加参数。
把下载的 lxr 文件包，解压缩到 /usr/local 下，默认解压缩后的文件夹包含版本号，但为了方便起见，将这个文件夹就命名为 lxr 。
下面步骤，基本按照 INSTALL 说明文档进行，有些步骤忽略掉，但不影响最后使用。
当相关软件都安装好，然后 cd 到 lxr 的目录下，登录 mysql ，默认的 mysql 用户 root 是不要密码的，这么登录：
登录好数据库后，执行命令：这条命令的意思是，执行 lxr 目录下的 initdb-mysql 脚本。其目的是创建相应的数据库以及用户名。所以，对于 mysql 用户，root 才有 GRANT 权限，也就是可以创建其他用户及为其分配相关权限。用下面命令看到了新建的数据库 lxr 的信息：mysql> show databases;
+--------------------+
| Database |
+--------------------+
| information_schema |
| lxr |
| mysql |
| test |
+--------------------+
4 rows in set (0.03 sec) mysql> show tables;
+------------------+
| Tables_in_lxr |
+------------------+
| lxr_declarations |
| lxr_files |
| lxr_indexes |
| lxr_releases |
| lxr_status |
| lxr_symbols |
| lxr_useage |
+------------------+
7 rows in set (0.01 sec)
现在，要把 LXR 用到的 perl 模块放到 mod_perl 能找到的地方。如果你用的是 Apache 1.x ，那么在说明文档上是这么做的：cp /usr/local/lxr/Local.pm /usr/lib/perl5/site_perl/
cp -r /usr/local/lxr/lib/LXR /usr/lib/perl5/site_perl
但是对于 Apache2 ，文档上说有 Apache2 对于外部模块有更聪明的处理办法，但没有具体说明。我对 Apache2 是不熟悉的，我也不知道这个聪明的办法是什么。于是根据 “要把 LXR 用到的 perl 模块放到 mod_perl 能找到的地方“ 这句话，在 /usr/local/perl5/ 目录下建立了一个 site_perl 目录( Mandriva 没有这个目录 )，然后按照上面说的，也把 Local.pm ，以及 LXR 目录搬到下面来。但是在后面的一连串试验中，似乎并不奏效，因为在浏览器里浏览那些 perl 文件时，浏览器却把这些源码都给原本不漏的当成纯文本的显示出来，基于此则推想，可能是 mod_perl 找不到这么些个 Local.pm 文件。于是，会过头来查看了一下 mod_perl 的安装路径：/usr/lib/apache-extramodules/mod_perl.so
/usr/lib/perl5/vendor_perl/5.10.0
/usr/lib/perl5/vendor_perl/5.10.0/i386-linux-thread-multi
... ...
为了方便，在 lxr 目录下建立一个 http 的目录，为的是放一些网页文件，于是执行命令：cp /usr/local/lxr/templates/* /usr/local/http/
最后，http 目录里被我放了一大堆文件：[beyes@localhost http]$ ls
diff* html-dir.html html-ident.html html-search-swish.html Local.pm lxr.css
find* html-find.html html-ident-refs.html html-tail.html lxr.conf search*
genxref* html-head.html html-search-glimpse.html ident* lxr.conf~ source*
上面所显示的文件，Local.pm ，genxref 文件也许不是必须的。但其它的都拷贝过去吧，少了会出问题，多了则不会：）
之后，在 lxr 目录下建立一个 http/lxr.conf 的一个软链接：ln -s /usr/local/lxr/http/lxr.conf /usr/local/lxr/lxr.conf
下面修改 lxr.conf 文件，具体修改就不列出，干脆把整个文件都呈现出来，以后也好比对：# -*- mode: perl -*-
# Configuration file
#
(
{
# Global configuration
# Path to glimpse executables.
# Define this OR the swish-e variable depending which search engine you want to use.
'glimpsebin' => '/usr/local/bin/glimpse',
'glimpseindex' => '/usr/local/bin/glimpseindex',
# Location of SWISH-E binary
#'swishbin' => '/usr/local/bin/swish-e',
# Path to Exuberant Ctags executable
'ectagsbin' => '/usr/bin/ctags',
# Place where lxr can write temporary files
'tmpdir' => '/tmp',
# Location of the Generic.pm config file
'genericconf' => '/usr/lib/perl5/vendor_perl/5.10.0/LXR/Lang/generic.conf',
# Location of the ectags.conf file
'ectagsconf' => '/usr/lib/perl5/vendor_perl/5.10.0/LXR/Lang/ectags.conf',
# Paths for CVS module
'cvspath' => '/bin:/usr/local/bin:/usr/bin:/usr/sbin',
},
{
# Configuration for http://192.168.1.3/lxr.
# baseurl is used to select configuration block.
'baseurl' => 'http://127.0.0.1/lxr', # Put your URL here
# baseurl_aliases allows other URLs to be used to reach the site
# comment this out if you do not want any aliases
'baseurl_aliases' => [ 'http://localhost/lxr', 'http://mydomain/lxr' ],
'virtroot' => '/lxr/', # The bit after the / above
'variables' => {
# Define typed variable "v". This is the list of versions to index.
'v' => {
'name' => '2.6.31.2',
# This can come from a file, a function or be explicitly
# ennumerated.
# From a file:
'range' => [ readfile('/usr/local/lxr/src/2.6.31.2') ],
# Explicitly:
# 'range' => [qw(v1 v2 v3.1 v4 experimental)],
# If files within a tree can have different versions,
# e.g in a CVS tree, 'range' can be specified as a
# function to call for each file:
#'range' => sub { return
# ($files->allreleases($LXR::Common::pathname),
# $files->allrevisions($LXR::Common::pathname))
# }, # deferred function call.
#
# When using a GIT repo, you can use its tags
# to create a list of versions. As an extra
# bonus, with the Linux::KernelSort Perl module,
# you can properly sort kernel versions. Just
# leave it off for generic projects.
#
# Simple example for any GIT repo:
# 'range' => sub {
# my $some_dir = "/path/to/project/.git/refs/tags";
# opendir (DIR, $some_dir) || die "cannot opendir $some_dir: $!";
# my @files = grep { -f "$some_dir/$_" } readdir (DIR);
# closedir DIR;
# return sort @files;
# },
#
# Advanced example for Linus's upstream
# linux-2.6 kernel repository:
# 'range' => sub {
# use Linux::KernelSort;
# my $kernel = new Linux::KernelSort;
#
# my $some_dir = "/path/to/linux-2.6/.git/refs/tags";
# opendir (DIR, $some_dir) || die "cannot opendir $some_dir: $!";
# my @files = grep { -f "$some_dir/$_" } readdir (DIR);
# closedir DIR;
# # Linus's tags all have a leading "v"
# # which we must cut off for sorting
# # add again afterwards.
# for (my $i = 0; $i < scalar (@files); $i++) {
# $files[$i] =~ s/^v//;
# }
# @files = $kernel->sort (@files);
# for (my $i = 0; $i < scalar (@files); $i++) {
# $files[$i] =~ s/^/v/;
# }
#
# return @files;
# },
# The default version to display
'default' => '2.6.31.2'
},
# Define typed variable "a". First value is default.
'a' => {
'name' => 'Architecture',
'range' => [qw(i386 alpha arm m68k mips ppc sparc sparc64)]
},
},
# These do funky things to paths in the system - you probably don't need them.
'maps' => {
'/include/asm[^\/]*/' => '/include/asm-$a/',
'/arch/[^\/]+/' => '/arch/$a/',
},
# Templates used for headers and footers
'htmlhead' => '/usr/local/lxr/http/html-head.html',
'htmltail' => '/usr/local/lxr/http/html-tail.html',
'htmldir' => '/usr/local/lxr/http/html-dir.html',
'htmlident' => '/usr/local/lxr/http/html-ident.html',
'htmlident_refs' => '/usr/local/lxr/http/html-ident-refs.html',
'htmlfind' => '/usr/local/lxr/http/html-find.html',
'htmlsearch' => '/usr/local/lxr/http/html-search-glimpse.html',
'sourcehead' => '/usr/local/lxr/http/html-head.html',
'sourcedirhead' => '/usr/local/lxr/http/html-head.html',
'stylesheet' => '/usr/local/lxr/http/lxr.css',
# sourceroot - where to get the source files from
# For ordinary directories, this specifies a directory which has each version as a
# subdirectory e.g.
# indexed-src/version1/...
# indexed-src/version2/...
# The names of the version directories must match the values for the Version
# variable above.
'sourceroot' => '/usr/local/lxr/src',
# Alternatively, this can specify a CVS repository by setting the value to "cvs:"
# followed by the path to the repository. Note this must be file accessible - remote
# server access does NOT work.
# 'sourceroot' => 'cvs:/hom/karsk/a/CVSROOT/linux',
# For a BitKeeper repository, specify bk: followed by the path to the
# repository. For BK it is also necessary to specify a 'cachepath' parameter in
# the 'sourceparams' value below. This should point to a directory where the
# code can write and read files.
# 'sourceroot' => 'bk:/some/repository/here',
# Finally, you can also use GIT, Linux Kernel's new
# distributes SCM to access the history:
# 'sourceroot' => 'git:/some/repository/.git'
# 'sourceroot' => 'git:/some/repostitory.git'
# The `sourceroot' config variable points to the directory
# in which you find the `objects', `refs', `index' etc.
# directories.
# The name to display for this source tree
'sourcerootname' => '/usr/local/lxr/src/$v',
#
# Any parameters to the source access method should be specified below.
#
# BK: 'cachepath' is used for temporary files
# GIT: if `do_annotations' is non-zero, the file revision that
# introduced each line is displayed.
# `do_blame' will add the author of each line to the
# output. Keep in mind that the core LXR code
# will only request author information if
# annotations are switched on, too!
#
'sourceparams' => {
'cachepath' => '/a/path/here',
'do_annotate' => 1,
'do_blame' => 1,
},
# The DBI identifier for the database to use
# For mysql, the format is dbi:mysql:dbname=<name>
# for Postgres, it is dbi:Pg:dbname=<name>
# for Oracle, it is dbi:Oracle:host=localhost;sid=DEVMMS;port=1521
'dbname' => 'dbi:mysql:dbname=lxr',
# If you need to specify the username or password for the database connection,
# uncomment the following two lines
# 'dbpass' => 'foo',
# 'dbuser' => 'lxr',
# If you need multiple lxr configurations in one database, set different table
# prefixes for them.
'dbprefix' => 'lxr_',
# For using glimpse, the directory to store the .glimpse files in is required
'glimpsedir' => '/usr/local/lxr/glimpse/databases',
# Location of swish-e index database files if using swish-e
#'swishdir' => '/a/directory/here/',
# where to look for include files inside the sourcetree. This is used to hyperlink
# to included files.
'incprefix' => [ '/include', '/include/linux' ],
# Which extensions to treat as images when browsing. If a file is an image,
# it is displayed.
'graphicfile' => '(?i)\.(gif|jpg|jpeg|pjpg|pjpeg|xbm|png)$', #'
# How to map files to languages
# Note that the string for the key and the first entry in the
# array MUST match
'filetype' => {
# Format is
# Language name, filepatten regexp, module to invoke,
# (optional )tabwidth
# Note that to have another language supported by Generic.pm,
# you must ensure that:
# a) exuberant ctags supports it
# b) generic.conf is updated to specify information about the language
# c) the name of the language given here matches the entry in generic.conf
'C' => [
'C', '\.c$|\.pc$' #'
, 'LXR::Lang::Generic', '8'
],
'C++' => [
'C++',
'\.C$|((?i)\.c\+\+$|\.cc$|\.cpp$|\.cxx$|\.h$|\.hh$|\.hpp$|\.hxx$|\.h\+\+$)' #'
, 'LXR::Lang::Generic', '8'
],
# Some languages are commented out until the relevant entries in generic.conf are made
# The list here is the set supported by ctags 5.0.1
# 'Beta' => [ 'Beta', '(?i)\.bet$', 'LXR::Lang::Generic' ],
# 'Eiffel' => [ 'Eiffel', '(?i)\.e$', 'LXR::Lang::Generic' ],
# 'Fortran' =>
# [ 'Fortran', '(?i)\.f$|\.for$|\.ftn$|\.f77$|\.f90$|\.f95$', 'LXR::Lang::Generic' ],
'Cobol' => [ 'Cobol', '(?i)\.cob$', 'LXR::Lang::Cobol' ],
'Java' => [ 'Java', '(?i)\.java$', 'LXR::Lang::Java', '4' ],
# 'Lisp' => [
# 'Lisp', '(?i)\.cl$|\.clisp$|\.el$|\.l$|\.lisp$|\.lsp$|\.ml$', 'LXR::Lang::Generic'
# ],
# No tabwidth specified here as an example
'Make' => [ 'Make', '(?i)\.mak$|(?i)\.mk$|makefile*', 'LXR::Lang::Generic' ],
# 'Pascal' => [ 'Pascal', '(?i)\.p$|\.pas$', 'LXR::Lang::Generic' ],
'Perl' => [ 'Perl', '(?i)\.pl$|\.pm$|\.perl$', 'LXR::Lang::Generic', '4' ],
'php' => [ 'php', '(?i)\.php$|\.php3$|\.phtml$', 'LXR::Lang::Generic', '2' ],
'Python' => [ 'Python', '(?i)\.py$|\.python$', 'LXR::Lang::Generic', '4' ],
# 'rexx' => [ 'rexx', '(?i)\.cmd$|\.rexx$|\.rx$', 'LXR::Lang::Generic' ],
# 'ruby' => [ 'ruby', '(?i)\.rb$', 'LXR::Lang::Generic' ],
# 'scheme' => [ 'scheme', '(?i)\.sch$|\.scheme$|\.scm$|\.sm$', 'LXR::Lang::Generic' ],
'shell' => [ 'shell', '(?i)\.sh$|\.bsh$|\.bash$|\.ksh$|\.zsh$', 'LXR::Lang::Generic' ],
# 's-Lang' => [ 's-Lang', '(?i)\.sl$', 'LXR::Lang::Generic' ],
'SQL' => [ 'SQL', '(?i)\.sql$|\.pks$|\.pkb$', 'LXR::Lang::Generic' ],
'VB' => [ 'VB', '(?i)\.bas$|\.cls$|\.ctl$|\.frm$|\.vbs$', 'LXR::Lang::Generic' ],
#'tcl' => [ 'tcl', '(?i)\.tcl$|\.wish$', 'LXR::Lang::Generic' ],
},
# Maps interpreter names to languages. The format is:
# regexp => langname
# regexp is matched against the part after #! on the first line of a file
# langname must match one of the keys in filetype above.
#
# This mapping is only used if the filename doesn't match a pattern above, so
# a shell script called shell.c will be recognised as a C file, not a shell file.
'interpreters' => {
'perl' => 'Perl',
'bash' => 'shell',
'csh' => 'shell',
'python' => 'Python',
'ksh' => 'shell',
'zsh' => 'shell',
'sh' => 'shell',
'ksh' => 'shell',
},
# a link of the form (prefix)($filepath)(postfix) is generated when viewing a file
# example for cvsweb:
#'cvswebprefix' => 'http://cvs.myhost.com/cgi-bin/cvsweb.cgi',
#'cvswebpostfix' => '?cvsroot=rootname',
# example for viewcvs:
#'cvswebprefix' => 'http://cvs.myhost.com/cgi-bin/viewcvs.cgi/myroot',
#'cvswebpostfix' => '',
# Directories to always ignore. These usually are the SCM's
# private directories which possibly may contain non-public
# project history.
'ignoredirs' => ['CVSROOT', '.git'],
}
)
建立源代码索引，用下面的命令：genxref --url=http://127.0.0.1/lxr --version=2.6.31.2
还需要把 lxr 目录下的 .htaccess-apache2 文件复制到 http 目录下，然后改名为 .htaccess ，按照网上说的，改了一下内容(说明文档上没有做修改，但这里还是改了)，完整内容如下：Options Indexes ExecCGI FollowSymlinks
order deny,allow
<Files lxr.conf>
deny from all
</Files>
<Files lib>
deny from all
</Files>
<Files ~ (find|search|source|ident|diff|cgi-bin)$>
SetHandler perl-script
PerlHandler ModPerl::RegistryPrefork
#PerlHandler ModPerl::Registry
<Files ~ (search|source|ident|diff|find)$>
SetHandler cgi-script
</Files>
ForceType text/html
PerlOptions +ParseHeaders
</Files>
这么做完了之后，还要再编辑一下 Apache 的 httpd.conf 文件，目的是要建立一下虚拟目录，修改的内容添加到文件最后，如下：#<VirtualHost *:80>
# ServerAdmin webmaster@dummy-host.example.com
# DocumentRoot /www/docs/dummy-host.example.com
# ServerName dummy-host.example.com
# ErrorLog /var/log/httpd/dummy-host.example.com-error_log
# CustomLog /var/log/httpd/dummy-host.example.com-access_log common
#</VirtualHost>
Alias /lxr/ /usr/local/lxr/http/
<Directory /usr/local/lxr/>
AllowOverride All
</Directory>
完后，重启一下 httpd 就可以使用了，截个图：
![]()
事实上，也许我的配置并不好，这个显示效果比
http://lxr.linux.no/
上的要差。
用 linux 大概就在于折腾吧。如果可以联网，就用：http://lxr.linux.no/ 这个吧，版本多，齐全，显示还更直观。 |
|
窥视卡
雷达卡
发表于 2009-10-21 19:36:00
收藏
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
千斤顶
显身卡